No, this is not an article about fecking – drilling for gas and oil in the shale. It is about “drilling down” in big data. We have been using this term for a long time and it provides a useful metaphor for data analysis. However, we have thought of limiting ourselves to only a superficial degree, and this requires rethinking.
When data was not large and analytics relied on less robust hardware, we were only able to scratch the surface of our data, a practice that survives to this day. “Scratching” often means looking for insights at the end of business processes. So, for example, we look for churn signs next week or the next best offer, or we try to forecast the next sales deadline. All this is valuable but not enough.
If we are doing our job properly, then we should use powerful analytics to perform root analysis for better forecasting events, so that we can either avoid it altogether or increase our chances of success further.
Our prediction problem
What if you could go further into your data so that instead of searching for someone or a business that is about to leave your service (churn, nonvegeal), you can find those moments of serious danger And can fix a problem at source? You can – but it requires changes, not more hardware or better software. Those things are always welcome, but you have a different way of preparing the challenge.
Very often, we make assumptions about some aspect of the business and then collect and analyze data about it. This is a good approach, as long as the assumptions are valid and accurate, but very often they are not.
When we eclipse something, we’re building a hawk model that we consider to be reality, and that’s a good thing. Modeling is the heart of all types of progress in any field of human endeavor, but it is not something we do particularly well in business – with a few exceptions.
“We have a problem and must accept it,” writes Nat Silver in Signal and Noise. “” We like to predict things – and we’re not very good at it. ”
You may remember that Silver correctly called 49 out of 50 states in the 2012 presidential election. This man does not have a prediction problem.
Retailers may be an exception; They model heavily and they do a good job. Those customers gather and store data so that they can model the way they set up stores and plan the classifications they stock. Those models show the customer and traffic for an individual store very closely.
When it comes to online business and B2B business, we are not there yet, because it is a different and difficult challenge.
Doing two steps
Finding a solution in the online world starts with locating your model, before you make an assumption, and before you apply something. (This is not your business model, but your approach to customers, which is part of the business model.) This is surprisingly easy if you take a two-step approach to analytics.
Step Number 1: Create a realistic model of your business by asking your customers. I call it “discovering the truth of your moments”, and I write about it in my new book, Solve for the Customer, which will be available soon.
As you know, if you read this place often, a “moment-of-truth” is just about any time your customers expect you to fulfill a promise – whether it’s a product, company, or brand promise – Promise whether express or implied.
Step Number 2: Build customer-facing processes based on your moment-to-truth. Your process and supporting software will meet customers where they live, so to speak.
The best way to do this is with travel-mapping software, as it allows you to check all contingencies and define subtypes appropriately. It is also the logical place to define a matrix that will tell you if you are making your goals come true in your moments.
For example, customer onboarding is a good example of a moment-of-truth, and there are many analytics vendors who focus on the health of the customer based on how quickly customers fall below your learning curve.
For example, the longevity of the customer and how fast the relationship is between them, people at Scout Analytics told me. Knowing this, smart vendors deploy customer success managers to ensure that onboarding is fast and hassle-free.
You can identify the truth of such moments during your customer life cycle, and often those moments do not automatically require costly human intervention. Nevertheless, a moment-to-truth approach, good analysis rather than assumptions, enables a vendor to deploy resources where they would be most beneficial to both the customer and the seller.