I recently wrote a detailed article about Large Agile Teams that was a detailed walkthrough of how to structure agile teams of various sizes. I suspect that this is the most comprehensive online discussion of this topic. The article addressed the following topics:
Organizing Agile Teams. The article starts with a summary of the results of some industry research that I've done regarding the size of agile teams, showing that agile techniques are in fact being successfully applied on a variety of team sizes. It then goes into detail describing the organization structure of agile teams at various sizes. The article starts with a discussion of small agile teams, covering the common rhetoric of how to organize such a team and then making observations about what actually happens in practice. It then walks through two approaches to organizing medium sized teams of 15 to 50 people - a structure for a single team and a structure for a team of teams. Finally, it walks through how to organize a large agile program of 50+ people, focusing a fair bit on the need for a leadership team to coordinate the overall activities within the program. This advice is similar to what is seen in the SAFe framework although proves to be a bit more flexible and pragmatic in practice.
Supporting Large Agile Teams. The leadership structure to support a large agile team is reasonably straightforward once you understand the issues that such a team faces. In this section the article overviews the need for three important sub teams within your overall leadership team: The Product Delivery Team, The Product Management Team, and The Architecture Team. It also describes the need for an optional Independent Testing/Integration Team, something misleadingly labeled an integration team in SAFe, that reflects some of the known agile testing and quality practices that I've been writing about for several years.
Organizing subteams. The article includes a detailed discussion for how to organize the work addressed by agile sub teams within a large agile program. These strategies include feature teams, component teams, and internal open source teams. As you would expect with the Disciplined Agile Delivery (DAD) framework, the article clearly summarizes the advantages and disadvantages of each approach on provides guidance for when (not) to apply each one. I suspect you'll find this portion of the article to be one of the most coherent discussions of the Feature vs. Component team debate.
Tailoring agile practices. The article provides a detailed overview of how the various DAD process goals are tailored to address the issues faced by large teams. This advice includes: Do a bit more up-front requirements exploration; Do a bit more up-front architectural modelling; Do a bit more initial planning; Adopt more sophisticated coordination activities; Adopt more sophisticated testing strategies; and Integrate regularly. My hope is that you find this part of the article very illuminating regarding how the DAD framework provides flexible and lightweight advice for tailoring your approach to address the context of the situation that you face.
Other Resources. The article ends with a collection of links to other resources on this topic.
I welcome any feedback that you may have about Large Agile Teams.
Modified by ScottAmbler
I was recently involved in an online discussion about how to calculate the benefits realized by software development teams. As with most online discussions it quickly devolved into camps and the conversation didn’t progress much after that. In this case there was what I would characterize as a traditional project camp and a much smaller agile/lean product camp. Although each camp had interesting points, the important thing for me in the conversation was the wide cultural and experience gap between the people involved in the conversation.
The following diagram summarizes the main viewpoints and the differences between them. The traditional project camp promoted a strategy where the potential return on investment (ROI) for a project would be calculated, a decision would be made to finance the project based (partly) on that ROI, the project would run, the solution delivered into production, and then at some point in the future the actual ROI would be calculated. Everyone was a bit vague on how the actual ROI would be calculated, but they agreed that it could be done although would be driven by the context of the situation. Of course several people pointed out that it rarely works that way. Even if the potential ROI was initially calculated it would likely be based on wishful thinking and it would be incredibly unlikely that the actual ROI would be calculated once the solution was in production. This is because few organizations are actually interested in investing the time to do so and some would even be afraid to do so. Hence the planned and actual versions of the traditional strategy in the diagram.
The agile/lean camp had a very different vision. Instead of investing in upfront ROI calculation, which would have required a fair bit of upfront requirements modelling and architectural modelling to get the information, the idea was that we should instead focus on a single feature or small change. If this change made sense to the stakeholders then it would be implemented, typically on the order of days or weeks instead of months, and put quickly into production. If your application is properly instrumented, which is becoming more and more common given the growing adoption of DevOps strategies, you can easily determine whether the addition of the new feature/change adds real value.
Cultural differences get in your way
The traditional project camp certainly believed in their process. In theory it sounded good, and I’m sure you could make it work, but in practice it was very fragile. The long feedback cycle, potentially months if not years, pretty much doomed the traditional approach to measuring benefits of software development to failure. The initial ROI guesstimate was often a work of fiction and rarely would it be compared to actuals. The cultural belief in bureaucracy motivated the traditional project camp to ignore the obvious challenges with their chosen approach.
The agile/lean camp also believed in their strategy. In theory it works very well, and more and more organizations are clearly pulling this off in practice, but it does require great discipline and investment in your environment. In particular, you need investment in modern development practices such as continuous integration (CI), continuous deployment (CD), and instrumented solutions (all important aspects of a disciplined agile DevOps strategy). These are very good things to do anyway, it just so happens that they have an interesting side effect of making it easy (and inexpensive) to measure the actual benefits of changes to your software-based solutions. The cultural belief in short feedback cycles, in taking a series of smaller steps instead of one large one, and in their ability to automate some potentially complex processes motivated the agile/lean camp to see the traditional camp as hopeless and part of the problem.
Several people in the traditional project camp struggled to understand the agile/lean approach, which is certainly understandable given how different that vision is compared with traditional software development environments. Sadly a few of the traditionalists chose to malign the agile/lean strategy instead of respectfully considering it. They missed an excellent opportunity to learn and potentially improve their game. Similarly the agilists started to tune out, dropping out of the conversation and forgoing the opportunity to help others see their point of view. In short, each camp suffered from cultural challenges that prevented them from coherently discussing how to measure the benefits of software development efforts.
How Should You Measure the Effectiveness of Software Development?
Your measurement strategy should meet the following criteria:
Measurements should be actioned. Both the traditional and agile/lean strategies described above meet this criteria in theory. However, because few organizations appear willing to calculate ROI after deployment as the traditional approach recommends, in practice the traditional strategy rarely meets this criteria. It is important to note that I used the word actioned, not actionable. Key difference.
There must be positive value. The total cost of taking a measure must be less than the total value of the improvement in decision making you gain. I think that the traditional strategy falls down dramatically here, which is likely why most organizations don’t actually follow it in practice. The agile/lean strategy can also fall down WRT this criterion but is much less likely to because the feedback cycle between creating the feature and measuring it is so short (and thus it is easier to identify the change in results to the actual change itself).
The measures must reflect the situation you face. There are many things you can measure that can give you insight into the ROI of your software development efforts. Changes in sales levels, how often given screen or function is invoked, savings incurred from a new way of working, improved timeliness of information (and thereby potentially better decision making), customer retention, customer return rate, and many others. What questions are important to you right now? What measures can help provide insight into those questions? It depends on the situation that you face, there are no hard and fast rules. For a better understanding of complexity factors faced by teams, see The Software Development Context Framework.
The measures should be difficult to game. Once again, traditional falls down here. ROI estimates are notoriously flakey because they require people to make guesses about costs, revenues, time frames, and other issues. The measurements coming out of your instrumented applications are very difficult to game because they’re being generated as the result of doing your day-to-day business.
The strategy must be compatible with your organization. Once again, this is a “it depends” type of situation. Can you imagine trying to convince an agile team to adopt the traditional strategy, or vice versa? Yes, you can choose to improve over time.
Not surprisingly, I put a lot more faith in the agile/lean approach to measuring value. Having said that, I do respect the traditional strategy as there are some situations where it may in fact work. Just not as many as traditional protagonists may believe.
A recurring discussion that I have with experienced agile developers is what it means to take a disciplined agile approach. The conversation usually starts off by some saying "but it already requires discipline to do agile", something that I fully agree with, followed by "therefore 'disciplined agile' is merely a marketing term", something which I don't agree with. The challenge with the "standard" agile discipline is that it is often focused on construction activities within a single project team, clearly important but also clearly not the full picture. There's more to an agile project than construction, and there's more to most IT departments than a single development project. In short, there are many opportunities for IT professionals to up their discipline, and thereby up their effectiveness, opportunities which we make explicit in the Disciplined Agile Delivery (DAD) framework.
Let's explore the many aspects to taking a disciplined agile approach:
You adopt "standard" agile discipline
. Aspects of agile which require discipline
include adopting practices such as test-driven development (TDD), active stakeholder participation, working collaboratively, shortening the feedback cycle
, and many more. These strategies are a great start to becoming disciplined IT professionals.
You take a goal-driven approach
. When we first started working on the DAD framework I didn't want to create yet another prescriptive framework, particularly given Rational's track record with the Rational Unified Process (RUP) framework. Rational has been pilloried for years for the prescriptive nature of RUP, which is unfortunate because there are a lot of great ideas in RUP that agile teams can benefit from, some of which we adopted in DAD and many of which are being actively reinvented with the agile community even as you read this. Furthermore, there are many prescriptive elements of the Scrum method that can get teams in trouble. For example, Scrum prescribes that you hold a daily stand up meeting, often called a Scrum meeting, where everyone should answer three questions. That's a great approach for teams new to agile, but it proves problematic in many situations due to it's prescriptive nature. Do you really need to do this once a day? I've been on teams where we held coordination meetings twice a day and others only once a week. Do you really need to stand up? I've been on geographically distrubited agile teams where many of us were sitting down during coordination calls. Do you really need to answer three questions, two of which are clearly focused on status regardless of claims otherwise? I've been on lean teams where we met around our Kanban board and focused on potential blockers. The answers to these questions depends on the context of the situation you find yourself in. The challenge, at least from the point of view of a process framework, is how do you avoid falling into the trap of being overly prescriptive. The strategy we adopted in DAD is to take a goal-driven approach. The observation is that regardless of the situation you find yourself in there are common goals your team will need to fulfill. For example, at the beginning of a project common goals include developing an initial plan, initially exploring the scope, initially identifying a technical strategy, and securing initial funding (amongst others). Throughout construction you should coordinate your activities, improve the quality of your ecosystem, and produce a potentially consumable solution on a regular basis (more on this below). So, instead of prescribing a daily stand up meeting the DAD framework instead indicates you should coordinate your activities, and gives several options for doing so (one of which is a Scrum meeting). More importantly DAD describes the advantages and disadvantages of your options so that you can make the choice that's best suited for the situation your team finds itself in (see this blog posting
for a detailed example of the types of tables included in the DAD book to help you through such process tailoring decisions). In short, our experience is that it requires discipline to take a goal driven approach
to agile delivery over the prescriptive strategies in other agile processes.
You take a context-driven approach
. There are many tailoring factors, which I describe in the Software Development Context Framework (SDCF)
, which you need to consider when making process, tooling, and team structure decisions. For example, a large team will adopt a different collection of practices and tools than a small team. A geographically distributed team will adopt a different strategy than a team that is co-located. You get the idea. Other tailoring factors include compliance, team culture, organization culture, technical complexity, domain complexity, and project type. It requires discipline to recognize the context of the situation you find yourself in and then act accordingly.
You deliver potentially consumable solutions
. One of the observations that we made early in the development of the DAD framework was that disciplined agile teams produce potentially consumable solutions, not just potentially shippable software. Although delivery of high-quality, working software is important it is even more important that we deliver high-quality working solutions to our stakeholders. For example, not only are we writing software but we may also be updating the hardware on which it runs, writing supporting documentation, evolving the business processes around the usage of the system, and even evolving the organizational structure of the people working with the system. In other words, disciplined agilists focus on solutions over software
. Furthermore, "potentially shippable" isn't sufficient: not only should it be shippable but it should also be usable and should be something people want to use. In other words it should be consumable (a concept DAD adopted from IBM's Outside In Development
). Minimally IT professionals should have the skills and desire to produce good software, but what they really need are the skills and desire to provide good solutions. We need strong technical skills, but we also need strong "
such as user interface design and process design to name just two.
The incremental delivery of potentially consumable solutions on an incremental basis requires discipline
to do successfully. DAD teams focus on repeatable results not repeatable processes
You are enterprise aware
. Whether you like it or not, as you adopt agile you will constrained by the organizational ecosystem, and you will need to act accordingly. It takes discipline to work with enterprise professionals such as enterprise architects, data admistrators, portfolio managers, or IT governance people who may not be completely agile yet, and have the patience to help them. It takes discipline to work with your operations and support staff in a DevOps
manner throughout the lifecycle, particularly when they may not be motivated to do so. It requires discipline to accept and potentially enhance existing corporate development conventions (programming guidelines, data guidelines, UI guidelines, ...). It requires discipline to accept that your organization has an existing technology roadmap that you should be leveraging, building out, and in some cases improving as you go. In short, enterprise awareness requires a level of discipline
not typically seen on many agile teams.
You adopt a full delivery lifecycle
. Empirically it is very easy to observe that at the beginning of an agile project there are some activities that you need to perform to initiate the project. Similarly at the end of the project there are activities that you need to perform to release the solution into production or the marketplace. The DAD process framework addresses the effort required for the full delivery effort, including project initiation, construction, and deployment. Our experience is that it requires discipline on the part of IT professionals to include explicit phases
for Inception/Initation, Construction, and Transition/Deployment and more importantly to focus the appropriate amount of effort on each. One danger of explicit phases is that you run the risk of taking what's known as a Water-Scrum-Fall
approach, a term coined by Dave West the person who wrote the forward for the DAD book, where you take an overly heavy/traditional approach to inception and transition in combination with a lighter agile approach to construction. Water-Scrum-Fall occurs because many organizations haven't made a full transition to agile, often because they think it's only applicable to construction. Our experience is that you can be very agile in your approach to inception and transition, experience we've built into the DAD framework. Having said that it clearly requires discipline to keep inception activities short
and similarly it requires discipline to reduce the "transition phase" to an activity
You adopt a wider range of roles
. An interesting side effect of adopting a full delivery lifecycle is that you also need to adopt a more robust set of roles. For example, the Scrum method suggests three roles - Scrum Master, Product Owner, and Team Member - a reflection of the Scrum lifecycle's construction focus. DAD suggests three primary roles - Team Lead, Product Owner, Team Member, Architecture Owner
, and Stakeholder - as well as five secondary roles which may appear at scale.
You embrace agile governance
. Governance establishes chains of responsibility, authority and communication in support of the overall enterprise’s goals and strategy. It also establishes measurements, policies, standards and control mechanisms to enable people to carry out their roles and responsibilities effectively. You do this by balancing risk versus return on investment (ROI), setting in place effective processes and practices, defining the direction and goals for the department, and defining the roles that people play with and within the department. It requires discipline to adopt an agile approach to governance
, and that's something built right into the DAD framework.
The basic idea behind DevOps
is that your development strategy and operations strategy should reflect one another, that you should strive to optimize the whole IT process. This implies that development teams should work closely with your operations staff to deliver new releases smoothly into production and that your operations staff should work closely with development teams to streamline critical production issues.
DevOps has its source in agile software development, and it is an explicit aspect of the Disciplined Agile Delivery (DAD)
process framework. As a result there is a collection of agile development strategies which enable effective DevOps throughout the agile delivery lifecycle. These strategies include:
- Initial requirements envisioning. Disciplined agile teams invest time at the beginning of the project to identify the high-level scope in a light-weight, collaborative manner. This includes common operations requirements such as the need to backup and restore data sources, to instrument the solution so that it can be monitored in real time by operations staff, or to architect the solution in a modular manner to enable easier deployment.
- Initial architecture envisioning. Disciplined agile teams will also identify a viable architectural strategy which reflects the requirements of their stakeholders and your organization’s overall architectural strategy (hence the need to work closely with your enterprise architects and operations staff). One goal is to ensure that the team is building (or buying) a solution which will work well with the existing operational infrastructure and to begin negotiating any infrastructural changes (such as deploying new technologies) early in the project. Another goal is to ensure that operations-oriented requirements are addressed by the architecture from the very start.
- Initial release planning. As part of release planning the disciplined agile team works closely with their operations group to identify potential release windows to aim for, any release blackout periods to avoid, and the need for operations-oriented milestone reviews later in the lifecycle (if appropriate).
- Active stakeholder participation. Disciplined agile teams work closely with their stakeholders, including both operations and support staff, all the way through the lifecycle to ensure that their evolving needs are understood.
- Continuous integration (CI). This is a common technical agile practice where the solution is built/compiled, regression tested, and maybe even run through code analysis tools. CI promotes greater quality which in turn enables easier releases into production.
- Parallel independent testing. For enterprise-class development or at scale, particularly when the domain or technology is very complex or in regulatory environments, disciplined agile team will find they need to support their whole team testing efforts with an independent test team running in parallel to the development team. These testing issues often include validation of non-functional requirements – such as security, performance, and availability concerns – and around production system integration. All of these issues are of clear importance for operations departments.
- Continuous deployment. With this practice you automate the promotion of your working solution between environments. By automating as much of the deployment effort as possible, and by running it often, the development team increases the chance of a successful deployment and thereby reduces the risk to the operations environment. Note that deployment into production is generally not automatic, as this is an important decision to be made by your operations/release manager(s).
- Continuous documentation. With this practice supporting documentation, including operations and support documentation, is evolved throughout the lifecycle in concert with the development of new functionality.
- Production release planning. This is the subset of your release planning efforts which focuses on the activities required to deploy into production.
- Production readiness reviews. There should be at least one review, performed by the person(s) responsible for your operations environment, before the solution is deployed into production. The more critical the system, the more product readiness reviews may be required.
- End-of-lifecycle testing. Minimally you will need to run your full automated regression test suite against your baselined code once construction ends. There may also be manual acceptance reviews or testing to be performed, and any appropriate fixing and retesting required to ensure that the solution is truly ready for production.
There’s more to it though than simply adopting some good practices. Your process must also embrace several supporting philosophies. The Disciplined Agile Delivery (DAD)
process framework not only adopts the practices listed above, and more, but it also promotes several philosophies which enable DevOps:
- Delivery teams should be enterprise aware, that they should work with people such as operations staff and enterprise architects to understand and work towards a common operational infrastructure for your organization.
- Operations and support people should be recognized as key stakeholders of the solution being worked on.
- The delivery team should focus on solutions over software. Software is clearly important, but we will often provide new or upgraded hardware, supporting documentation (including operations and support procedures), change the business/operational processes that stakeholders follow, and even help change the organizational structure in which our stakeholders work.
- Your process should include an explicit governance strategy. Effective governance strategies motivate and enable development teams to leverage and enhance the existing infrastructure, follow existing organizational conventions, and work closely with enterprise teams – all of which help to streamline operations and support of the delivered solutions.
For more detail about this topic, I think that you will find the article I wrote for the December 2011 issue of Cutter IT Journal entitled “Disciplined Agile Delivery and Collaborative DevOps
” to be of value.
In November 2011 Paul Gorans, the Accelerated Solution Delivery (ASD) practice lead in IBM GBS, and I ran an agile adoption survey
. The survey explored a range of issue, including the factors that appear to be associated with the success and failure of agile project teams. Paul wrote up his thoughts in his Agile State of the Art Survey
article on ibm.com and I did the same for Dr Dobb's Journal in Agile Success Factors
. This blog posting summarizes the results of the survey.
Factors which appear to accelerate agile adoption include:
- People are assigned to a single team
- Development teams have easy access to business expertise
- Development teams are organized for agile delivery (not traditional)
- Your organization has an agile support group/community of excellence
- Your organization is explicitly addressing barriers to agility
- There is executive sponsorship for agile
- Agile teams are measured on value creation
- Your organization's IT governance strategy includes an agile path
Factors which appear to decelerate agile adoption include:
- Agile teams are measured using traditional metrics
A few days later someone asked a series of questions that I thought would make an interesting blog posting, so here goes:How much of IBM's projects (in percentage) are agile at the moment?
I don’t have exact numbers, but I believe that 90%+ of our teams in SWG are applying agile techniques in practical ways that make sense for their projects. The primary goal is to be effective – in frequent releases, higher quality, and happy customers – not just agile. By the way, there is roughly 30,000 developers in SWG.Can all of IBM's projects work with an agile methodology?
It’s certainly possible, but it may not always make sense. Products that are in maintenance mode with few bugs or feature requirements may not benefit as much from agile practices -- those teams will likely continue to do whatever it is that they have been doing. Having said that, it's still highly desirable to apply agile techniques on maintenance projects.
Also, agile methods can be harder to use on some projects than others, for example, around hardware development. As a general rule, I believe that the majority of software projects can benefit from agile techniques. The primary determinant of whether a team can adopt agile techniques is culture and skill – not team size, the domain, or the degree of geographic distribution. That notion surprises many people who think that large agile teams
or geographically distributed agile teams
can’t succeed in adopting agile practices.Are agile projects sub-parts of large waterfall projects?
In some cases, that may happen. I’m sure it’s also true in reverse. We see many customers who are migrating from waterfall projects to a more agile way of doing things, and they often start this migration with smaller sub-projects. At IBM, we have tens of thousands of developers worldwide on hundreds of teams, so we have examples of pretty much any combination of agile, iterative, and traditional practices that you can imagine. There’s definitely not one size that fits all, which is a key aspect of the Disciplined Agile Delivery (DAD)
process framework.What do you think the impact of these numbers will be on the PM community?
The IBM PM community is embracing agile. And the reality is that a majority of development organizations around the world are moving to agile now as well (as much as 80% in some of the recent studies I’ve seen). I look forward to the increased adoption of agile methods by the PM community in general. The fact that PMI now offers an Agile Certified Practitioner training program certainly underscores the fact that agile practices are being adopted widely in the mainstream which is a great thing to see.
During the second week of August the Agile 2011 conference was held in Salt Lake City (SLC). As you likely know the Agile Manifesto was formulated 10 years ago in SLC so it was apropos to hold it there. There was some excitement around the 10 year anniversary of the manifesto, with a panel session with the 17 authors of it. Sadly there seemed to be little excitement around the efforts of the 10th anniversay agile workshop
in February which proposed a potential path forward for the agile community. I found the conference to be an evolutionary improvement over the conferences of the past few years, which is a very good thing because the focus since 2008 has moved beyond the "cool" new programming techniques to include the issues that enterprises face.
Starting at the Agile 2008 conference I've seen an uptick in interest in what I would consider some of the more mature topics in agile development, although I'm unfortunately still seeing significant confusion out there too, in part due to over-exuberence of people new to agile. For example, there's people still asking about basic issues about agile architecture
and agile database
techniques, although I was really happy to see more coherent discussions around scaling agile
. My own presentation about the Agile Scaling Model
was well attended and I suspect I opened a few people's eyes regarding the realities that we face (yes, there's a lot more to it than holding a "scrum of scrums", yeesh). We have a long way to go until people really start to understand scaling issues, but we're clearly on the path to getting there.
The conference show floor was interesting, with a wide range of vendors offering services and products focused on agile and lean. One thing that I noticed was many vendors had large monitors showing off their ability to support lean task boards, which for the most part they all looked the same. At the IBM booth we were showing off some of the Jazz tools
, in particular Rational Team Concert (RTC)
. For a long time now we've been giving away fully functional, with no time limit, licenses of RTC for teams of up to 10 people. Something worth checking out.
The Agile 201x conferences hosted by the Agile Alliance are always a good investment of your time and money, and Agile 2011 was no exception. See you at Agile 2012 in the great state of Texas!
My new white paper, Disciplined Agile Delivery: An Introduction
, is now available free of charge from IBM.com. The paper overviews the Disciplined Agile Delivery (DAD) process framework, a hybrid comprised of strategies from Scrum, XP, Agile Modeling, and other agile methods which is people first, learning oriented, and enterprise aware. DAD is the basis from which you can scale agile.
- Context counts - The Agile Scaling Model
- People first - People, and the way they interact with each other, are the primary determinant of success for a solution delivery project.
- Learning-oriented - The DAD process framework promotes the ideas that team members should collaborate closely and learn from each other, that the team should invest effort to learn from their experiences and evolve their approach, and that individuals should do so as well.
- Hybrid - DAD adopts and tailors strategies from Scrum, XP, Agile Modeling, UP, Kanban, and many others. It addresses many of the issues Mark Kennaley discusses in SDLC 3.0.
- IT solution focused - DAD teams produce potentially consumable solutions every construction iteration. This extends Scrum's "potentially shippable" strategy to explicitly address usability/consumability plus the fact that we're really delivering full solutions not just software.
- Goal-driven delivery life cycle - The DAD lifecycle is focused on delivery, not just construction. Furthermore it is goals-driven, the DAD process framework suggests various strategies to fulfill those goals but does not prescribe specific practices.
- Risk and value driven - The DAD lifecycle is risk and value driven. It extends Scrum's value-driven lifecycle which produces potentially shippable software each sprint/iteration so that it explicitly includes light-weight milesstones such as ensuring stakeholder consensus as to the scope of the project early in the lifecycle, proving the architecture with working code early in the lifecycle, ensuring sufficient functionality exists before transition, and ensuring production readiness before actual release of the solution.
- Enterprise aware - The DAD process framework promotes the ideas that DAD teams should work closely with their enterprise architecture groups to ensure they leverage and evolve the existing infrastructure, adopt and follow corporate guidelines, and work to the overall organizational vision. DAD teams are self organizing with appropriate governance.
There is a distinct rhythm, or cadence, at different levels of the agile process. We call this the agile 3C rhythm, for coordinate, collaborate, and conclude (which is sometimes called stabilize). The agile 3C rhythm occurs at three levels in Disciplined Agile Delivery (DAD):
- Day. A typical day begins with a short coordination meeting, called a Scrum meeting in the Scrum method. After the daily coordination meeting the team collaborates throughout most of the day to perform their work. The day concludes with a working build, hopefully you had several working builds throughout the day, which depending on your situation may require a bit of stabilization work to achieve.
- Iteration. DAD construction iterations begin with an iteration planning session (coordinate) where the team identifies a detailed task list of what needs to be done that iteration. Note that iteration modeling is often part of this effort. Throughout the iteration they collaborate to perform the implementation work. They conclude the iteration by producing a potentially consumable solution, a demo of that solution to key stakeholders, and a retrospective to identify potential improvements in the way that they work.
- Release. The DAD lifecycle calls out three explicit phases - Inception, Construction, and Transition – which map directly to coordinate, collaborate, and conclude respectfully.
The agile 3C rhythm is similar conceptually to Deming’s Plan, Do, Check, Act (PDCA) cycle:
- Coordinate maps to plan
- Collaborate maps to do
- Conclude maps to check and act
There is a fair bit of rhetoric surrounding agile methods, some of which we subscribe to and some of which we don’t. We’d like to briefly examine the rhetoric which we’ve found to be the most misleading for people trying to be effective at adopting agile techniques. The following list is in the format X but Y, where X is the rhetoric and Y is the strategy promoted by the Disciplined Agile Delivery (DAD) process framework. This includes:
- Requirements evolve throughout the lifecycle BUT the scope should still be agreed to at the beginning of the project. There has to be an initial vision for a project, a vision which your stakeholders should help define and then agree to, and to come to that vision you will need to perform some initial requirements envisioning. A list of high level features is part of this initial vision. Yes, the details are very likely to evolve over time but the fundamental goals of your project and scope of your effort needs to be defined early in your project. In a very small minority of situations you may not be able to get the right people together, either physically or virtually, to define the initial vision – this should be seen as a significant project risk.
- Simple designs are best BUT the architecture should be thought out early in the lifecycle. Too many developers interpret the advice to focus on simple designs to mean that they should build everything from scratch. Yet more often than not the simplest design is to take advantage of what is already there, and the best way to do that is to work closely with people who understand your existing technical infrastructure. Investing in a little bit of architectural envisioning early in the lifecycle enables your team to identify existing enterprise assets that you can leverage, to identify your architectural options, and to select what appears to be the best option available to you. The details will still emerge over time, and some decisions will be deferred until a later date when it’s more appropriate to make them, but the bottom line is that disciplined agilists think before they act.
- Teams should be self organizing BUT they are still constrained (and enhanced) by your organizational ecosystem. Intellectual workers, including IT professionals, are most effective when they have a say in what work they do and how they do it. IT professionals can improve their productivity by following common conventions, leveraging and building out a common “dev-ops” infrastructure, building towards a common vision, and by working to common business and technical visions. In short, disciplined agile professionals are "enterprise aware".
- Delivery teams don’t need prescriptive process definitions BUT they do need some high-level guidance to help organize their work. Individual IT professionals are typically highly-skilled and highly-educated people often with years of experience, and teams of such people clearly have a wide range of knowledge. As a result of this knowledge it is incredibly rare for such people to read detailed procedures for how to do their work. However, they often still require some high-level advice to help them to organize their work effectively. Teams can often benefit from techniques and patterns used by other teams and this knowledge sharing should be encouraged.
- IT professionals know what to do BUT they’re still not process experts. A decade ago the strategy was to provide detailed process advice to teams, but recently the pendulum has swung the other way to provide little or no defined process at all. Over the last few years there’s been a trend within the agile community to advise teams to define their own process so that it’s tailored to their own unique situation. While this clearly strokes people’s egos, it’s relatively poor advice for several reasons. First, although every team is in a unique situation there is significant commonality so having at least a high-level process framework from which to start makes sense. Second, although these teams have a wide range of knowledge it might not be complete, nor consistent, nor is it clear what the trade-offs are of combining all the really good techniques that people know about. There is significant benefit in having a flexible process framework such as DAD which shows how everything fits together.
- IT professionals should validate their own work to the best of their ability BUT they likely aren’t testing experts so therefore need help picking up the appropriate skills. The mantra in the agile community is to test often and test early, and better yet to test first. As a result agile teams have adopted a “whole team” approach where the development team does its own testing. This works when there are people on the team with sufficient testing skills and more importantly can transfer those skills to others. Minimally you will need to embed testers into your delivery teams, but you should also consider explicit training and mentoring of everyone on the team in testing and quality skills. You may find my agile testing and quality strategies article to be an interesting read.
- Disciplined agile teams work in an iterative manner BUT still follow a lifecycle which is serial over time. On any given day people on a DAD project team may be performing analysis, testing, design, programming, deployment, or a myriad of other activities and iterating back and forth between them. But, the DAD lifecycle includes three distinct phases which are performed in order. So, DAD is both iterative in the small but serial in the large.
Modified by ScottAmbler
IBM Rational recently published an update to my Agility@Scale e-book, which can be downloaded free of charge. The e-book is a 21 page, 2.3 meg PDF (sorry about the size, guess the graphics did it) . It overviews the Agile Scaling Model (ASM) (which has since been replaced by the Software Development Context Framework (SDCF) ), Disciplined Agile Delivery (DAD), the scaling factors of agility at scale, and ends with some advice for becoming as agile as you can be. In short it's a light-weight coverage of some of the things I've been writing about in this blog the past couple of years. Could be a good thing to share with the decision makers in your organization if they're considering adoption agile strategies.
When it comes to testing on agile projects it is common practice for agile teams to adopt a "whole team testing
" approach where the team itself does its own testing. To accomplish this agile teams will often embed testers in the development team. Programmers will work closely with the testers, often via non-solo development
strategies such as pair programming, to pick up their valuable testing skills. The testers will in turn pick up new skills from the programmers, and in effect both groups will move away from being just specialists (testers or programmers) to being what's called generalizing specialists
. Whole team testing can be very different from traditional approaches where programmers may do some testing, often unit testing of their own code, and then throw it over the wall to testers and quality assurance (QA) professionals for verification and validation.
The types of testing that the parallel independent test team performs may include:
Pre-production system integration testing. Does the solution work within your overall organizational ecosystem? Importantly, if this is one of several teams currently developing new solutions, does this team's solution work with what will be in production (including the work in progress of other teams) when they go to release? In mid-to-large organizations the only economical way to do this sort of testing is via an independent, centralized team.
. Although it's possible to do usability testing on the development team, the reality is that usability testing is a specialized skill that few people have (although could pick up via non-solo development). Furthermore, particularly for solutions with many potential users, you may want to invest in a usability testing
lab. This is a centralized resource, or an outsourced resource these days, which is shared across many teams.
. Security testing is also a specialized skill, albeit one well supported with sophisticated security testing tools such as the Rational Appscan
suite which can be included in your continuous integration (CI) strategy. Many organizations will centralize their security testing efforts.
Exploratory testing. The fundamental goal of exploratory testing is to discover where the solution breaks, as opposed to confirmatory testing which focuses on showing that the solution conforms to the requirements (this is the type of testing the development team typically focuses on). Exploratory testing is also a skill, a good one which everyone should strive to pick up, but exploratory testers are often few in number in many organizations. So, to leverage their skills effectively you may want to have some of them on the independent test team while they mentor others while doing so.
Non-functional testing. Non-functional requirements have a tendency to fall through the cracks on some development teams. Knowing this the independent test team will often "test to the risk" and focus on non-functional issues.
And much more. The above points are just exemplars, not an exact list. Please follow some of the links above for greater detail.
I'd like to leave you with several important thoughts:
The developers still do the majority of the testing. Just because there's an independent test team it doesn't imply that they are the ones doing all the testing. In fact, nothing could be further from the truth. They should be doing the minority of the testing effort, albeit the more difficult forms of it.
An independent test team will support multiple dev teams. For example, a test team of 5-6 people could support several development teams totalling 70 to 80 people. I typically look for a 15:1 or 20:1 ratio of developers to independent testers, hopefully even higher than that.
- You need to consider better tooling. Although the development team will still be using common agile testing tools such as the xUnit and FIT frameworks the independent test team (ITT) will need more sophisticated tooling. First, the ITT will need to be able to report defects back to the team easily. When the development team is using a Jazz-based tool such as Rational Team Concert (RTC) then this can easily be done using either RTC (the web interface may be sufficient) or another Jazz-enabled product such as Rational Quality Manager (RQM). Second, the ITT will likely need more sophisticated testing tools, such as Rational Appscan for static and dynamic security testing and Rational Performance Tester (RPT) for performance testing (just two of several software quality management tools you should consider).
Independent testing is economical. Although I listed several tools in my previous point (hey, I do work for a vendor after all) an "unfortunate" implication of my advice (unfortunate for IBM at least) is that you can reduce the number of licenses that you require and still get this critical testing done by centralizing their use.
It may be a bit more complicated in regulatory environments. In a strict regulatory environment the independent test team may need to repeat, or at least validate, the testing efforts of the development team. In regulatory environments my fundamental advice is always this -- Have practical people, including yourself, read and interpret the regulations. If you leave it to the bureaucrats you'll get a bureaucratic solution.
This is an important scaling technique. Parallel independent testing, when done in an agile manner, is an important technique which you should consider when scaling agile strategies to meet the uniques needs of the situation that you find yourself in.
Modified by ScottAmbler
One of the scaling factors called out in the Software Development Context Framework (SDCF) is domain complexity. The general idea is that agile teams will find themselves in different situations where some teams are developing fairly straightforward solutions, such as an informational website, whereas others are addressing very complex domains, such as building an air-traffic control system (ATCS). Clearly the team building an ATCS will work in a more sophisticated manner than the one building an informational website. I don't know whether agile techniques have been applied in the development of an ATCS, although I have to think that agile's greater focus on quality and working collaboratively with stakeholders would be very attractive to ATCS delivery teams, I do know that agile is being applied in other complex environments: The 2009 Agility at Scale Survey found that 18% of respondents indicated that their organizations had success at what they perceived to be very complex problem domains,.
Increased domain complexity may affect your strategy in the following ways:
Reaching initial stakeholder consensus becomes difficult. One of the risk reduction techniques called out in Disciplined Agile Delivery (DAD) is to come to (sufficient) stakeholder consensus at the beginning of the project during the Inception phase (called Sprint 0 in Scrum or Iteration 0 in other agile methods). Stakeholder consensus, or perhaps "near concensus" or "reasonable agreement" are better terms, can be difficult to come to the more complex the problem domain is because the stakeholders may not fully understand the implications of what they're making decisions about and because there is likely a greater range of stakeholders with differing goals and opinions. The implication is that your project initiation efforts may stretch out, increasing the chance that you'll fall back on the old habits of big requirements up front (BRUF) and incur the costs and risks associated with doing so.
Increased prototyping during inception. It is very common for disciplined agile teams to do some light-weight requirements envisioning during inception to identify the scope of what they're doing and to help come to stakeholder consensus. The greater the complexity of the domain, and particularly the less your team understands about the domain, the more likely it is that you'll benefit from doing some user interface (UI) prototyping to explore the requirements. UI prototyping is an important requirements exploration technique regardless of paradigm, and it is something that you should consider doing during both initial requirements envisioning as well as throughout the lifecycle to explore detailed issues on a just in time (JIT) manner.
Holding "all-hands reviews". One strategy for getting feedback from a wide range of people is to hold an "all hands review" where you invite a large group of people who aren't working on a regular basis with your team to review your work to date. This should be done occasionally throughout the project to validate that the input that you're getting from your stakeholder represenatives/product owners truly reflects the needs of the stakeholders which they represent. The 2010 How Agile Are You? Survey found that 42% of "agile teams" reported running such reviews.
Increased requirements exploration. Simple modeling techniques work for simple domains. Complex domains call for more complex strategies for exploring requirements. The implication is that you may want to move to usage scenarios or use cases from the simpler format of user stories to capture critical nuances more effectively. A common misunderstanding about agile is that you have to take a "user story driven approach" to development. This is an effective strategy in many situations, but it isn't a requirement for being agile.
The use of simulation. You may want to take your prototyping efforts one step further and simulate the solution. This can be done via concrete, functional prototypes, via simulation software, via play acting, or other strategies.
Addition of agile business analysts to the team. Analysis is so important to agile teams we do it every day. In situations where the domain is complex, or at least portions of the domain is complex, it can make sense to have someone who specializes in exploring the domain so as to increase the chance that your team gets it right. This is what an agile business analyst can do. There are a few caveats. First, even though the domain is complex you should still keep your agile analysis efforts as light, collaborative, and evolutionary as possible. Second, this isn't a reason to organize your team as a collection of specialists and thereby increase overall risk to your project. The agile analyst may be brought on because their specialized skills are required, but the majority of the people on the team should still strive to be generalizing specialists. This is also true of the agile analyst because their may not be eight hours a day of valuable business analysis work on the team, and you don't want the BA filling in their time with needless busy work.
The important thing is to recognize that the strategies which work well when you're dealing with a simple domain will not work well for a complex domain. Conversely, techniques oriented towards exploring complex domains will often be overkill for simple domains. Process and tooling flexiblity is key to your success.
One of the scaling factors
of the Agile Scaling Model (ASM)
is technical complexity.
The fundamental observation is that the underlying technology of solutions varies and as a result your approach to developing a solution will also need to vary.
It’s fairly straightforward to achieve high-levels of quality if you’re building a new system from scratch on a known technology platform, but not so easy when there are several technologies, the technologies are not well known, or legacy assets are involved.
There are several potential technical complexities which a Disciplined Agile Delivery (DAD) team may face:
- New technology platforms. Your team may choose to work with a technology platform which is either new to the team or sometimes even new to the industry. In the past few years new technology platforms include the Android operating system, Apple’s iPad platform, and various cloud computing (http://www.ibm.com/ibm/cloud/) platforms. Working with these platforms may require you to adopt new development tools and techniques, not to mention the need to train and mentor your staff in their usage. Furthermore, your team may need to allocate time for architectural spikes to explore how to use the new technology and to prove the overall architecture with working code early in the project lifecycle (this is a DAD milestone).
- Multiple technology platforms. IT solutions often run on multiple platforms. For example, a system’s user interface (UI) could run in a browser, access business logic implemented using J2EE on Websphere which in turn invokes web services implemented in COBOL running on a Z-series mainframe, and stores data in an Oracle database, a DB2 database, and in several XML files. Implementing new business functionality, or updating existing functionality, could require changes made on several of these platforms in parallel. The implication is that you’ll need to adopt tools and strategies which enable your team to develop, test, and deploy functionality on all of these platforms. Testing and debugging in particular will become more difficult as the number of technology platforms increases, potentially requiring you to adopt the practice of parallel independent testing. The Agility at Scale survey found that 34% of respondents indicated that their agile teams were working with multiple technology platforms.
- Legacy data. IT solutions should leverage existing, legacy data wherever possible to reduce the number of data sources and thereby increase data quality within your organization. Also, using existing data sources can potentially speed up development, assuming your team has a good relationship with the owners of the legacy data sources (sadly, this often isn’t the case as the Data Management Survey found). Working with legacy data sources may require improved database regression testing, practices, database refactoring practices, and agile approaches to data administration. The Agility at Scale survey found that 42% of respondents indicated that their agile teams were working with legacy data sources (personally, I’m shocked that this figure is so low, and fear that many agile teams are contributing to data quality problems within their organization as a result).
- Legacy systems. There are several potential challenges with legacy systems. First, the code quality may not be the best either because it was never really that good to begin with or because it’s degraded over the years as multiple people worked with it. You know you’ve got a quality problem if you’re either afraid to update the code or if when you do so you have to spend a lot of time debugging and then fixing problems revealed when doing the update. If the legacy system is a true asset for your organization you will want to pay off some of this technical debt by refactoring the code to make it of higher quality. Second, you may not have a full regression test suite in place, making it difficult to find problems when you do update the code let alone when you refactor it. Third, your development tools for your legacy code may be a bit behind the times. For example, I often run across mainframe COBOL developers still working with basic code editors instead of modern IDEs such as Rational Developer for System Z. Some of the strategies to deal effectively with legacy systems are to adopt a modern development toolset if you haven’t already done so (better yet, if possible adopt a common IDE across platforms and thereby reduce overall licensing and support costs) and to adopt agile practices such as static code analysis, dynamic software analysis, and continuous integration (CI). The Agile Project Initiation Survey found that 57% of respondents were integrating their new code with legacy systems and 51% were evolving legacy systems.
- Commercial off-the-shelf (COTS) solutions. COTS solutions, also called package applications, can add in a few complexities for agile teams. The packages rarely come with regression test suites, they often have rules about what you can modify and what you shouldn’t (rules that are ignored at your peril), and they’re often architected with the assumption that they’re the center of the architectural universe (which is a valid assumption if they’re the only major system within your organization). As I describe in my article Agile Package Implementations it is possible to take an agile approach to COTS implementations, although it may require a significant paradigm shift for the people involved. The Agility at Scale survey found that 15% of respondents indicated that their agile teams were working with COTS solutions.
- System/embedded solutions. For the sake of simplicity, if your team is developing a solution with both hardware and software aspects to it then you’re a systems project. Embedded systems are a specialization where the system has a few dedicated functions often with real-time constraints. Bottom line is that systems/embedded projects are typically more challenging than software-only projects – it gets really interesting when laws of physics starts to kick in, such as when you’re building satellites or space probes. I highly suggest Bruce Douglass’s book Real-Time Agility if you are interested in taking an agile approach to systems/embedded solution delivery.
The technical complexity faced by a project team is contextual – Working with four technology platforms is straightforward for someone used to dealing with seven, but difficult for someone used to dealing with just one. Recommended Reading:
Modified by ScottAmbler
One of the scaling factors called out in the Software Development Context Framework is “geographic distribution". As with the other scaling factors the level of geographic distribution is a range, with co-located teams at one extreme and far-located at the other. When your team is co-located the developers and the primary stakeholders are all situated in the same work room. If you have some team members in cubicles or in separate offices then you're slightly distributed, if you're working on different floors in the same building you're a bit more distributed, if you're working in different buildings within the same geographic area (perhaps your team is spread across different office buildings in the same city or some people work from home some days) then your team is more distributed, if people are working in different cities in the same country you're more distributed, and finally if people are working in different cities around the globe you're even more distributed (I call this far located).
As your team becomes more distributed your project risk increases for several reasons:
Communication challenges. The most effective means of communication between two people is face-to-face around a shared sketching space such as a whiteboard, and that requires you to be in the same room together. As you become more distributed you begin to rely on less effective communication strategies.
Temporal challenges. When people are in different time zones it becomes harder to find common working times, increasing the communication challenges. One potential benefit, however, is the opportunity to do "follow-the-sun" development where a team does some work during their workday, hands off the work to another team in a significantly different time zone, who picks up the work and continues with it. This strategy of course requires a high degree of sophistication and discipline on the part of everyone involved, but offers the potential to reduce overall calendar time.
Cultural challenges. As the team becomes more distributed the cultural challenges between sites typically increases. Different cultures have different work ethics, treat intellectual property differently, have different ideas about commitment, have different holidays, different approaches to things, and so on.
As you would imagine, because the project risk increases the more distributed your team is, the lower the average success rates of agile projects decrease as they become more distributed. The 2008 IT Project Success Survey found that co-located agile teams has an average success rate of 79%, that near located teams (members were in same geographic area) had a success rate of 73%, and that far-located agile teams had a success rate of 55%. The success rate decreases similarly for project teams following other paradigms.
The practices that you adopt, and the way that you tailor the agile practices which you follow, will vary based on the level of geographic distribution of your team. For example, a co-located team will likely do initial architecture envisioning on a whiteboard and keep it at a fairly high-level. A far-located team will hopefully choose to fly in key team members at the beginning of the project, at least the architecture owners on the various sub-teams, to do the architecture envisioning together. They will likely go into greater detail because they will want to identify, to the best of their ability, the interfaces of the various subsystems or components which they'll be building.
Interestingly, the Agility at Scale 2009 survey found that it was quite common for agile teams to be geographically distributed in some manner:
45% of respondents indicated that some of their agile teams were co-located
60% of respondents indicated that some of their agile teams had team members spread out through the same building
30% of respondents indicated that some of their agile teams were working from home
21% of respondents indicated that some of their agile teams had people working in different offices in the same city
47% of respondents indicated that some of their agile teams had team members that were far located
The bottom line is that some organizations, including IBM, have been very successful applying agile techniques on geographically distributed teams. In fact, agile GDD is far more common than mainstream agile discussion seem to let on.
People who are new to agile are often confused about how agile teams address architecture, but luckily we're seeing more discussion around agile architecture
now in the community so this problem is slowly being addressed from what I can tell. But, what I'm not seeing enough discussion about, at least not yet, is how is enterprise architecture addressed in the overall agile ecosystem. So I thought I'd share some thoughts on the subject, based on both my experiences over the years (see the recommended resources at the bottom of this posting) as well as on an enterprise architecture survey
which I ran in January/February 2010.
My belief is that effective enterprise architecture, particularly in an agile environment, is:
- Business driven. Minimally your EA effort should be driven by your business, not by your IT department. Better yet it should be business owned, although this can be a challenge in many organizations because business executives usually aren't well versed in EA and view it as an IT function. Yes, IT is clearly an important part of EA but it's not the entirety of EA nor is it the most critical part. In many organizations the IT department initiates EA programs, typically because the business doesn't know to do so, but they should quickly find a way to educate the business in the need to own your organization's EA efforts.
- Evolutionary. Your enterprise architecture should evolve over time, being developed iteratively and introduced incrementally over time. An evolutionary approach enables you act on the concrete feedback that you receive when you try to actually implement it, thereby enabling you to steer its development successfully.
- Collaborative. The EA survey clearly pointed to "people issues" being critical determinants of success, and of failure, of EA programs. My experience is that the best enterprise architects, just like the best application architects, work closely with the intended audience of their work, both on the business side of things as well as on the IT side. They will "roll up their sleeves" and become active members of development teams, often in the role of Architecture Owner on agile teams or Architect on more traditional teams. Their mission is to ensure that the development teams that they work with leverage the EA, to mentor developers in architecture skills, and to identify what works and what doesn't in practice so that they can evolve the EA accordingly. Enterprise architects, architects in general, who don't participate actively on development teams (holding architecture reviews isn't active participation) run the risk of being thought of as "ivory tower" and thus easy to ignore.
- Focused on producing valuable artifacts. The most valuable artifacts are useful to the intended audience, are light weight, and ideally are executable. Many EA programs run aground when the enterprise architects focus on artifacts that they've always wanted but that development teams really aren't very excited about -- yes, it might be interesting to have a comprehensive comparison of cloud technologies versus mainframe technologies, but a collection of reusable services would be fare more interesting to them. A detailed enterprise data model indicating suggested data attributes would be intellectually interesting to develop, but a list of legacy data sources with a high-level description of their contents would be immediately valuable to many development teams. A detailed model depicting desired web services would be useful, but an actual collection of working services that I can reuse now would be even better.
- An explicit part of development. In Disciplined Agile Delivery (DAD) architectural activities are an explicit part of the overall delivery process. Part of the architectural advice is that delivery teams should work closely with their organization's enterprise architects so that they can leverage the common infrastructure, and sometimes to help build it out, effectively. Disciplined agile teams realize that they can benefit greatly by doing so.
The Agile Scaling Model (ASM)
calls out addressing enterprise disciplines, such as enterprise architecture, as one of eight scaling factors which may apply to a given project. The interesting thing about this scaling factor is that it's the only one where things get potentially easier for development teams when we move from the simple approach, having a project focus, to the more complex approach, where we have an enterprise focus. By having a common infrastructure to build to, common guidelines to follow, and valuable artifacts to reuse project teams can benefit greatly. So, I guess my advice is to seriously consider adding enterprise disciplines to your agile strategy.Recommended Resources:
My new paper Scaling Agile: An Executive Guide
is now available. As the title suggests the paper overviews how to scale agile strategies to meet your organization's unique needs. The executive summary:
Agile software development is a highly collaborative, quality-focused approach to software and systems delivery, which emphasizes potentially shippable working solutions produced at regular intervals for review and course correction. Built upon the shoulders of iterative development techniques, and standing in stark contrast to traditional serial or sequential software engineering methods, agile software delivery techniques hold such promise that IBM has begun to adopt agile processes throughout its Software Group, an organization with over 25,000 developers. But how can practices originally designed for small teams (10-12) be “scaled up” for significantly larger operations? The answer is what IBM calls “agility@scale.”
There are two primary aspects of scaling agile techniques that you need to consider. First is scaling agile techniques at the project level to address the unique challenges individual project teams face. This is the focus of the Agile Scaling Model (ASM).
Second is scaling your agile strategy across your entire IT department, as appropriate. It is fairly straightforward to apply agile on a handful of projects, but it can be very difficult to evolve your organizational culture and structure to fully adopt the agile way of working.
The Agile Scaling Model (ASM) defines a roadmap for effective adoption and tailoring of agile strategies to meet the unique challenges faced by a software and systems delivery team. Teams must first adopt a disciplined delivery lifecycle
that scales mainstream agile construction techniques to address the full delivery process, from project initiation to deployment into production. Then teams must determine which agile scaling factors
– team size, geographical distribution, regulatory compliance, domain complexity, organizational distribution, technical complexity, organizational complexity, or enterprise discipline, if any — are applicable to a project team and then tailor their adopted strategies accordingly to address their specific range of complexities.
When scaling agile strategies across your entire IT organization you must effectively address five strategic categories — the Five Ps of IT
: People, principles, practices, process, and products (i.e., technology and tooling). Depending on your organizational environment the level of focus on each area will vary. What we are finding within many organizations, including IBM, is that the primary gating factor for scaling agile across your entire organization is your organization’s ability to absorb change.
Modified by ScottAmbler
An imporant step in scaling your agile strategy is to adopt a Disciplined Agile Delivery (DAD)
approach instead of one which is just focused on agile construction. One aspect of adopting a DAD approach it to mature your focus from just producing software to instead providing a solution which meets the needs of its stakeholders within the appropriate economic, cultural, and technical constraints. The fundamental observation is that as IT professionals we do far more than just develop software. Yes, this is clearly important, but in addressing the needs of our stakeholders we will often:
Provide new or upgraded hardware
Change the business/operational processes which stakeholders follow
Change the organizational structure in which our stakeholders work
Update supporting documentation
And yes, develop high-quality software
Although delivery of high-quality, working software is important it is even more important that we deliver high-quality working solutions to our stakeholders. Minimally IT professionals should have the skills and desire to produce good software, but what they really need are the skills and desire to provide good solutions. We need strong technical skills, but we also need strong "soft skills" such as user interface design and process design to name just two.
The shift to a solution-oriented focus from a software-oriented focus requires your agile teams to address some of the software-oriented prejudices which crept into the Agile Manifesto
. The people who wrote the manifesto (which I fully endorse) were for the most part software developers, consultants, and in many cases both. It is little wonder that this group would allow a bias towards software development creep into the language of their manifesto.
My January 2010 DDJ Agile Update, Tragic Mistakes When Adopting Test Driven Development (TDD)
, is now online. In the article I summarize what I consider to be common, and tragic, mistakes that I'm seeing organizations make when they attempt to adopt TDD.
These mistakes include:
The article also goes into potential benefits of TDD as well as potential challenges that you're face when adopting it.
activities are evolutionary (iterative and incremental) and highly collaborative in nature. Initially requirements are explored at a high level via requirements envisioning
at the beginning of the project and the details are explored on a just-in-time (JIT) basis via iteration modeling
and model storming
activities. The way that you perform these agile practices, and the extent to which you do so, depends on the situation in which a project team finds itself. The Agile Scaling Model (ASM)
is a contextual framework for effective adoption and tailoring of agile practices to meet the unique challenges faced by a system delivery team of any size. To see how this works, let's apply the concepts of the ASM to see how we would scale our agile approach to requirements.
First, let's consider how a small, co-located team would work. The first two categories of the ASM are core agile development and disciplined agile delivery
, the focus of both are small co-located teams in a fairly straightforward situation. In these situations simple techniques such as user stories
written on index cards and sketches on whiteboards
work very well, so the best advice that I can give is to stick with them. Some teams will take a test-driven development
(TDD) approach where they capture their requirements and design in the form of executable specifications
, although this sort of strategy isn't as common as it should be (yet!), likely because of the greater skill and discipline that it requires. Traditionalists often balk at this approach, believing that they need to document the requirements in some manner. But, for a small co-located team working in a collaborative manner, requirements documentation proves to be little more than busy work, often doing nothing more than justifying the existence of a business analyst who hasn't made the jump to agile yet. Don't get me wrong, there are good reasons to write some requirements documentation, and we'll see this in a minute, but you should always question any request for written specifications and try to find more effective ways to address the actual goal(s) motivating the request. Never forget that written documentation
is the least effective communication
option available to you.
Although inclusive tools
such as whiteboards and paper work well for requirements, for development activities you will need electronic tools. You will either put together an environment from point-specific tools or adopt something more sophisticated such as IBM Rational Team Concert (RTC)
which is already fully integrated and instrumented. RTC is a commercial tool, but luckily you can download a 10-license environment free of charge, which is just perfect for a small team. Larger teams, of course, will need to purchase licenses. One of the things that a disciplined agile delivery approach adds to core agile development is it addresses the full delivery life cycle, which is important because it explicitly includes pre-construction activities such as requirements envisioning. The first step in scaling agile techniques is to adopt a full delivery life cycle which covers the full range of activities required to initiate a project, produce the solution, and then release to solution to your end users.
More interesting is the third category of the ASM, Agility@Scale, and how its eight agile scaling factors
affect the way that you tailor your process and tooling strategy. Let's explore how each one could potentially affect your agile requirements strategy:
- Geographical distribution. The majority of agile teams are distributed in some manner -- some people are working in cubicles or private offices, on different floors, in different buildings, or even in different countries -- and when this happens your communication and coordination risks goes up. To counter this risk you will need to perform a bit more requirements envisioning up front to help ensure that everyone is working to the same vision, although this doesn't imply that you need to write detailed requirements speculations which would dramatically increase the risk to your project. Remember, agilists do just barely enough modeling and are prepared to iteratively elicit the details when they need to do so. The more distributed the team is the more likely they will need to adopt software-based requirements modeling tools such as IBM Rational Requirements Composer (RRC) which supports streamlined, agile requirements elicitation throughout the delivery life cycle. Index cards and whiteboards are great, but they're difficult to see if you're outside the room where they're posted. I've written a fair bit about distributed agile development in this blog.
- Team size. Some organizations, including IBM, are successfully applying agile techniques with teams of hundreds of people. A team of one hundred people will naturally work much differently than a team of ten people, or of one thousand people. Large teams are organized into collections of smaller teams, and the requirements for the overall project must be divvied up somehow between those teams. The implications are that as the team size grows you will need to invest a bit more time in initial requirements envisioning, and in initial architecture envisioning for that matter; you will need to use more sophisticated tools; and may need to use more sophisticated modeling techniques such as use cases and functional user interface prototypes. See large agile teams for more advice.
- Compliance requirement. When regulatory issues – such as Sarbanes Oxley, ISO 9000, or FDA CFR 21 – are applicable you are likely going to be required to capture requirements specifications in some manner and to enact traceability between those requirements. However, I highly recommend that you read the actual regulations yourself and don't let bureaucrats interpret them for you (doesn't it always seem that their interpretation always results in an onerous, documentation heavy solution?) because I have yet to run into a regulation which required you to work in an ineffective manner. Managing your requirements as work items in RTC can often more than meet your regulatory requirements for documentation and traceability, although you may want to consider a tool such as IBM Rational RequisitePro for complex regulatory situations.
- Domain complexity. The manner in which you elicit requirements for a data entry application or an informational web site will likely be much simpler than for a bio-chemical process monitoring or air traffic control system. More complex domains will require greater emphasis on exploration and experimentation, including but not limited to prototyping, modeling, and simulation. Although user stories may be effective as a primary requirements artifact in simple domains, in more complex domains you are likely to find that you need to drive your requirements effort with more sophisticated modeling techniques.
- Organization distribution. Sometimes a project team includes members from different divisions, different partner companies, or from external services firms. In these cases, particularly where the work is strictly organized between the various organizations (perhaps for security concerns), you may need a more sophisticated approach to managing the requirements. RTC enables you to organize the requirements between teams, and then to automatically track progress in real time via the RTC project dashboard.
- Technical complexity. The technical complexity of a solution can vary widely, from a single platform silo application to a multi-platform application working with legacy systems and data to a full-blown systems engineering effort. Complex technical domains, just like complex business domains, require more complex strategies for requirements elicitation and management. The requirements for your legacy systems are likely to have been captured using tools and techniques appropriate for that platform, for example the requirements for your COBOL application may have been captured using data flow diagrams and data models, whereas the requirements for your Java legacy application where captured using UML diagrams. The subteam working on the COBOL system might be using IBM Rational Application Developer (RAD) and RTC for Z whereas the Java subteam may use Eclipse with RTC. Because systems engineering projects can stretch on for years, particularly when the hardware is being developed in parallel to the software, sophisticated tooling such as IBM Rational DOORS is often used in these situations. For more information about systems engineering, see the IBM Rational Harmony process.
- Organizational complexity. Your approach to requirements elicitation and management will be affected by a host of organizational complexities, including your corporate culture. When the culture is flexible and collaborative you can be very agile in your approach to requirements, but as it becomes more rigid you become more constrained in what is considered acceptable and thus take on greater project risk. For example, many organizations still struggle with their approach to funding projects, often demanding that the project team provides an "accurate" estimate up front to which they will be held to. This in turn motivates risky behavior on the part of the development, including a "big requirements up front (BRUF)" approach where a detailed requirements speculation is developed early in the project. This is just one example of how questionable corporate culture can impact the way in which an agile team works.
- Enterprise discipline. Some organizations have enterprise-level disciplines, such as enterprise architecture, enterprise business modeling, strategic reuse, and portfolio management in place. These disciplines can easily be agile and from what I can tell the more successful efforts appear to lean more towards the agile end of the spectrum rather than the traditional end. Having an enterprise business modeling effort underway will affect your project-level requirements strategy -- you'll be able to leverage existing models, have access to people who understand the domain at an enterprise level, and will likely need to map your project efforts back to your enterprise models. The enterprise modelers will likely be using tools such as IBM Rational System Architect or IBM Websphere Business Modeler.
It is important to note that the way that you tailor the agile practices that you follow, and the tools that you use, will reflect the situation that you find yourself in. In other words, you need to right size your process and the Agile Scaling Model (ASM) provides the context to help you do so. As you saw above, in simpler situations you will use the simpler tools and techniques which are commonly promoted within the core agile development community. But, when things become a bit more complex and one or more of the scaling factors applies you need to modify your approach -- just don't forget that you should strive to be as agile as you can be given the situation that you find yourself in.
On Tuesday, Dec 1, 2009 Philippe Kruchten
, Bruce MacIsaac
, and myself participated on two virtual panels about the future of the Unified Process (we did two to support callers from around the globe) for the Global Rational User's Group (GRUG)
. During the panel sessions we discussed a bit of the history of the Unified Process, some of the misconceptions people have with it, some of the common mistakes people made implementing it (instantiating it to be documentation heavy and/or serial) due to those misconceptions, how it can be very agile if you choose to instantiate it that way, the OpenUP
, the AUP
, how UP relates to the IBM Practices
, and other topics.
The links to the recordings are:
Hope you find it interesting. As I've written in the past, the RUP can be as agile as you want to make it. Furthermore, there are a lot of really good ideas in the RUP that the agile community can and should choose to mine, although sadly I see far too many teams doing things the hard way and reinventing the process wheel on their own. I hope they're enjoying themselves, because it clearly isn't a very efficient way for them to go about process improvement.
The Agile Scaling Model (ASM) is a contextual framework for effective adoption and tailoring of agile practices to meet the unique challenges faced by a system delivery team of any size.
The ASM distinguishes between three scaling
- Core agile development. Core agile methods, such as Scrum and Agile Modeling, are self governing, have a value-driven system development lifecycle (SDLC), and address a portion of the development lifecycle. These methods, and their practices, such as daily stand up meetings and requirements envisioning, are optimized for small, co-located teams developing fairly straightforward systems.
- Disciplined agile delivery. Disciplined agile delivery processes, which include Dynamic System Development Method (DSDM) and Open Unified Process (OpenUP), go further by covering the full software development lifecycle from project inception to transitioning the system into your production environment (or into the marketplace as the case may be). Disciplined agile delivery processes are self organizing within an appropriate governance framework and take both a risk and value driven approach to the lifecycle. Like the core agile development category, this category is also focused on small, co-located teams delivering fairly straightforward systems. To address the full delivery lifecycle you need to combine practices from several core methods, or adopt a method which has already done so.
- Agility at Scale. This category focuses on disciplined agile delivery where one or more scaling factors are applicable. The eight scaling factors are team size, geographical distribution, regulatory compliance, organizational complexity, technical complexity, organizational distribution, domain complexity, and enterprise discipline. All of these scaling factors are ranges, and not all of them will likely be applicable to any given project, so you need to be flexible when scaling agile approaches to meet the needs of your unique situation. To address these scaling factors you will need to tailor your disciplined agile delivery practices and in some situations adopt a handful of new practices to address the additional risks that you face at scale.
The first step in scaling agile approaches is to move from partial methods to a full-fledged, disciplined agile delivery process. Mainstream agile development processes and practices, of which there are many, have certainly garnered a lot of attention in recent years. They’ve motivated the IT community to pause and consider new ways of working, and many organizations have adopted and been successful with them. However, these mainstream strategies (such as Extreme Programming (XP) or Scrum, which the ASM refers to as core agile development strategies) are never sufficient on their own; as a result organizations must combine and tailor them to address the full delivery life cycle. When doing so the smarter organizations also bring a bit more discipline to the table, even more so than what is required by core agile processes themselves, to address governance and risk.
The second step to scaling agile is to recognize your degree of complexity. A lot of the mainstream agile advice is oriented towards small, co-located teams developing relatively straightforward systems. But once your team grows, or becomes distributed, or you find yourself working on a system that isn’t so straightforward, you find that the mainstream agile advice doesn’t work quite so well – at least not without sometimes significant modification. Each of the scaling factors introduces their own risks, and when addressed effectively can actually reduce project risk, and for your project team to succeed you will want to identify the scaling factors applicable to the situation that you face and act accordingly. Unfortunately, this is a lot easier said (OK, in this case blogged about) than done.
IBM Rational advocates disciplined agile delivery as the minimum that your organization should consider if it wants to succeed with agile techniques. You may not be there yet, still in the learning stages. But our experience is that you will quickly discover how one or more of the scaling factors is applicable, and as a result need to change the way you work.
When you’re inside, safe in the warmth of your home watching snow fall on your driveway outside, all snowflakes look the same. But, when you look at a snowflake up close, particularly when you do so under a microscope, you quickly discover that all snowflakes are in fact unique.
It’s the same with IT projects.
When you look at them from afar, particularly
from a very high level, they all look the same.
However, when you look at them up close, you quickly discover that they too
The agile scaling factors
, which are really just general scaling factors applicable to all types of IT
project regardless of paradigm, help to make this very clear.
For example, when it comes to team size some teams
are small, less than ten people perhaps, some are medium sized, and some are very
large (with hundreds of people).
comes to distribution some teams are co-located in the same room, some teams
have team members in different cubicles in the same building, some have people
working in different buildings, and some even have people working in different
Many agile teams work in
regulatory environments, in fact the July 2009 DDJ State of the IT Union survey reports that one third of agile teams must
comply to industry regulations, although clearly many agile teams do not have
this as a concern. That’s only three
scaling factors. The point is that a
small, co-located team working in a non-regulatory environment will work much
differently than a fifty-person team working in three different locations,
which in turn works differently than a two hundred person team in the same
building working in a regulatory situation.
Different teams, facing different scaling issues will work in different
ways – unique snowflakes from a process point of view.
At IBM Rational we define disciplined agile delivery as:
Disciplined agile delivery is an evolutionary (iterative and incremental) approach which regularly produces high quality solutions in a cost effective and timely manner via a risk and value driven life cycle. It is performed in a highly collaborative, disciplined, and self-organizing manner within an appropriate governance framework, with active stakeholder participation to ensure that the team understands and addresses the changing needs of its stakeholders to maximize business value provided. Disciplined agile delivery teams provide repeatable results by adopting just the right amount of ceremony for the situation which they face.
Let’s explore the key points in this definition:
- Full delivery life cycle. Disciplined agile delivery processes have life cycles which are serial in the large and iterative in the small. Minimally they have a release rhythm which recognizes the need for start up/inception activities, construction activities, and deployment/transition activities. Better yet, they include explicit phases as well. It is very important to note that these are not the traditional waterfall phases – requirements, analysis, design, and so on – but instead different “seasons” of a project. The point is that we need to look beyond agile software development and consider the full complexities of solution delivery. Adopting a full delivery life cycle, not just a construction life cycle, is arguably the “zeroth” agile scaling factor.
- Evolutionary. Agile strategies are both iterative and incremental in nature. Iterative means that you are working in a non-serial manner, on any given day you may do some requirements analysis, some testing, some programming, some design, some more testing, and so on. Incremental means that you add new functionality and working code to the most recent build, until such time as the stakeholder determines there is enough value to release the product.
- Regularly produces high quality solutions. Agilists are said to be quality focused. They prefer to test often and early, and the more disciplined ones even take a test-first approach where they will write a single test and the just enough production code to fulfill that test (then they iterate). Many agile developers have adopted the practice of refactoring, which is a technique where you make simple changes to your code or schema which improves its quality without changing its semantics. Adoption of these sorts of quality techniques seems to work – it appears that agile teams are more likely to deliver high quality systems than traditional teams (according to the DDJ 2008 Project Success survey). Within IBM we take it one step further and focus on consumability, which encompasses quality and other features such as ease of deployment and system performance. Furthermore, although some agile methods promote the concept of producing “potentially shippable software” on a regular basis, disciplined agile delivery teams produce solutions: a portion of which may be software, a portion of which may be hardware, and a portion of which will be the manner in which the system is used.
- Cost effective and timely manner. Agile teams prefer to implement functionality in priority order [http://www.agilemodeling.com/essays/prioritizedRequirements.htm], with the priority being defined by their stakeholders (or a representative thereof). Working in priority order enables agile teams to maximize the return on investment (ROI) because they are working on the high-value functionality as defined by their stakeholders, thereby increasing cost effectiveness. Agile teams also prefer to produce potentially shippable solutions each iteration (an iteration is a time-box, typically 2-4 weeks in length), enabling their stakeholders to determine when they wish to have a release delivered to them and thereby improving timeliness. Short iterations reduce the feedback cycle, improving the chance that agile teams will discover problems early (they “fail fast”) and thereby enable them to address the problems when they’re still reasonably inexpensive to do so. The DDJ 2008 Project Success survey found that agile teams are in fact more likely to deliver good ROI than traditional teams and more likely to deliver in a timely manner.
- Value driven life cycle. One result of building a potentially shippable solution every iteration is that agile teams produce concrete value in a consistent and visible manner throughout the life cycle.
- Risk and value driven life cycle. Core agile processes are very clear about the need to produce visible value in the form of working software on a regular basis throughout the life cycle. Disciplined agile delivery processes take it one step further and actively mitigate risk early in the life cycle – during project start up you should come to stakeholder concurrence regarding the project’s scope, thereby reducing significant business risk, and prove the architecture by building a working skeleton of your system, thereby significantly reducing technical risk. They also help with transition to agile, allowing traditional funding models to use these milestones before moving to the finer grained iteration based funding that agile allows.
- Highly collaborative. People build systems, and the primary determinant of success on a development project is the individuals and the way that they work together. Agile teams strive to work closely together and effectively as possible. This is a characteristic that applies to both engineers on the team, as well as their leadership.
- Disciplined. Agile software development requires greater discipline on the part of practitioners that what is typically required by traditional approaches.
- Self organizing. This means that the people who do the work also plan and estimate the work.
- Self-organization within an appropriate governance framework. Self-organization leads to more realistic plans and estimates which are more acceptable to the people implementing them. At the same time these self-organizing teams must work within an appropriate governance framework which reflects the needs of their overall organizational environment. An “appropriate governance framework” explicitly enables disciplined agile delivery teams to effectively leverage a common infrastructure, to follow organizational conventions, and to work towards organizational goals. The point is that project teams, regardless of the delivery paradigm they are following, need to work within the governance framework of their organization. More importantly, effective governance programs should make it desirable to do so. Our experience is that traditional, command-and-control approaches to governance where senior management explicitly tells teams what to do and how to do it don’t work very well with agile delivery teams. We’ve also found that lean development governance, an approach which is based on collaboration and enablement, is far more effective in practice. Good governance increases the chance that agile delivery teams will build systems which fit into your overall organizational environment, instead of yet another stand-alone system which increases your overall maintenance burden and data quality problems.
- Active stakeholder participation. Agile teams work closely with their stakeholders, who include end users, managers of end users, the people paying for the project, enterprise architects, support staff, operations stuff, and many more. Within IBM we distinguish between four categories of stakeholder: principles/sponsors, partners (business partners and others), end users, and insiders These stakeholders, or their representatives (product owners in Scrum, or on-site customers in Extreme Programming, or a resident stakeholder in scaling situations), are expected to provide information and make decisions in a timely manner.
- Changing needs of stakeholders. As a project progresses your stakeholders will gain a better understanding of what they want, particularly if you’re showing them working software on a regular basis, and will change their “requirements” as a result. Changes in the business environment, or changes in organization priority, will also motivate changes to the requirements. There is a clear need for agile requirements change management [http://www.agilemodeling.com/essays/changeManagement.htm] on modern IT projects.
- Repeatable results. Stakeholders are rarely interested in how you delivered a solution but instead in what you delivered. In particular, they are often interested in having a solution which meets their actual needs, in spending their money wisely, in a high-quality solution, and in something which is delivered in a timely manner. In other words, they’re interested in repeatable results, not repeatable processes.
- Right amount of ceremony for the situation. Agile approaches minimize ceremony in favor of delivering concrete value in the form of working software, but that doesn’t mean they do away with ceremony completely. Agile teams will still hold reviews, when it makes sense to do so. DDJ’s 2008 Modeling and Documentation Survey found that agile teams will still produce deliverable documentation, such as operations manuals and user manuals, and furthermore are just as likely to do so as traditional teams. The DDJ September 2009 State of the IT Union survey found that the quality of the documentation delivered by agile teams was just as good as that delivered by traditional teams, although iterative teams (e.g. RUP teams) did better than both agile and traditional.
At Agile 2009 in August Sue McKinney, VP of Development Transformation with IBM Software Group, was interviewed by DZone's Nitin Bharti about IBM's experiences adopting agile techniques. There are over 25,000 developers within IBM Software Group alone. Follow the link to the interview
to view it online (there is also a text transcript posted there. There's some great insights into the realities of scaling agile in large teams, in distributed agile development, and in particular how to transform a large organization's development staff.