KPI execution - business case

Adam Beniowski
Adam Beniowski
20 October 2020

How does Artificial Intelligence support searching for growth in mature organizations?

The template for business development looks very similar in most production companies. With product development and marketing as the foundation, it is then necessary to firstly build the distribution of products, and then take care of the quality of the exposition. This means that we start by getting our product to the point of sale, and then we have to make sure that the shopper sees it in the store, preferably in the right place and neighborhood, and with the right communication. The store’s “shelf” is the spot where the act of purchasing takes place, and which builds the real strength of our products and companies.

That is why a number of organizations invest a great deal of their financial and human resources in defining and implementing quality KPIs, which describe the most effective model of exposition, both standard and promotional. These are periodically repeated internal processes that last for multiple months, concerning topics such as product distribution matrixes (i.e. what type of product should be found in which type of store), a schedule for introducing and distributing new products, or targets describing the expected shelf shares. These discussions engage all market departments, Marketing, Trade Marketing, Category Management, and finally Operations – or in other words Sales.

While product distribution to stores can be verified at the level of sales systems or through data provided by companies such as Nielsen etc., the exposition quality parameters are most often based on measurements taken by Sales Representatives or Merchandisers.

Problems begin when both defined distribution and exposition targets are executed, but sales or market shares decrease. One of our clients encountered such a situation. After a few conversations, he decided to test the PRO.Display application, which, with the use of the Photo Recognition technology supported by Artificial Intelligence algorithms, verifies the given exposition parameters, based on photo analysis, with an accuracy of over 99%. The goal was to verify the declarative quality of the promotion’s execution vs the reality in the store.

Was it worth it? It turned out that in the month during which the pilot of PRO.Display use was carried out, the execution of quality KPIs, which was often over 90% according to the sales representatives’ previous self-assessment, was proved by PRO.Display to be, in some cases, only about 20%. This explained the reasons for decline in market share, and at the same time defined real areas for improvement and growth.

The reasons behind the differences between the report prepared based on Artificial Intelligence and to the previous reports prepared on the basis of the self-assessment of sales representatives, is a topic for a separate post.

This pilot program and verification of the real market situation is one thing, but bringing about actual change is another. It turned out that the precise determination of the initial situation allowed our client to prepare effective correction plans. In turn, thanks to the support of PRO.Display - an application which verifies the actual progress of the project, this translated into real exposition improvement.

The attached chart shows our client’s results right after the pilot.

Wykresy-aniomowane---PRO-v2

The first month of systematic work with the use of the PRO.Display application was December 2015. We compared the results during that month to the results from June of the following year, i.e. after 6 months. The target execution level improved in every measured KPI category. Shelf shares: the number of stores meeting the target increased from 44% to 65%, the number of stores with timely introduction of new products increased from 45% to 51%, and the presence of products in the hot zone increased from 63% of shops to over 91%.

In our client’s case, this of course translated into an increase in sales. As the client himself calculated, stores that met the quality KPIs showed a 20% higher sales dynamic.

Our main conclusions from this project are:

  • Defining proper exposition KPIs and preparing an effective implementation plan requires knowledge of the real situation in stores
  • The implementation itself requires support in the form of systematic and reliable tools that verifies real changes in stores.
  • Only then can we expect the development of our business, as exemplified by our client’s hard market data.

 

TAGS: photorecognition, prodisplay, aiinretail, ai, businesscase, kpi