By Todd Johnson, Executive VP and Chief Operating Officer, Saama Technologies
In the last 9 weeks I have spoken to 27 large companies in the retail, high tech, insurance, food and beverage, automotive, banking, and transportation industries. These meetings have been mostly at the “C” level, VP and above. From my conversations about their big data strategy, three common themes have lifted from these discussions:
Looking for Trends from Disparate Data Sets
[av_dropcap1]W[/av_dropcap1]hether you call it BI, analytics or advanced analytics, more companies are looking for ways to bring together disparate data sources to show the trends in the business to the C Suite. The problems are old school methods, disparate data, data integration challenges, and data quality challenges. The output is mostly old school dashboards… but this time it has an emerging twist: companies are making a real effort to employ predictive analytics tools to “guide their actions.” Companies want to go beyond reporting the facts and to make suggestions on how to address predicted issues. We have seen a very popular example in analyzing a company’s installed base of customers looking for likely candidates for churn. We have completed projects in the financial services industry, insurance and technology sectors with really good results. The initial analysis work identifies churn patterns based on historical data. Then profiles are built based on the analytics and are applied on an ongoing basis to the current customer pool. Likely churn candidates are identified and then based on their profile; pro-active marketing campaigns are implemented to encourage retention. Our success rates have been quite high, and the ROI can be staggering.
Giving Visual Data Access to Decision Makers
[av_dropcap1]C[/av_dropcap1]ompanies are working hard to move the data and the analytics processes to the edge of the organization where the decision makers are. Companies want financial analysts, HR managers, supply chain specialist, claims processors, product managers, etc. to have access to analytical tools to do the “what if” analysis themselves – not having to submit a report request to IT. The rise in popularity of tools like Tableau (http://www.tableausoftware.com/) QlikView (http://www.qlikview.com/) and Spotfire (www.tibco.com) are the direct result of these trends. Most of our major customers have multiple projects active today that support this trend. It is a very important part of many of engagements we are working on. In addition to deploying new tools, we are working on a number of other approaches looking to achieve the same outcomes – analytics driven tools at the point of business decision maker. Many projects we are running are looking to accomplish this by mobile enabling their suit of dashboards. The growth of tablets has really driven this trend in the enterprise. Again, there is a very high occurrence of this type of projects in our installed base and in our pipeline.
Everyone is Evaluating a Big Data Strategy
[av_dropcap1]C[/av_dropcap1]ompanies of all shapes and sizes are in various states of evaluating or implementing Big Data capabilities. Everything from fraud analytics, to churn analysis, to employee sentiment, to key opinion leader reporting, predictive failure analysis via log files; people are looking at how this relatively new set of data tools can add value to the businesses. People are thinking of Big Data as way to add new dimensions to existing analytical processes by incorporating unstructured or very large data sets, and are looking to correlate these new sources with their existing structured data. I see lots of “proof of concepts” that are focused on understanding the infrastructure like Hadoop (www.hortonworks.com) or Google’s Big Query (https://cloud.google.com/products/big-query), but the projects that I think show the most promise are the ones that are use case focused. Market basket and product trends based on point of sale data, correlating social data with CRM data to better assess channel effectiveness for retail products, employee sentiment correlating Glass Door, Facebook unstructured data with internal Jive or Yammer unstructured data – the list is getting long, but the people trying to solve to real business questions are making the most progress.
It may only be a snippet in time reflecting where we are in the market, but the consistency of these three topics really surprised me.