Most business analytics today is useless. Not because the data is bad or the conclusions are wrong—but because managers don’t act on either. For analytics to reach its full potential, this situation is going to have to change, and fast.
To be fair, this problem isn’t really managers’ fault. Instead of providing useful, easy-to-apply insights, most analytics software drowns them in a deluge of charts and graphs they don’t have the training or the time to interpret. Sure, those charts and graphs might tell a sales manager that a particular employee underperformed last quarter. But rarely do they give any insight into why, or what (if anything) he or she can do about it.
This is the biggest missed opportunity in analytics today. Analytics-driven decision-making should be fast enough and accessible enough for the lowest-level employees to employ in their day-to-day work. We have the data to make this happen. Now we need frameworks to make that data actionable.
Let My Dataset Change Your Mindset
Before data can change how people act, however, it needs to be able to change how people think. As Hans Rosling said in his many excellent TED Talks: “does my dataset match your mindset? If not, one of them needs upgrading.” And in most cases, it’s the mindset that’s fallen behind, not the facts.
Today, people look at data analytics dashboards as a way of finding answers. But good data analytics should provoke questions and dialogue as well. One product is not doing as well in one region vs. another—why? Some members of the sales team are performing way better than the others—why? Good analytics software should point out these trends itself, and then suggest ways the manager can find out an explanation. It should encourage curiosity and flexibility, not just reinforce ideas that are already there.
But in the long process of turning analytics into action, this is only step one.
Information is Food—and We All Need a Diet
In another famous TED talk, technologist JP Rangaswami argued that to humans, information is food—we gather it, prepare it, and consume it just like we do our daily meals. The corollary, of course, is that we need to monitor the quality and quantity of the information we consume. We can’t gobble up the data equivalent of three full bags of Doritos each day and expect our brains to remain in good shape.
For many of us, that also means going on an information diet. A lot of the data our technology collects is the equivalent of empty calories, useless for creating action. For data to change your mindset, you don’t have to see all of the data—just whatever metrics and suggestions are most relevant. Analytics software should be able to answer the question: what is the most important thing to show the user?
At SpringML, we use next-generation technologies such as Salesforce Analytics and Apache Spark to transform data analysis and help users drive towards meaningful action.
Which brings us to my next point: the importance of mobility in analytics.
Bringing Analytics “Within Arm’s Reach of Desire”
In the first half of the 20th century, Coca-Cola president Robert Woodruff made his brand a household name by promising to bring Coke “within arm’s reach of desire” for soda drinkers across the world. That is, he planned to make his product widely accessible by meeting customers where they were at. To permeate all aspects of business decision-making, analytics needs to bring itself within arm’s reach of desire as well. And in the 21st century, that means moving to mobile.
Mobile is simply how business is conducted these days. If a manager wants to see how much business they did last quarter, they shouldn’t have to email someone to run that report. They should just be able to access it via a mobile app. Anything else will feel cumbersome to a worker used to accessing everything from their email to their bank account balance on their phone.
The corollary is, of course, that once you give managers access to analytics software on their phones, they’ll be that much more likely to use it regularly. Put analytics within arm’s reach, and its integration into everyday decision-making will happen naturally.
The Analytics Assistant
Imagine the below two scenarios.
Scenario 1: A sales rep chats with her trusted advisor before an important prospect meeting to brainstorm the best way to position a deal—she chats about the pain points of the prospect, their short term and long term objectives, preferences of their decision makers, etc. This conversation provides the sales rep with valuable insights and forces her to think through aspects she hadn’t considered before. The result is that she’s able to put the best possible proposal forward.
Scenario 2: A technical support engineer has just received a high severity case. This engineer is facing a demanding customer who needs the case resolved immediately in order to avoid business impact. As the engineer starts to research the case, she gets a message from a colleague who suggests a list of similar cases that were solved in the recent past and relevant knowledge base articles. She also discusses potential solutions that can help resolve this customer case. The support engineer is able to use this information and presents a satisfactory solution to the customer in a timely manner.
One might think that both the above scenarios are typical in any sales or service organization. However, the difference is that the trusted sales advisor and the support colleague are not humans, but intelligent virtual assistants capable of engaging humans in a conversational manner and providing valuable insights and suggestions.
Today most people are familiar with consumer technologies like Google Now, Apple Siri or Amazon Echo. However, these technologies have not made inroads into the Enterprise space. At SpringML, we are continuously investigating ways to bring rich machine-learning based insights to business users in an easy to consume manner. This includes providing a natural language interface so that a user could get those insights in a conversational style. SpringML is leveraging recent advancements in technologies around natural language processing and bots (e..g Microsoft Bot Framework) to make this possible.
Programmed to translate data analysis into actionable advice, these virtual assistants will make using insights from Big Data easier than ever. And all of them will help us reach our ultimate goal: translating analytics into action.
By Prabhu Palanisamy with contribution from the Hippo Reads research network.