Tackling analytics in a big data world
For years, midmarket companies have invested in bolstering their analytics capabilities, and in the process have gained valuable insight into their business and their customers. While these midsize firms have made undeniable progress with data inside their organizations, tackling the greater volume and new types of information associated with big data still remains a challenge for many such companies.
Why is this? To start, according to a new IBM report titled "Analytics: A blueprint for value in midmarket organizations," many midsize companies lack some of the key ingredients to unlocking the value of big data. IBM conducted interviews with companies across the globe to identify the common patterns found in organizations that are successfully analyzing big data. The report delves into why midmarket companies are lagging behind larger enterprises and what they can do about it.
We spoke with Rebecca Shockley, Global Leader for Technology and Data research at the IBM Institute for Business Value, about some of the report's key findings and what midmarket companies can do to improve their ability to leverage big data in their businesses. Following are highlights from the interview.
You've been researching the role and impact of analytics and big data in business for several years. How has the role of analytics changed in that time? What further changes should we be ready for?
In the past, we've seen companies primarily focus on using very structured data from within their organization to really understand what has happened with their business or potentially what's happening today. Those are things that are very descripive: How many sales have you made in the past week or in the past day? How many customers do you have and what do those customers look like?
Most organizations are now moving into an era where they can be much more predictive –- so that is, looking at what's going to happen based on what we know and what's going to happen tomorrow.
The early adopters and leaders within the analytics industry have moved into an era that is much more prescripive. So not only the predictive: What do we think is going to happen tomorrow? But also the prescripive: What should we be doing about it?
Let's talk a little bit about your latest report, which studies how successful companies are infusing analytics and big data into their businesses. What were some of the key findings?
For midmarket [companies], what we found is that they have a very clear sense of the actions that they can take with analytics. But as of yet, they haven't put measurement processes in place and their platform is still lagging in its ability to deal with the volume and variety of big data. We also find that culturally they are doing their best to make data available and to use it in their decision-making. There's a high level of organizational trust; but where they're lacking is in the data management practices [needed] to really organize, manage and govern that data.
"There's a high level of organizational trust; but where they're lacking is the data management practices to really organize, manage and govern that data."
The last thing that we found in our study is a set of levers that really boost the ability to create value from information in analytics. And these are executive sponsorship, a rigorous funding process, and the expertise within the organization to really be able to analyze and manage the data. And in these three areas – sponsorship, funding and expertise – we see that [the] midmarket is still lacking compared to our global leader pool.
More than 450 midmarket executives and managers were surveyed for your report. What surprised you about the adoption of analytics and big data among midsize companies?
We see a lot more of the entrepreneurial spirit within midmarket companies, and I think that leads them to say, "I'm willing to take whatever [steps] I need in order to make my business better." And having that willingness and openness and risk-taking ability within an organization is critically important to being able to really leverage analytics to create value. So that's been really encouraging. And I think because in this era when we're talking about big data, so many people tend to think that [big data] is just for big companies, and that's not what we're seeing at all.
It seems that midsize companies are really only scratching the surface of what analytics can do.
What we see in the majority of midmarket organizations is that they're so very comfortable with structured data, which is – if you think about an Excel spreadsheet – data that could easily fit into rows and columns. So it's transactional data. You know what product was sold, you know what account it went to, you know what customer it went to, because I can put those in neat rows.
Where we aren't seeing midmarket organizations progress to yet is being able to use what you call more unstructured data and that's things like free-form text. Looking at social media, for example, or a call log, or notes that they've made about customers – being able to look at a wider variety of data such as video or sensor data, and things like that.
If I had to make a recommendation for where midmarket organizations should start their transition from looking at structured data to looking at more unstructured data, I would really recommend that they start with something that will help them understand free-form text. I think that's the biggest bang for the buck because once they can start to analyze free-form text, it opens up call logs, it opens up social media, it opens up all of those customer notations, and it opens up the ability to look at emails.
And these are things that already exist within an organization. Some, not all, midmarket organizations are already collecting or storing this data; so I would say look for ways that you can begin to analyze more of the data that you already have within your organization, because there's a wealth of insights buried within that data.
More from ForwardView
Join the conversation
Smarter backup and recovery
Top 6 reasons to move to DB2