However, the analytics market is still noticeably dominated by visualization and business intelligence tools like Tableau, Qlik, and Birst, and not by prediction tools.
If predictive analytics is the next best thing, why isn’t everyone using it? What’s the solution?
The Value of Prediction
Simply put, predictive analytics tells you what will happen in the future. And in some cases, simply knowing what’s ahead is enough to help your business better position itself to succeed in that future.
For example, if you’re in the business of making home loans and you predict a potential borrower will default, you can decline to make the loan or increase the interest rate to minimize your risk.
If you’re an airline and you predict that tickets for an upcoming flight will sell out, you can increase the ticket price to maximize your revenue.
Or, if you’re a hedge fund manager and you predict a big sell-off in global markets, you can pare down stocks or boost defensive ones to protect investments.
Where Prediction Fails
The dirty secret about predictive analytics is that, for the majority of businesses, prediction alone holds little value.
That’s because the actions a business should take in response to a prediction are typically less cut-and-dried than in the cases of the loan company, airline, or hedge fund. Raising prices on an in-demand flight is a no-brainer. So is declining a risky loan or shifting investment to avoid losses from stock market turbulence.
But what should you do when a customer is predicted to cancel their subscription to your service? You need to convince them to stay, but how? How about when a machine is predicted to fail? Fix it fast, sure. But how? How about when a delivery to a customer is going to be late?
In each of these examples — and countless more — there are many potential solutions and it’s not clear which will work best.
And if you don’t know how to act on a prediction to resolve an imminent problem (or take advantage of an opportunity) what good is it to have the prediction in the first place?
An Example: Customer Churn
Consider the simple case of customer churn. Predictions will tell you which customers are likely to take their business elsewhere.
Knowing this, you can certainly accept your fate and just make changes at the strategic level of the business to cushion your losses.
For example, you could choose to shift your investments (ad spend, promotions, service, etc.) away from those at-risk customers. Somewhat risky, as you may just accelerate their departure.
You could also step up new customer acquisition efforts to offset your predicted customer losses. Getting new customers, of course, is difficult and very expensive.
Because of these strategic changes, you may be more profitable at the end of the quarter or year than you would have been if you didn’t know who was going to churn.
But most business don’t consider “may be more profitable at the end of the quarter” a winning strategy in this day and age. And simply trading existing customers for new ones each year is not every leadership team’s favorite path to building a long-lasting business. Especially in smaller markets (like B2B) with a relatively finite universe of clients.
A much sounder and more profitable long-term strategy than acceptance and strategic mitigation of future losses is to proactively prevent those losses by stopping the predicted churn in its tracks.
To do that, you need to know what solution or “remedy” will succeed with each individual customer. But if you don’t know why they are leaving, you won’t know the right actions to take to ensure loyalty.
If, for instance, a customer is leaving because her product keeps failing — but you don’t know that — your story will play out in one of two common scenarios.
One scenario is you guess what might make her, and the rest of your at-risk customers, stay. Maybe you offer a longer contract and a free add-on. This “guess then spray-and-pray” approach will fail to keep her and your other customers because you haven’t addressed their actual pain points. Who wants a longer contract or free add-on for a faulty product?
Or you can try to find the right solutions by digging through the data — combing through spreadsheets or looking for “insight” in charts and graphs — to determine what churning customers have in common, why they’re leaving and therefore what interventions might be the most effective. This requires expertise and effort that most businesses can’t afford, and time that customers aren’t likely to give you.
So what’s the solution?
Prescriptive Analytics: Taking Prediction a Step Further
Prescriptive analytics goes beyond prediction, telling you what specific solutions or “remedies” to implement to prevent predicted problems or threats.
When you have predictions and remedies, suddenly every business — not just the few examples with cut-and-dried solutions — can capture the immense value prediction can add to the enterprise.
That said, not all prescriptive analytics tools are created equal. Many rely on remedies that people within the business have supplied and inputted into the predictive analytics software.
For example, a leader in the customer service department decides that all at-risk customers in a certain price tier should be given a 20% discount. So when those customers call into the service center with an inquiry or problem, the agent’s software tells them the customer is at-risk and to offer the 20% discount.
This is still the “guess and spray-and-pray” approach, but now it’s programmed into the software for greater convenience. But it’s not more effective at keeping customers.
A select few prescriptive analytics tools don’t rely on guesswork programmed into their software. Instead, they analyze your data to determine the most effective solution for each problem.
For example, Emcien finds the reasons behind every prediction and shows them to you in plain English so you know exactly what remedy to provide each customer.
So in the case of the customer with the faulty product, Emcien would tell you the reason she is predicted to churn is the 6 calls to tech support in the last 2 weeks to fix the SIM card in her mobile device.
Armed with that knowledge, your team can proactively send her a new SIM card and $10 off her next month’s bill. You’ll probably manage to keep her (saving yourself the high cost of finding a replacement customer) while doing it for much less than the 20% off guessed by the service department.
What’s more, you will unlock the true potential of predictive analytics that everyone is talking about, but that few have actually harnessed.