The Case Against Quick Wins in Predictive Analytics Projects

By Greta Roberts, Feb 09, 2016

When beginning a new predictive analytics project, the client often mentions the importance of a “quick win.” It makes sense to think about delivering fast results, in a limited area, that excites important stakeholders and gains support and funding for more predictive projects. A great goal.

However, implementing a quick win for a predictive analytics project can be difficult. There are at least two challenges, which I’ll describe, when taking a traditional quick win approach to predictive analytics projects.

Challenge #1: Predicting Something that Doesn’t Get Stakeholders Excited

Almost daily I hear of a new predictive project that was limited in scope, allowed people to dip their toe in the predictive water, and get a quick win. Too often, these projects predicted something that stakeholders didn’t care about, or couldn’t act upon.

Examples include the following:

  • Predicting which Colleges and Universities yield the highest performers. The problem here is that results of this prediction can lead to uncomfortable questions – such as are they also the most expensive schools? Does only a certain economic class of person attend these schools.

Instead of gaining a quick win, these predictions lead to discussion of issues like economic discrimination, and make HR and executives nervous. They often decide to ignore their newfound ability to predict performance, they don’t implement the prediction and the project doesn’t advance the case for more predictive projects.

  • Predicting a “Middle Measure” Like Engagement. What is the problem with this quick win? While HR thought the project was a winner, the results failed to earn excitement from business stakeholders and didn’t advance the goal of gaining additional support and resources for more predictive projects.

In this instance, executives had seen little or no correlation between engagement and actual business results at their own firm. Imagine trying to sell the VP of Sales on predicting engagement of their sales reps. At the end of the day their employees aren’t hired to be engaged, they are hired to do their job and sell.

Challenge #2. Quick Wins Shouldn’t Mean Tiny Data

In non-analytics projects you’re able to do a pilot with a small amount of people and data. You can focus on a small piece, a sample, something light, less expensive, less risky and less time consuming before you fully commit. An example would be piloting a piece of software. You could install it for a small number of people and gain their feedback before making a broader commitment. Pilots work great for small sample sizes and testing things with just a few people.

When you think about predictive analytics projects, though you want a quick win, you still need to find a project with enough data to conduct a great predictive experiment. To be predictive your models need to find patterns, and patterns require sufficient data. It doesn’t make sense to do a predictive analytics pilot on tiny a bucket of data. Rather than reducing the amount of data, it is better to reduce the scope of the prediction.

An example: Instead of doing a predictive analytics pilot project to predict flight risk for all jobs in your Chicago office, maybe it would yield better results to keep the scope small and targeted by predicting flight risk for a single role that has a lot of people in it.

Ask your data scientist for their guidance on how to frame your quick win project to keep the project scope smaller, while giving the data scientist a reasonable amount of data to optimize your chance for success.

For your predictive projects, “quick” isn’t enough of a win

Instead, you want a quick, implementable and exciting win that people care about.

The only way to get a quick, exciting win is to start with a project that predicts something that either saves or makes money for your company. Find a project that solves an existing business problem. Remember what predicting does for your organization. Accurate predictions do a better job at decision making so that you have better end results. End results are the only thing that will get people excited and will be implemented.

Think of banks that try to “predict” whether or not to give you a mortgage. They want to do a better job of extending credit only to people that can pay their mortgages. They’re not doing this to predict who will be engaged as a customer.

All your predictive projects should be ones where you are saving or making money. Do a project where you can demonstrate that your model worked and saved money on an important measure. Often this is a line of business problem, not an HR problem.

Results are the only kind of win that will get business stakeholders excited and move your efforts forward.

About the author

Author photo

Greta Roberts is an influential pioneer of the emerging field of predictive workforce analytics where she continues to help bridge the gap and generate dialogue between the predictive analytics and workforce communities.

Since co-founding Talent Analytics in 2001, CEO Greta has successfully established the firm as the recognized employee predictions leader, both pre- and post-hire, on the strength of its powerful predictive analytics approach and innovative Advisor™ software platform designed to solve complex employee attrition and performance challenges. Greta has a penchant for identifying strategic opportunities to innovate and stay ahead of the curve as evident in the firm’s early direction to use predictive analytics to solve “line of business” challenges instead of “HR” challenges and model business outcomes instead of HR outcomes.

In addition to being a contributing author to numerous predictive analytics books, she is regularly invited to comment in the media and speak at high end predictive analytics and business events around the world.

Follow Greta on twitter @GretaRoberts.


Six Models for Organizing Analytics

Best Practices

Why organize for analytics? Because companies that think strategically about analytics achieve the biggest impact from analytics. See which model is right for your organization.

Read the ebook »