To realize value from retail analytics, context is critical

Over the course of the past 13 years, we have worked on many analytical projects and programs delivering real value for some of the world’s most advanced retail and consumer goods organizations.

Over that same period, we have also seen many technological and analytical terms, trends and themes appear and then fade with little lasting impact other than client fatigue as they search for that silver bullet that was promised.

We can all piggyback on what is hot – “we are a [insert one of Machine Learning/Artificial Intelligence/Deep Learning/Blockchain/Predictive/Unicorn] business now”, in the hope of tapping into the technological zeitgeist or boosting our value. However, the truth is that, no matter what we might say we are, if we are in the business of providing expert retail analytics then there are usually three simple rules to ensure the successful development of an analytical solution^*:

1. Frame the challenge with context

• Identify challenges that are important for the business’ success.

• The tools are constantly improving, but they can only help you get to the target faster, not create the target for you, so context is critical.

• Ensure success can be measured by current key performance indicators that the business manages and tracks.

2. Ensure we have the right data for the challenge

• This stage can sometimes take longer than it does to build the analytical models, but garbage in does equal garbage out no matter how good the modeling is in between.

• The right data is foundational. Have people capable of understanding what is core, what is important, and what is nice to have.

3. Apply the right tools and technologies to solve the specific challenge

• Ensure that our data and decision scientists have a good grasp of the market, or can collaborate effectively with subject matter experts who understand the market, to develop the right solution for the problem.

The only way we can truly set ourselves up for success is when we invest in people capable of gaining a deep appreciation for the problem’s context, usually gleaned from extensive domain expertise, and in combining this with analytical talent capable of finding and harnessing the right data, not all data, and the right tools, not any tool, to solve the problem.

As we continue to be challenged to deliver effective solutions to complex retail and consumer goods challenges, requiring ever-more sophisticated modeling techniques, our expert team will continue to bear these three simple rules in mind as we focus on creating client value from retail analytics.

*Please note, this should not be confused with what it takes to successfully deliver and implement a retail analytics solution within the business, but rather to develop working prototypes that can, with the right implementation wrapper, solve the client’s challenge.