Use Cases

How can Watson Analytics help you

 

 

Blog

Analyze Supplier Delays

Supply chain effectiveness and material sourcing are critical to Retail operations and is getting more strategic by the day to gain that elusive competitive edge. Let's say that I'm a store manager for a large retailer with operations across US, and that we're having some problems with shipping delays. These delays are affecting our business and I’ve been given the job to get to the bottom of it. So what’s in the data? I have our Supplier dataset with details of the deliveries along with other pieces of information. I want to focus on the shipment column that tells me whether a delivery was made on-time or delayed. Let’s see how Watson Analytics can help me with this critical analysis. After I upload this dataset, I click on it to begin my analysis. Even without my typing a question, I see that Watson Analytics has surfaced these interesting and ‘unbiased’ cognitive starting points for me. Getting the overall sense – I click on the first recommendation to look at the spread of opportunity costs across the states. I quickly add Margin and I can see that while Wisconsin has the highest average opportunity costs, Iowa gives us the highest average margins. With that quick understanding, I open a new tab and this time type my own question in natural language to understand the opportunity cost incurred on delayed vs on-time shipments. I select the most relevant insight and realize the dire impact of lost revenue owing to delayed shipments. I now want to know who is responsible for this and type my next question to look at the breakdown of opportunity cost by suppliers only for delayed shipments. I can see that the most severe offender here is a supplier called Rich Industries and maybe we should replace them. Delving a little deeper into our suppliers – I first need to dig deeper and understand what products they supply. So I add Products as ‘rows’ on top of this visualizations and filter it to show only the top 3 suppliers for each product. I can now starting thinking about replacing Rich Industries with other viable suppliers like Acme and Oscorp with relatively lesser opportunity costs for each of these products. But before I pull the trigger on this, I decide to delve a little deeper into our Suppliers. I use the dynamic discoveries ribbon on the right for some inspiration and click on the first insight. I replace the Equipment value with Margins and looks like Rich Industries have the best average Margins and hence getting rid of them might not be such a slam dunk decision. What really drives shipment delays? This now brings me to a point where I want to understand the WHY behind my delayed shipments. I type my next question to know more about what really drives my shipment deliveries and select the most relevant spiral visualization. This easy to understand spiral visualization shows me the factors that drive Shipment and ranks them in order of their predictive strengths. I click to dig deeper into the top factor combination of ‘Weather and Origin State’. This visualization shows how the interplay between ‘Weather and Origin State’ has a bearing on whether a shipment is delivered on-time or otherwise. Similarly, I can view the associated detailed insights for each of the other drivers. I type my next question as though I were talking to an expert – “I want to know more about shipments” and select the predictive model. Here I first review the statistically derived decision rules for the most likelihood for a Shipment to be ‘Delayed’. I can look at the various profiles to understand the conditions that lead to delayed shipments and then use this knowledge to streamline our approach to minimize such delays going forward. Similarly I can analyze such profiles for other important KPIs as well. Bottom-line, while it’s great to have a handle on what’s happening, I now know the WHY behind what’s happening. Wrapping it up in a Display – Confident and excited about the progress I made, I created this interactive Display based on the insights I gleaned so far. Creating interactive and multi-tabbed Displays is really easy – I can either create new visualizations from within Display OR I can reuse those that I created earlier by just dragging them in. I filter this Display on Delayed shipments and Weather conditions of Snowstorm and Thunderstorm before sharing it with my boss. Finally, I wrap it up with a quick email to my boss! So to recap, I was able to analyze my raw Supplier shipment delay dataset and not only found out more about the Suppliers but also understood the real factors that led to delayed shipments. Thanks to Watson Analytics, with these findings, I look forward to streamlining our supply chain operations, optimizing our assortment planning and minimizing shipment delays going forward! You can take a look at the demo video for this usecase here.

Blog

Analyze Sales Effectiveness

Sales effectiveness through self-service efficiency!!! Sales effectiveness is about the relentless focus on ensuring that the sales professionals are winning deals and attaining their quotas. Contemporary sales organizations need to leverage self-service analytics to not only be on top of sales performance but also to understand their real drivers. And do this real quick and on a consistent basis! In this example I’m going to play the role of a Sales Enablement Manager and I’ve been given this Sales and Training data to analyze. Some worthy pursuits – Among other things, I’ll try to find answers to the following pressing questions pivotal to sales effectiveness – Does source of a new-hire help determine his/her success? Which skills correlate with high performance? Which training and sales enablement programs are working well? How important is cultural fit? I’ll use Watson Analytics, and leverage its guidance based ‘smarts’ to take a stab at the above questions. I upload this dataset on Watson Analytics and click on it. This is one of the manifestations of the built in ‘smarts’ that even without my typing a word, I’m presented with these ‘unbiased’ cognitive starting points. I click on one of them to begin my analysis. This packed bubble chart gives me a quick overview of my regional earnings. I see a validation of the three American regions bringing in the bulk of our earnings. Do our hiring sources have a bearing on our quota attainment levels? Next I open a new tab and type my first question in natural language to understand our average quota attainments across the different hiring sources we use. I see that Trade Magazines and direct College recruitment give us the bulk of our top performers. I add Roles to the visualization and notice that while these two give us the best Major Account Managers and System Engineers, these might not be ideal to hire Territory Account Managers. At the least, this begs for further investigation! I add Geo into the mix to check if this trend holds true across the board. This view tells me that the pattern is not consistent and I can further drill down into each of the Geos to explore more. But I do realize that certain hiring sources in certain geographies are more suitable for certain roles than the others. This insight has a huge potential to pursue! A culturally fit seller attains more! I ask my next question as though I were talking to an expert – “Show average of attainment and culture fit compare by Geo” and select the most relevant combination chart. The intent here is to validate if our recent focus on assessing the cultural fitness of the new recruits in America is actually translating into higher quota attainments. With the 'Geo - Region' hierarchy automatically created for me, I drill down into America to find that out. It’s evident is that a great cultural fit results in greater attainment levels. Going forward, we need to expand our screening for cultural fitness across all our regions. At this point, I look for inspiration in the discoveries ribbon on the right which surfaces dynamic insights based on the visualization being discussed here. This is another manifestation of the built-in cognitive smarts! What really drives sales quota attainment levels? Clicking on one of the insights reveals this easy to understand spiral visualization that shows me the factors that drive our sales quota attainment levels and ranks those factors in order of their predictive strengths. Among many other factors, I can see that the combination of ‘Days to Close’ a deal and ‘Oral Communication Proficiency’ skill has the biggest bearing on our Attainment levels i.e. it drives Attainment level with a predictive strength of 50%. The associated detailed insight tells me that the best Attainment levels are achieved when a sale is closed in less than 9 days by a seller whose Oral Communication assessment score is 4 or above. This is great insight! And I can review the associated detailed insights for each of the factors listed. I continue my analysis by leveraging the dynamic discoveries once again. This time I take a look at the suggested Predictive Model for Attainment. Here I review the statistically derived decision rules and look at the various profiles to better understand the conditions that lead to high Attainment levels. While Oral Communication skills and Interpersonal Relationships stand out as the most important skills, I also notice that Sales Training Classes 2 and 3 have been very effective. Bottom-line, while it’s great to have a handle on what’s happening, I now know the WHY behind what’s happening. Bring it all together – At this point, I’ll save my analysis as a Discovery Set. Based on all these insights I gleaned, I can bring them all together in a Display. Here is the first tab. Here is the second tab. And this is the third tab in my effort to summarize all the insights. Creating such interactive multi-tabbed dashboards is really easy in Display – I can either create new visualizations from scratch from within Display (using the ‘+ New Discovery’ option) OR I can reuse those that I created earlier by just dragging them in (from a single or from multiple Discovery sets). In this case, I drag in the packed bubble we used earlier. I also filter this dashboard to look at only our Active employees and a specific Product group. Please note that since ‘Active’ is a global filter, it’ll impact all tabs in my dashboard. Finally, I wrap it up with a quick email to my boss! Effective sales through the efficient use of analytics – So to recap, I was able to analyze my raw sales & training data and quickly discovered the need to prioritize our sources of recruitment based on Geos, validated our practice of screening new recruits on cultural fitness, understood specific skills needed to ensure higher attainment levels and found out about the training programs that were working well. Thanks to Watson Analytics, going forward, we’ll not only attract the right talent but also effectively enable them to ensure that they are closing more deals and bringing in more revenue! Watch the demo video - here

Blog

All your Watson Analytics use-cases in one spot

We've made a small - but significant - change to your Watson Analytics community. Our use-cases have proven popular over the past few months that we've made them easier to find amongst all the blog articles in the community. So, how do you get there? From the Watson Analytics Community menu, select Use Cases. Here's the direct link to get you started - https://www.ibm.com/communities/analytics/watson-analytics/usecases/ Also notice that from that page you can also drill into the content by selecting the Industry or Business Roles links.

Blog

Demystifying Incentive Plan Effectiveness

Contemporary Sales operations is evolving to become more strategic by the day. Balancing incentive compensation and sales quota management on one hand and maximizing bookings leading to improved earnings on the other is ever more critical now! In this example I’m going to play the role of a Sales Performance Manager working in the Sales Operations organization. I’ve been given the job to analyze our increasing overall commission costs and to look for insights to bolster our earnings. I’ll use IBM Watson Analytics to help me on this. It all begins with a dataset! I begin with the Opportunity Payment dataset with details of 35,000 closed opportunities. I want to know how much of the bookings is being paid out in the form of commissions or incentives across all plan participants. Once I upload this dataset, I click on it to begin my analysis. While I can use one of the ‘unbiased’ cognitive starting points, I go ahead and type my question in natural language to look at the spread of Compensation Cost of Sales or CCOS across our operating Geos. What on earth is CCOS? Expressed as a percentage, CCOS for an opportunity is essentially total bookings divided by total incentives paid for that opportunity. Hence, it’s the cost incurred on incentivizing a sale. I see that America has the highest overall average CCOS of just over 8%. I drill down to look into the American regions to find the Western region having the highest CCOS of over 10%. Further drilling down into the Districts under the Western region, we see that Pacific district has the highest average CCOS of more than 13%. The question is whether such high costs of making a sale is leading to proportionate high bookings that eventually lead to earnings? Is the high cost of making a sale actually leading to overall high bookings? I open a new tab and this time create my own visualization – a Tree Map. I quickly fill the data slots with columns from the data tray to focus on credited Bookings and CCOS across Regions within AMER & EMEA geographies. While the West American region carries the highest CCOS at over 10%, it does not yield the highest proportion of Bookings. Similarly, while we found Pacific district with the highest CCOS earlier, we now know it’s the worst performing district in terms of bookings! Digging deeper into the Pacific district lets us hone in on the area of concern – Named Accounts in the Bay Area. At this point, I realize we need to revisit our Account Planning and Forecasting strategy that could be contributing to inaccurate quota-setting for this territory. Understanding the WHY behind the WHAT – I continue my analysis by typing my next question as though I were talking to an expert – “Tell me more about CCOS” and select the most relevant spiral visualization. This easy to understand spiral visualization shows me the factors that drive CCOS and ranks them in order of their predictive strengths. I can view each of the associated detailed insights to delve deeper. I see that the combination of SF_Type or Sales force type and Region has a 31% bearing on CCOS and click on the associated detailed insight. Analyzing their interplay in this heat map, I can see specific SF_Types having disproportionately high costs of making a sale across specific regions. However, I want to further customize this view based on some of the insights I gleaned earlier. So I multiply this heat map over quarters and filter on the Pacific which we found to have the highest CCOS in the American Western region. While I see high CCOS for Resellers during Q1, Q3 and Q4, it’s surprising to see that during Q2 it’s the Wholesalers that have the highest average Compensation Cost of Sales. The mystery behind sales quota attainment levels – I continue my analysis by bringing in quota Attainment details of all our sales employees or Payees. Essentially, I want to check if our quotas are distributed accurately and fairly across our sales territories and if attainment of quota is leading to proportionate bookings. Note that if too many Payees are over-attaining, then we are paying too much in incentives (hence high CCOS) and our quotas were not accurately aligned with the market and if too few are meeting quota, then that means the targets were too high. Both of these situations are not ideal for business and our sellers! First I take a look at our average Earnings and credited Bookings for the last six quarters. Note that credited or validated bookings and the speed with which they are pursued towards making a sale define the quality of our sales pipeline. Filtering on America, I see this steady decline in both since the last three quarters – not a good sign at all! The distinction between over-attaining and over-achieving – Next, looking at the quota Attainment levels – it’s concerning to note that the largest number of commissioned employees is falling into the over 150% category, while the highest average for credited Bookings was found in the 100-105% category. The implication is that we are paying bulk of our commissions for significantly lower revenue results. Also note the cognitive insights bar on the right that dynamically generates insights based on the visualization being discussed here. Next, this view indicates the Territories that exceeded the quota by 50% or more over the last three quarters in America. Ideally, more sellers ought to lead to more Bookings at such high attainment levels. Given that’s not the case here points to a serious under-estimation of forecast business and the need to reassess both the coverage model and individual team quota setting to make sure they align with the sources of revenue. Bringing it all together – Based on the insights I gleaned thus far, I created this interactive display with one tab each for the two datasets. Before sharing my findings, I decide to slightly modify the second tab by adding this combination chart. Wrapping it up – Finally, I wrap it up with a quick email to my boss! So to recap, I was able to analyze my raw datasets and quickly noticed the disparity between Bookings & Compensation Cost of Sales and was able to zero in on one of the problem sales territories. I also found that the actual Bookings and quota Attainment levels of our sellers need to be realigned across regions, roles, sales force types and quarters to drive Bookings. With my recommendations, I look forward to optimizing our compensation cost of sales, streamlining our quota attainment levels and improving our earnings! Use Watson Analytics to discover, display and share your findings with ease and agility! Watch the demo video - here

Blog

Analyze your sales wins and losses

Download the Dataset Converting opportunities in the pipeline into successful deals is the primary focus of the sales organization and is critical to any company! In this example I’m going to play the role of a sales executive at an automotive supply wholesaler and I’m trying to investigate a sales execution issue. We have not been converting enough opportunities lately. I want to better understand our sales pipeline and which deals our sales team can expect to win or lose based on data that I’ve pulled out of our pipeline database. What’s in the data? Apart from doing some ad-hoc exploration, I’ll use the predictive capability of Watson Analytics to understand the most important elements that go into a successful deal. Ask your question right here on the Watson Analytics main page: I begin by typing my question in natural language right here on the Watson Analytics main page. The built-in cognitive ‘smarts’ looks across all datasets to fetch the results and then ranks those insights in order of their relevance. I choose the most recommended Bar chart from the Sales Win Loss dataset to look at my top 5 regions based on opportunity amounts. I can immediately see that Midwest and Pacific are my leading regions in terms of opportunity amount. I drag and drop ‘opportunity result’ from the data tray on top of this visualization and format it accordingly to understand the deal conversion efficiency across these regions. There is surely a huge room for improving our sales execution efficiency across all these regions! I continue my analysis with a new discovery tab and while I can use one of the dynamic and ‘unbiased’ starting points, I type another question and choose one of the more relevant visualizations. Here I take a look at the spread of opportunity amount across my top 5 route to markets and regions. While Field Sales & Resellers are significant, I can further fine-tune this visualization to focus on each of my route to markets and their performance across regions or say across Opportunity Result. Also note the cognitive insights bar on the right that dynamically generates insights based on the visualization being discussed here. What really drives whether you’ll win or lose a sales deal? While it’s great to have a handle on what’s happening, I now want to understand the WHY behind what’s happening. With that intent, I ask my next question as though I were talking to an expert – “I want to understand Opportunity Result” and choose the recommended spiral visualization. This easy to understand spiral visualization shows me the factors that drive opportunity result and ranks them in order of their predictive strengths. Looks like the combination of ‘revenue from client since the past two years’ and ‘total days identified through qualified’ has the most significant impact on whether we win or lose deals. I click on the detailed insight for the top combination driver to dig deeper! Guidance based self-service predictive analytics brings out the latent insights buried deep within your data: This detailed insight essentially shows ‘revenue from client past two years’ along the y-axis and ‘total days identified through qualified’ along the x-axis. In general, looking from left to right I realize that my chances of winning a deal decreases as it stays longer in the pipeline. I also see that an opportunity is more likely to result in a loss if the client didn’t buy anything from us within the last 2 years. Biggest chance of winning a deal is when a client did business with us in the last two years - esp. to the tune of 0-25K USD (blue area in the tooltip). However, although the deal will likely result in a win if they have bought within the last 2 years, that chance of a win decreases as the sales deal rises (look at the red areas from top to down). That was great insight and more importantly, in such short time! Now I want to look at the predictive model for opportunity result. I take a look at these statistically derived decision rules. The simplest rule tells me that if a client’s purchase history with us in the last two years is less than 25000 dollars, there is an 83% chance of successfully closing that opportunity. Similarly, we can look at the more complex deal profiles below. These profiles are extremely valuable in the hands of my sellers and sales managers when reviewing their deal pipeline, anticipating pipeline gaps and course correct our sales strategy accordingly. Wrapping it up in a Display: I then use the Display option to open an existing Sales Win Loss dashboard. Looking from left to right, this Scatter chart shows that irrespective of opportunity amounts, we start losing deals as they stay longer in the pipeline. This could help formulate threshold levels for each supplier based on how many days a deal is in the pipeline and create alert mechanisms to expedite its progression. I click to edit this dashboard so that I can add more content to it. I drag and drop two visualizations from the Discovery Set I just created along with two filters viz. ‘competitor type’ and ‘client size by revenue’. Based on the insights I gleaned so far, I know we need to take a second look at our pipeline strategy so that we focus on the right deals and then work to maximize their progression through the pipeline. Once I’ve assembled my dashboard, added filter elements and customized its look and feel, I can share it with others using one of the multiple sharing options available to me. So to recap, I started with just a question and within minutes I was able to understand more about our opportunities, sales pipeline and more importantly, what drives our wins and losses.

Blog

Analyze Service Agent Performance

Service Agent Performance dataset   The success of a customer service organization depends on the performance of its agents! The devil is in the details - so what's in the data? In this example I’m going to don the hat of an analyst trying to investigate this critical issue of our dipping customer service satisfaction ratings since the past few months. I have been given a subset of data pulled from our case management repository to analyze the last 10,000 cases that our agents resolved. Apart from doing some ad-hoc exploration, I want to use the self-service predictive capability of Watson Analytics to understand what really drives Service Satisfaction and more importantly, how to improve on it. Ask your question as though you were talking to an expert: Once my data is uploaded, I want to start by looking at the spread of cases across its types. So I type my question in natural language right here on the main page and choose the most relevant insight. I see that Systems along with Access and Login lead in terms of volume while Hardware cases were relatively fewer. Customer is king - so were they delighted? I drag and drop Service Satisfaction on top of this visualization to better understand how we have performed across each of these four Case Areas. Minimizing 'Unsatisfied' ratings is a business imperative. At this juncture, it’s important to note that the performance of our service agents is measured based on their average call duration and the service satisfaction ratings on the cases they have resolved. I replace the number of Cases with Case Call Duration and instantly realize that Access and Login case area had more number of cases but takes less call time whereas Hardware case area had less number of cases but takes relatively more time to resolve. This is a good start! Understanding the WHY vs just knowing the 'what':  I open a new discovery tab and use one of the ‘unbiased’ cognitive starting points to understand more about Service Satisfaction ratings. This easy to understand spiral visualization shows me the factors that drive Service Satisfaction and ranks them in order of their predictive strengths. Looks like the combination of ‘Agent Training Level’ and ‘Case Severity’ has the most significant bearing on customer ratings. I can see that the predictive strength of this combination is almost 60%. I can also see that Case Call Duration, Case Area, Case Priority and even Requester Seniority are other important factors. This is great insight and I can look at each of these factor combinations and their associated detailed insights to delve deeper. With my curiosity piqued, I click on the detailed insights to know more about this top driver combination! A well trained agent is like a well trained soldier: Looking across Case Severity (from left to right), especially for the Critical cases, I see that most of the ‘Unsatisfied’ ratings is due to ‘No training’ while most of the ‘Highly satisfied’ ratings is the result of ‘Sufficient Training’. Despite our recent aggressive hiring of new agents, we need to have a strategy in place and to map the right agents to the right case based on their training levels and the case severity. This will have a direct bearing on bringing 'Unsatisfied' ratings down! I open a new discovery and type my next question to look at the predictive model for Service Satisfaction. Here I first review the statistically derived decision rules for ‘Highly satisfied’ ratings and then those for the ‘Unsatisfied’ ratings. This further corroborates the pivotal role that Training plays towards our agents’ ability to resolve cases and delight our customers. Bottom-line, while it’s great to have a handle on what’s happening, I now know the WHY behind what’s happening! Brevity is the soul of wit: I go ahead and ask a new question to look at the spread of avg Call Durations across Agent Training Levels and Case Severity and click on the most relevant insight. This visualization shows how the lack of training results in higher average call duration across case severities. This insight further emphasizes how training can help cut Call Durations and hence minimize cost while expediting successful resolution of cases. When it all comes together like a charm: With my quick exploration done, I can bring it all together in a Display! I open the dashboard that I created based on my analysis. In the first tab, we can clearly see how lack of training significantly increases the average call duration. Great insights but where is the action plan? In the second tab, while the scatter chart and the heat map further underlines what we discovered, I want to modify the Service Agents by Avg Call Duration tree map (on the right) to arrive at a priority list of those agents who should be immediately enrolled in the first phase of our training roll-out. The numbers in the tree map correspond to agent IDs. This is the action plan I want to derive from my insights! To get to this priority list, I apply a bunch of filters as follows - Service Satisfaction is 'unknown' or 'unsatisfied', Agent Training Level is 'no training', Case Priority is high, Case Severity either 'major' or 'critical', Case Type being 'issue' and Requester Seniority being either 'senior' or 'management'. I then distribute this list of agents across Case Areas for better focus during the training program. While the agents delight the customers, I delight my boss :-) I wrap it up with a quick email to my boss! So to recap, I started with a simple dataset and an intent to get to the bottom of our dipping customer satisfaction ratings issue. Thanks to the cognitive smarts of Watson Analytics, I was able to quickly analyze the status quo, understand the latent cause of the problem and concluded with an action plan for its immediate resolution. Check out the narrated video for this usecase - here.

Blog

Using Customer Behavior Data to Improve Customer Retention

Telco Customer Dataset This demo uses the the sample data within Watson Analytics. Please use the sample dataset.   What’s in the Protect Your Customer data set? This data set provides info to help you predict behavior to retain customers. You can analyze all relevant customer data and develop focused customer retention programs. A telecommunications company is concerned about revenue and the number of customers leaving their landline business for cable competitors. They need to understand who is leaving. Imagine that we are analysts at this company and we have to find out who is leaving and why. The data set includes information about: Customers who left within the last month –this column is called Churn Services that each customer has signed up for – phone, multiple lines, internet, online security, online backup, device protection, tech support, and streaming TV and movies Customer account information – how long they’ve been a customer, contract type, payment method, paperless billing, monthly charges, and total charges Demographic info about customers – gender, age range, and if they have partners and dependents Getting the data                          Under the Data tab in Watson Analytics, tap + New Data button. Tap Import > Sample Data and then select and import the Protect Your Customer dataset. The data set appears as a tile under the Personal folder within the Data tab and you’re ready to get to work. It may take a couple of seconds as Watson Analytics is analyzing the data to aid your journey in using this dataset. Which customers have high value? To find the answer to this question, tap the Protect Your Customers CSV data set tile. You want to know where the revenue comes from and what you want to protect.   To better understand the business you may want to look at total charges by internet service type by asking “What is the average TotalCharges by InternetService?” You will want to select the first tile as best represents the line of inquiry.  Note that the image on the tile indicates you should expect a bar chart for this comparison.  In looking at the results, we see that Fiber Optic is clearly  the main internet service that gets the bulk of the revenue. Next, you want to find out about the total charges by contract type. Press the Plus button circled below and we will add another tab to your discovery set.   We want to investigate the total charges by contract type.  Enter the question “What are the average TotalCharges by contact?” and select the first suggestion (tile) from Watson Analytics.  We see the result that 2 year contracts generate more income whereas month-to-month is the lowest.  Typically, I would have thought average charges would be lower with longer contracts.  This is a little surprising. Clearly we want to protect customers with Fiber Optic and longer service contracts.  Lets add to the discovery set again with the plus button and find out how long customers stay with the services for each contract type by asking “What is the average tenure by contract type?”  Again the first suggestion from Watson Analytics is exactly the line of inquiry we want to explore, so we will select the first tile. When reviewing the results, we see that month-to-month contracts stay with the service on average 18 months whereas customers stay 42 months and 56 months on average for one year and two year contracts respectively.  You can hover over the bars of the chart to get the actual numbers.  The month-to-month contracts are not leaving immediately, but we should be thinking about how we can move these customers into longer term contracts. What drives customer tenure and churn? In thinking this through, we want to nail down the factors that drive customer tenure.  Let's add to the discovery set and ask “What drives Tenure?”.  The first suggestion fromWatson Analytics brings you to a spiral diagram which highlights TotalCharges and InternetService as the key factors for Tenure with a predictive strength of 91%. Looking at the relationships further down the list, I see that churn also affects tenure. This makes a lot of sense. Let’s see what drives churn by adding to the discovery set and asking “What drives Churn?”.  This time we will look at the second tile as it shows a decision tree for Churn.  By scrolling downward on the decision tree and hovering over each of the tree nodes, we can see that  customers with a month-to-month contract with less than six months tenure and Fiber Optic services churn 75% of the time. This occurrence is very high and we need to understand this better.  Perhaps the service is weaker than what our competition is providing and these new customers see the difference.  In any case, we need to speak to customer services and our hardware team with this finding as this directly impacts Fiber Optic revenue which is key to our business. Again, you can watch the narrated video for this use case here.  

Blog

Analyzing Customer Campaigns (US PacWest usecase)

Download the Dataset The challenge: One of the biggest challenges of any business is to grow their business through effective marketing. As a Marketing manager, you want to understand your customers, what they want and what really drives them, so that you can sell to them more effectively. And you want to do this quickly and without having to depend on IT or pre-requisite expertise. Watson Analytics can help you do that and much more without any expensive or time consuming setup or training. Analyze using natural language: Let’s see how that works with an example! I want to assess the performance of my email campaign in the Western Pacific states in US. I also want to take a look at my customers’ lifetime values and more importantly understand what drives those values. I begin my discovery by typing my question in natural language right here on the Watson Analytics welcome page. The built-in cognitive ‘smarts’ looks across datasets to fetch my answer and then ranks them in order of their relevance. I choose the Tree Map recommendation from the Pacific West Campaign dataset to look at the spread of my customers by their response to my email campaign and their Renew Offer Types. Find who are responding: I immediately realize that the campaign was not a huge success and that’s not very encouraging! At this point, I could also change the focus of my analysis by using the interactive titles – like I can change Response to Education or say Coverage, etc. However, I want to delve deeper to get a sense of the Yes responders shown in green. I keep the ‘Yes’ respondents and keep digging to see what is driving them. By changing the view I can quickly see that the second offer is actually more important to focus on. Something that in a spreadsheet would be hard to discern and because of the powerful charting capabilities in WA, that becomes very obvious very quickly. Know where they are responding from: Let me focus on Offer 2 and Drill Across to see what States those customers were from. Watson Analytics automatically rendered a tree map that shows me the distribution of my Customers who responded as ‘Yes’ to Offer-2 across States. California and Oregon stand out among the other states. This is helpful but I wonder if there is more to the story. Adding data to the chart is easy, I’ll add Policy Type – Policy roll up that the smarts of Watson Analytics dynamically created for me. I’ll then want to drill down into my biggest Policy Type which is Personal Auto while keeping all my existing filters. Drilling down I see that Personal 2 and Personal 3, which are part of my basic categories have the highest response and interestingly California and Oregon are competing for top spot. I open a new tab in my discovery set to explore further. This time I use on one of the suggested starting points that Watson Analytics has generated for me and click on the Income by State map visualization. I modify the map to now look at the average Customer Lifetime Values and the number of Customers across these states. While Oregon has a slightly higher average Customer Lifetime value, California has the higher concentration of customers. Also note the cognitive insights bar on the right that dynamically surfaces insights relevant to the visualization that’s being discussed. I click on first recommendation (as highlighted in the image below) – ‘Top Drivers of Customer Lifetime Value’. Understanding the real drivers will not only help me retain my customers but also enable me up-sell and cross-sell to them. Understand what really drives customer lifetime value (CLTV): I’m getting a picture of my campaign results, but of course, what I really want to know is what I should do looking forward. Who are my most profitable customers and what drives them? Understanding these aspects will not only help me retain them but also to up-sell and cross-sell to them. That’s where Watson Analytics’ built in predictive analytics come in. This easy to understand spiral visualization shows me the factors that drive my target, in this case, Customer Lifetime Value, in order of their importance. Looks like the combination of vehicle class and number of policies has the most significant impact on my target and drives it with 65% predictive strength. Identify the right target market and course-correct (if required): The devil is in the details and I surely want to dig deeper! Clicking on the associated detailed insight for the top combination, I can see the sweet spot for Customer Lifetime Value is at the intersection of the Luxury Vehicle Class, which comprises Luxury SUVs and Cars, and the Number of Policies. So essentially that’s my target market and that’s great insight! Bring it all together: Finally, after I explored my data, I can bring it all together in Display. I create a new Display by choosing an appropriate template and start building my dashboard by dragging and dropping visualizations from my saved discovery sets – in this case I use the discovery set that I just created. I can choose from a variety of layout and formatting options to create interactive multi-tabbed dashboards. I can assemble, format and even edit my saved discoveries and visualizations to generate a better understanding of how my campaign is performing. Here I added two local filters viz. Location Code and Gender and filter my dashboard on only on females from non-urban locations. Looks like Nevada could be a huge potential for us to grow this market! To summarize, I started with just a question and within minutes I was able to understand more about our email campaign, how they were received, our policies & offers and more importantly, what really drove our Customer Lifetime Values. With these insights, I’m better placed to focus on our strengths and maximize our marketing and campaign ROI! Use Watson Analytics to discover, display and share your findings with ease and agility!

Blog

Exploring Banking Loss Event data with Watson Analytics

Download the Dataset   This IBM Watson Analytics use case shows you how you can analyze loss event data from IBM OpenPages GRC using the updated Watson Analytics user experience.  (If you haven’t already signed up for Watson Analytics, you can do so here for free.) For the purposes of this use case, We are working on the risk team of a financial services enterprise and we need to review and analyze 7 years of loss events recorded in OpenPages which can be download from here. After we login we will see three main tabs and am currently positioned within the Data tab.  The Discover tab is where you will explore and discover the data you have in Watson Analytics.  The Display tab takes the discoveries and assembles them into rich stories, dashboard and infographics to share.  It all begins with your data, so the first thing to do is import the spreadsheet in Watson Analytics. 1.    Tap the New Data button. 2.    Tap the Local file tab, then tap the Browse button to select the spreadsheet you downloaded for the win/loss analysis. 3.    Tap the Import button. After importing the spreadsheet into Watson Analytics, where we can directly access the Excel file that I exported from the Loss Events pages in OpenPages . We can analyze the data in five steps. Step 1. Discover your Data When we click the tile created by the import and Watson Analytics immediately positions me into the “Discovery” functionality which provides me with a set of suggested questions or starting points that you can use right away.  You could also type your own question here too. If I am seeing a question in the tiles that make sense for me, I could simply click on the tile to get the result.  Note that each tile has a graphic showing you what to expect in the result. Let’s start by looking into the trend of the net loss by year. Watson Analytics presents a list of possible interpretations of what you wanted.  As it turns out, the first tile identifies exactly what we want, let’s select this question.  When I look at the figures, it appears that the budget safeguard we put in place in early 2014 worked as expected after that big loss in 2013. Next, we should check the trend of net loss by region by dragging the “Region” field (circled above in red) from the data tray directly onto the data visualization. Now we see the safeguard also worked as expected for all the countries: a great result. Step 2. Be open-minded and take suggestions from Watson Analytics Watson Analytics provides suggested lines of inquiry on the right based on interesting data distributions it finds adjacent to our current analysis.  These suggestions change as I change my line of inquiry.  I note that Net loss by business is very relevant, so lets evaluate this discovery by clicking on this tile. When we do that, we learn that our Corporate Finance and Retail Banking businesses account  for close to half of the net loss of my company.  Yikes! Lets tweak the visualization to use a treemap by clicking on the left “Visualization” icon”. And then select the Treemap visualization. We now see an interesting view of net loss by business. Step 3. Use Predict to review a model with net loss as the target Watson Analytics discovery capabilities also apply to predictive analytics.  Next, we ask the question: “What drives Net Loss”. Watson Analytics creates a spiral diagram where the factors most likely to correlate with the outcome (called “predictors” or ”drivers”) appear closest to the center. Here, business unit and risk sub-category are the top predictors with a predictive strength of 75%. By clicking the   icon next to an item in the list of drivers we are able to zoom into the details of this model.   We can see that the top issue with Net Loss is the relationship with Vendors or Suppliers with our Corporate Finance.  Mouse over the cells to get details as shown below. Let’s rename the tabs for our Discovery Set and then save it: 1.    Click on a tab name and then click the “pencil” icon to edit the tab name. 2.    Click on the disk icon on the top right (   ) and provide a name for the Discovery Set. 3.    Close the Discovery Set using the drop menu as shown below. Step 4. Assemble the data within a Display We can quickly put these findings into an interesting Display such as a dashboard or Infographic to share with others.  Click on the Display tab and then click “+ New display”. Select the Dashboard option and then select the four quadrant display template. On the left, we will be able to locate the previously saved Discovery set in the personal folder assuming we saved in the default location.  Expand your Discovery Set and select a visualization. Drag or click the four visualizations from the Discovery set onto each of the quadrants of the template (the blue box will glow to show you when to drop) and save the Display using the disk icon as you had done before. Step 5. Share the new insights! These findings are significant and we will want to share them with the VP of Risk.  We could use shared folders if the VP is also a user in the same Watson Analytics account or we can share with anyone using PDF, Powerpoint or Image files via email of download. Click on the share    icon, select Email and then select “PDF”.  Using download or email, the person you are sharing with does not even need Watson Analytics to benefit from my analysis. Great Job!  Don’t stop there apply these analytics to your own data!

Blog

What will a graduate degree give me? Exploring the American Time Use Survey data set

American Time Survey data The American Time Survey data is included within Watson Analytics as a sample data set called American Time Use Survey.csv Imagine you’re a university student thinking about going to graduate school and wondering what the impact would be on your income and how this affects your free time over the long term. The American Time Use Survey data set contains data about the amount of time people spend doing various activities, such as paid work, volunteering, childcare, and socializing. This demographic data is about a subset of Americans but can be applied more widely.   It all starts with your Data!   In Watson Analytics, click the New Data button. Click Sample Data icon.   Select American Time Use Survey.csv, scroll down and then click the Import button. The data set appears as a tile in your Personal data folder. Watson Analytics analyzes the data and metadata when uploading the csv file to provide smarter data discovery and analysis. In this process, Watson Analytics identifies field names and concepts, possible measurements and hierarchies in your data and captures metadata including data quality, data distributions, skewness and missing values. Let’s ask our first question.   Does higher education lead to higher earnings? Tap the American Time Use Survey data set tile. You are taken into a new Discovery set. This is where you start interacting with the data. That single tap gave you a list of Starting points, which are different ways to launch yourself into data analysis and visualizations.   Let’s enter our question: does higher education lead to higher earnings, and then press Enter. You now see different Starting points based on your question and these are ranked by relevance.  The most relevant inquiries bubble to the top of the list.   Select the Starting point: What is the breakdown of Weekly Earnings by Education Level? The results are shown in a treemap visualization. The size of each rectangle below indicates the relative size of weekly earnings by education level. The largest rectangles are for those with advanced degrees. This visualization is for all ages.  Let’s see how weekly earnings by education level breaks down when ages are added in. At the very bottom of the window is the Data Tray showing all the column headings in the data set. Add Age Range to the visualization. Just drag it from the data tray (the grey strip on the bottom) and drop it anywhere on the visualization. Note: you can also drop it on the Data Slot beside the drop down for Education Level on the bottom left just below the visualization. There’s a lot more detail in the visualization now, perhaps too much. Let’s focus in on people with college or university degrees. Below the visualization, you can modify what is displayed. Select Education Level and tap the items listed from 9th grade down to Some College to remove them from the visualization. You may need to scroll down in the box to complete this. Some of the smaller rectangles are for age groups that aren’t really relevant to the question that we’re exploring. People aged 0-19 have generally not completed university or college, and those aged 70 and older have generally retired from paid work. Let’s filter out these groups: Tap Age Range at the bottom of the Visualization Select 0-19, 70-79, and 80+ to remove them.   Then tap Done or outside the Age Range list to close it.   Try a different visualization type Different visualization types communicate information about data in different ways. Let’s see what else we can learn by using a different visualization type. Tap  to the left of your visualization to see what Watson Analytics recommends. You can, of course, pick any type you want. Tap the first recommended visualization: the Bar chart.   You see that earnings peak when people are in their 30s and 40s, regardless of education level. But what about work-life balance? Earnings is one way to look at it.  However, life is about more than how much money you earn. Does someone with more education work longer hours? Do they have time to spend with their families and friends? Lets add to this discovery set with a simple click on the plus button circled below and then ask the question “How do weekly hours worked compare by education level?” By clicking on the insight tile circled above you will see the treemap.  We can see that people with more advanced education level spending more time working.   In the previous inquiry on weekly hours worked by education level, I see that there are other questions we could ask that are more predictive in nature. Similar to Step 8 lets add to the Discovery Set and determine “What drives Weekly Earnings?”  Select the circled insight tile. It may take a few minutes for this insight to process as it is going through many predictive models to determine what drives weekly earnings.   Once it evaluates thousands of models, it will present us with a short list of predictive relationships. Not surprising -based on what we have already seen that weekly hours worked and education level have relatively strong relationship with weekly earnings with a predictive strength of 45%. If you wanted to see more drivers, you can tap the link for “Show more drivers”.  If we tap the button to the right of the driver we can see more details on the driver. As we mouse over the blue blocks in this heatmap, which show the key elements of the relationship, the cell values for weekly income (shown as color intensity) are generally higher earnings as you move your cursor up and to the right. What did we learn? These findings show us that working hard to get good marks in school to attain a higher education does not stop there.  We will need to keep working after we have attained our advanced degree to continue in building up the weekly earnings.  This of course affects our free time. Don’t stop there - Try this type of analysis with your own data set!

Blog

Brand Analysis for Telecom companies with Watson Analytics for Social Media

Dataset Dataset is in WASM cannot be put into a shared folder and will need to be created in a user account. In the personal folders of wa2pro_demo*@yopmail.com accounts, we have put the Telco WASM project. One of the most common uses of Watson Analytics for Social Media is brand analysis.  This blog will walk you through six considerations when producing analysis on a brand using social media data. The blog is complemented by this short video.  The goal here is that you will get an idea of how you would go about doing a brand analysis on a Telecom company after reviewing these considerations. 1. Getting your data right When starting an analysis on a brand, ask yourself the question, can the brand be confused with something else?  For example, while Sprint is a company providing Telco services, it could be referenced in a running race.  Perhaps the brand is so pervasive such that it is referenced in ways that are outside the scope of your analysis. Consider the AT&T Center, the multi-purpose sporting arena; people referencing the AT&T Center may be referencing a sport rather than the actual brand. The most typical situation which presents a challenge is if you did not immediately consider the possibilities for polysemy.  Watson Analytics for Social Media topic suggestions is design to call out these possibilities.  As seen in the video link, the AT&T sporting event is called out within the topic suggestions so you can address this in your configuration. So how do you correct for this problem?  You have two ways to get to the data you want:  context terms and exclude terms are a means to narrow the topic area.  Exclude terms will remove conversations from the data set.  Context terms require additional context to the data to be included.  In the video there is a short list of context terms for each of the topics to ensure we are capturing the brand itself.  As you add terms to the configuration, your topic suggestions should provide topics more aligned to what you want to capture for your analysis. 2. Open your mind to alternative vernacular. When configuring your topics, consider the topic suggestions which expose you to a list of topics that your configuration is capturing.  Are there words, hashtags or jargon that would allow you to better target your topic as well as expand what you are looking for.  Using topic suggestions, you should see terminology that you can add to your include and context rules. Topic suggestions uses a sampling of data based on the settings in the project, such as date ranges, languages specified as well as sources selected.  It is a good idea to try topic suggestions with different date ranges, sources to get as many suggestions as you can before running your analysis. 3. Determine how to breakdown the topics What topics matter to you? If you want to compare the brands on author’s perceptions of pricing, then you should be adding a theme which captures conversations on pricing.  As shown above, pricing can be identified many ways: bucks, dollars, price, cost, paid are some ways you may reference the pricing. Pricing is a common perspective to analyze a brand, but there are many others for the Telco business:  Quality of Service, Coverage, Customer Service, Loyalty, Plans, Bandwidth, Speed, Phones, Internet, Deals, Unlocked sim cards, etc...  I am sure you can add your own.  It is important to note that many of the topics listed here were not listed because they are a given, but in many cases, topic suggestions helped build the list.  Building the list based on topic suggestions will reflect what people matter as it is a sampling of the actual conversations.  Using topic suggestions for the themes can be quite useful to expand list of themes and the rules within each of the themes.  In the video you will see the word outage added to the service theme as an example. Other terminology you may want to put into the rules for quality of service would be along the lines of “dropping calls”, “static”, “dropped calls”, “my service” etc.,  Topic suggestions should also call out terminology from competitors which will allow you examine the perceptions of strengths and weaknesses of each of the brands.  The sentiment can be different for the brands for given themes which magnifies the importance of themes. 4. Impact of the reports In the video, we have multiple brands being analyzed.  Including multiple brands is a good practice to follow as it provides a benchmark for high and low watermarks on volume, sentiment and other measures.  The reports available within the social media project are great for doing brand analysis.  It allows for head to head comparison on Share of Voice and Trends in the Share of Voice.  You can quickly see which brand has the biggest mind share. Again the theme breakdown really enables you to review the brands’ strengths and weakness. As the video calls out you can filter on anything in the data tray which makes this particular perspective valuable. The sentiment report helps you with benchmarking the sentiments for each of the brands with can also be useful to capture the strengths and weaknesses of a brand. The reports on Geography, Authors, Behavior and Demographics call out who the authors are when referencing the brands, where are they from, where do they communicate.  This information will support a lot of the outbound activities to make sure the messaging is right for the target audience using the right social channels. The reports call out the brands’ influencers in the social space and these individuals/groups are key people you want to engage with to maximize your digital presence. Current users of a brand are the life blood of brand so knowing what they are talking about is very important to understand.  The theme breakdown helps the brand know what particular elements matter to the core customer. 5. Using Watson Analytics for more insights When you run an analysis, you create a dataset within Watson Analytics.  Using this dataset, you can create your own reports by asking questions about the social data that are not available in the default reports.  You can review authors who are talking about being a prospective customer, or target the competitors’ audience who are talking about leaving the brand.  Using the full features of Watson Analytics, you can choose different metrics (other than mention counts) such as follower counts, document counts, share counts for brands and themes and quickly put together your favorite visualizations within your own Watson Analytics displays. 6.  Combine other data Social data gets more compelling when view in context with other data.  You can add other datasets to the mix, such as your CRM, Digital Analytics or e-Commerce data to glean more ideas on how go to market.  Combining this data in a single view opens up new opportunities to answer questions relating to the “why” people are responding to the messaging, pricing, branding that we put into the marketplace. At this point within the blog, you should having a better view as to how you can create a project within Watson Analytics for Social Media for brand analysis on Telecommunications companies.  Similar approaches can be used for brand analysis on other industries as well.  Try it out with your brand!

Blog

Watson Analytics Use Case for IT Helpdesk: Minimize resolution times and maximize satisfaction

Download the Dataset IT Helpdesk is one of the core pillars of any technical support organization. For this usecase, let’s assume that our IT department’s help desk customer satisfaction rating has dipped for three straight quarters and I know we need to reverse this trend. I’ve been given data about hundred thousand closed tickets with information about customers who submitted the tickets, the IT analysts who were assigned the tickets, the type of issue, in what areas, the severity and the priority of the tickets, the days that it was open for and the satisfaction of the customer after it was resolved. I’m going to leverage the powerful cognitive, predictive and guided visual analytics capabilities of Watson Analytics to investigate what really drives our ticket resolution times and ways to improve on them to bolster our customer satisfaction metrics. I upload the dataset and after Watson Analytics has ingested and analyzed the underlying data, it is assigned a quality score which helps me understand if there are any irregularities within the data. I start my investigation by just clicking on the data asset. Even without my typing a question, I’m presented with intuitive starting points based on this dataset in the discovery mode. While I can start with any of these recommended and unbiased starting points that include predictive insights as well, I can also ask my own question. With my intent to understand the drivers of Ticket Resolution times, I type my query in natural language – “what drives days open”. Watson Analytics immediately responds with a set of insights ranked in order of their relevance. I click on the most relevant spiral diagram visualization to begin my analysis. The spiral diagram shows me all the factors that typically drive the number of days that a ticket was open for or in other words the Ticket Resolution Time. The closer a factor is to the center of the spiral, the stronger is the prediction. I change the aggregation type to show me average times and then look at the drivers list. I immediately realize that the combination of the function to which the ticket was assigned to along with the type of the ticket is the most important driver of Ticket Resolution Time. I can also make use of the insights bar on the right that surfaces dynamic suggestions based on the query being discussed here. I quickly give this discovery a name and continue my analysis by looking at the detailed insight for the most important factor combination. Using the Drivers List, I can delve deeper into each of these insights to know more. Here I look at the interplay between two of my more important drivers of Days Open – the function type the ticket was submitted against and the type of ticket. It’s evident that Hardware Requests take the longest time to resolve. It’s important to note that Watson Analytics suggested this insight to me based on statistical relevant patterns that it discerned from my dataset. Moving on, while I can use the suggested insights or start with a visualization, I type my next query to understand more about the functions that the tickets are submitted against. I open the suggested Tree Map to understand how the volume of tickets is spread across my functions. I add Severity and then replace it with Priority as I review and analyze this spread. Systems function shows the highest volume of tickets across all Priority levels. So far we found out what drove our Ticket Resolution Times and the spread of ticket volume across Functions and Priorities. I continue my analysis with my next question to understand more about Requester Seniority as it was deemed as an important factor earlier in the spiral visualization. After sorting on Ticket volume, I drag and drop Ticket Type on top of the visualization and also format it. I save my discovery set and at this point, I can also share it with my colleagues or even give them access to the actual asset for us to co-create going forward. After I explored my data, I can bring it all together in Display. I create a new Display by choosing an appropriate template and start building my dashboard my dragging and dropping visualizations from my saved discovery sets. I can assemble, format and even edit my saved discoveries and visualizations to generate a better understanding of our IT helpdesk usecase. So I use the discovery set that I just created to populate my Display by just dragging and dropping the required discoveries. I can also choose from a variety of layout and formatting options to further customize my dashboard. While I’m fine with all the other elements of the dashboard, I open the bar chart in edit mode and replace the existing measure which was ‘Number of Tickets’ with ‘Days Open’ and further change the aggregation to show average Days Open. This helps us to have a closer look into the Ticket Resolution Times across Requester Seniority and Ticket Types. Finally, I add the filter elements to my dashboard and adjust them to show only those tickets that were ‘Critical’ in Severity and resulted in ‘Unsatisfied’ Customer Satisfaction. It is pivotal that we minimize unsatisfactory resolution of critical tickets especially those submitted by Management and Senior requesters. I already see areas of improvements in select functions like Access/Login and Systems with regards to ticket volume and Hardware with regards to resolution time. So to recap, I started with a data about helpdesk tickets and within moments Watson Analytics was able to interpret it and presented me with a visualized list of interesting insights with plain language interpretations that I can easily understand. I was able to understand more about Ticket Resolution Times and that how that is impacted by the Type of Ticket, the function area Seniority of the requester and the experience of the employee it is assigned to. With these insights, I can better allocate my resources to critical and previously unsatisfactorily resolved tickets and provide additional training to those employees that take longer to resolve tickets.

Blog

Watson Analytics Use Case for HR: Retaining your valuable employees

Download the Dataset Employee attrition is an established reality and ways to minimize it a pivotal business imperative. As an HR analyst and I’m trying to understand what the key drivers are for employee attrition in my organization. I want to ensure that we attract and retain our top talent and take remedial action if it looks like we’re going to lose good people. What's in the data? I’ve been given a data set about all of my past and current employees, so here I have access to some demographic information like age, gender and marital status along with their compensation and performance related details. But the field that I’m most interested in here is whether they’re a current employee or whether they’ve left for greener pastures. I want to use this data to understand what really drives attrition in my workforce and then use that knowledge to prevent my top performers from leaving. Fortunately I have access to Watson Analytics to help me on this journey. I upload my spreadsheet and after Watson Analytics has ingested and analyzed the underlying data, it is assigned a quality score which helps me understand if there are any irregularities within the data. I begin my analysis by just clicking on the data asset. Watson Analytics immediately presents me with intuitive starting points based on this dataset in discovery mode. Natural language is the best language: While I can begin with any of these recommended starting points that include predictive insights as well, I can also ask my own question in natural language. By the way, I can even start with asking a question on the Mainpage and Watson Analytics will look across the uploaded datasets to find the insights for me. Here I ask a simple question about looking at the spread of my employees across departments. Not only am I presented with answers but I see these insights have been ranked in order of their relevance. I start by selecting the most relevant insight. As I look at the bar chart, I drag and drop the Attrition indicator on top of it to understand attrition across departments. I use the data tray to replace Attrition with Job Roles. This gives me a good overview of attrition across the different job roles in our organization. I quickly stack these values and give this discovery a name before moving forward. While I am able to clearly see now where the Attrition is, I now want to understand why our employees are leaving us! What REALLY impacts attrition? So I type my next question as if I were talking to an expert – “I want to understand Attrition”. I instantly get the resultant insights and I choose the Spiral visualization to continue my analysis. In this easy to understand Spiral diagram, I can immediately see those factors that drive Attrition in my organization. The closer a factor is to the center of the spiral, the stronger is the prediction. I see that the combination of Over Time and Monthly Income has a huge bearing on Attrition and so does the combination of Work Life Balance and Job Role. Using the Drivers List, I can delve deeper into each of these insights to know more. I quickly add another tab and ask a question to understand Over Time across Job Roles and I select one of the highly recommended Tree Maps. While the Tree Map shows me the spread of Over Time across Job Roles, I use the data tray to understand this spread across Departments. I further look at this spread for only those employees who left us. I can also make use of the insights bar on the right that surfaces dynamic suggestions based on the query being discussed here. Armed with my knowledge about the drivers of Attrition and my intent to prevent my top performers from leaving us, I ask my next question to look at all my employees by Total Working Years, Monthly Income and Years at our Company. Given there are three numeric columns in my question, Watson Analytics presents the most relevant insight as a Bubble Chart. This bubble chart shows me all my employees color coded by Departments along with their Monthly Incomes, Total Experience and Years in our Company. I save my Discovery Set and will now consolidate my findings in a dashboard in Display. While this bubble chart shows a positive correlation between Monthly Income and both Total Experience and Years at our Company, I’ll save this discovery set to use it in my Display. At this point, I can also share my findings with my colleagues. Bringing it all together: Finally, after I explored my data, I can bring it all together in Display. I create a new Display by choosing an appropriate template and start building my dashboard my dragging and dropping visualizations from my saved discovery sets. I can also leverage an entire gamut of formatting and other options to customize my visualizations and my overall Display. I use the Employee Attrition discovery set that I just created to build my Display. It’s as easy as dragging and dropping these visualizations to populate my dashboard. I populate my first tab to serve as an overview and once created, I give my first tab a name and proceed to create a second one. I drag in the bubble chart and add in the required filter elements. Most valuable employees who are at the highest risk of quitting: From this exhaustive bubble chart comprising all my employees, I want to zero in on only our more valuable employees who are at the highest risk of leaving us. So I use a combination of filters to get to that – high on Over Time, Awaiting a Promotion since at least 3 years, high Job Levels and Job Involvement along with relatively lower Monthly Incomes. As easy as it sounds, I have all those employees that we must take immediate action to retain. I’ll recommend that we first focus on those employees (as highlighted in blue) who have a lot of industry work experience as well as have been veterans in our own company. I will share my findings and the necessary call to action with my boss through an email to take this further. I’ll also share my discovery set and display with my analyst colleagues granting them suitable access control so that we can co-create and delve deeper into this usecase going forward. So to recap, thanks to Watson Analytics’ guided data discovery capabilities, I was not only able to understand the factors that drove Attrition in our organization but also derived a tangible list of those high performers that we must take immediate action to retain.

Blog

Watson Analytics Use Case for Marketing: Driving the Success of your next Marketing Promotion Campaign

Quick access to the right insights between your marketing promotions and resulting sales has been elusive for most marketers—unless you happen to be a data scientist or business analyst. Now, with the power of self-service analytics and smart data discovery in Watson Analytics everyone can harness their inner data scientist. In this use case I’m developing a marketing campaign for a fast food restaurant franchise. We have been experimenting with the roll out of a new menu item and we just completed a 4-week pilot promotion across different store locations and markets. Now I need to analyze the results of the promotions and find the connections between the campaign and the related sales.  How did the promotion do during the different weeks? Did the promotion do better in one market or store location compared to others? Finding the right marketing promotion is a challenge, but when done right, it can lead to a whole new audience of hungry customers walking through your doors. Follow this use case to see how Watson Analytics can help you find the most important insights from your customer data so you can implement the right promotion plan and draw in the customers and sales you are looking for. See the video version of this use case here: https://www.youtube.com/watch?v=Urur84JY-h4 STEP 1: Add the sample data This example is based on the “Analyze Test Market Campaigns“ sample data available directly within Watson Analytics. To get started, add the sample data to your Watson Analytics environment. On the Watson Analytics > Data home page, tap New data. On the Import tab, tap Sample Data. Select Analyze Test Market Campaigns and tap Import. You can also find out more about this data set and download a local version of it here: SAMPLE DATA: Marketing Campaign, Promotion Effectiveness – Fast Food Chain https://www.ibm.com/communities/analytics/watson-analytics-blog/marketing-campaign-eff-usec_-fastf/ After the data is loaded, it appears as a data set tile in your Personal folder. STEP 2: Ask questions to explore your sales and promotions data Asking questions about your data is the key step to diving in, exploring and visualizing your data. To see how the promotion campaign went, I start by asking some questions about the data just using plain language questions based on the column titles in the data - no SQL or complex queries needed here. For example, I enter the question:  “Show me Sales by Week”. Right away, Watson Analytics does some of the thinking for me (that’s the cognitive part) and offers a number of starting points based on the Sales and Week columns.  This helps to lighten the cognitive load on my own brain while allowing me to interact with the data in a more natural way. I decide to look at the Sales by Week bar chart suggestion, but there’s not too much there - sales looks pretty flat across each week. That’s ok, because I’m just iterating and exploring here. Let’s try something different. From here I have a couple different paths I could take. I could look at the discoveries listed on right side, change things in the current visualization or ask a new question. Let’s stay here and change the current visualization using the interactive title to change the fields in the visualization. How about sales by promotions? I tap on Week and change the column to Promotion. The results are a little better here and I can start to see some interesting results between the different promotions. Using the interactive title I can quickly iterate through different visualizations to explore different combinations of fields. Now I try visualizing sales by these other fields: Sales by Market Size Sales by MarketID Sales by Age of Store TIP: Instead of over-writing the same visualization and tab, use the tab copy feature to duplicate the current visualization on a new tab and then change the field there. Step 3: Customize your visualizations to take a deeper look After asking questions and using the suggested visualizations, I start building some of my own custom visualizations to explore deeper into the data. I’m interested in how the three different promotions breakdown between the different market sizes and market IDs, so I try the following combinations: Sales by MarketSize with Promotion set as the Color by option. Sales across the 10 different market IDs with Promotion set as the Color by option. Multiplier option: Then I try the multiplier option with a packed bubble visualization to plot Market ID (bubble label), Sales (bubble size) and average Age of Store (bubble color intensity) across each of the three Promotions. In this case, promotion 3 and market ID 3 stand out with the highest sales. The lighter color intensity of the bubble also shows that this combination has a lower average store age. I might want to think about evaluating and possibly updating some of my older stores to increase sales. STEP 4: Use predictive analytics to find the top drivers of sales OK, so I spent some time exploring and visualizing. Now it’s time to run some predictive analytics on the data to see how Watson Analytics can tell me which parts of the promotion were the main drivers of sales. Was it market size?  Store location? Or something else? Let’s see. I ask a new question: “What drives Sales?” The resulting spiral diagram shows the predictive insights for the top drivers of sales during the promotions. Watson Analytics also displays a table of the top results. I scroll the list to see the top one and two-field combinations that impact sales. Scrolling through the list of drivers, I see the top single fields that impact sales: MarketSize MarketID LocationID Here are the top two-field combinations that impact sales: MarketSize and MarketID LocationID and MarketSize Promotion and MarketSize Let’s take a closer look at these highlights to see the relationship between market size, market ID, promotion and sales. Looking at the impact of MarketSize and MarketID on sales, I can see that small and large market size have the highest sales overall for market ID 3 and lower. A certain set of location IDs (405 and below) in the large market size have the highest amount of sales overall. Finally, looking at promotions and market size, I see that these drive the most sales when promotion 1 and 3 were used in the large market size stores. STEP 5: Build a dashboard and share your insights Now that I’ve explored the data and ran a predictive model, I can pull all that together in a dashboard to communicate the results with others. The Display builder gives me access to all of the visualizations and discoveries I found along the way. I browse and drag the key visualizations onto the canvas to create the following dashboard. TIP: Need another visualization? No problem, add one on-the-fly using the Add discovery button. To share these insights with my team, I simply click the Share icon. I can email or download the dashboard as different file types. I can also post a tweet about the dashboard or share it as a link with other users in my Watson Analytics account. So just by starting with some simple, plain language questions I was able to explore and visualize my marketing campaign data to find valuable insights between the promotions and the sales that were generated. I can now use these insights and visualizations to drive my next marketing campaign to even more impressive returns by using what I learned right here in Watson Analytics. Now that you’ve seen the power of Watson Analytics in action, try using your own data sets and continue exploring the capabilities of Watson Analytics for your next marketing campaign. If you haven’t already, you can register and learn even more about the free version of IBM Watson Analytics here. http://www.ibm.com/analytics/watson-analytics/