Sign up

How to freshen up your profits with big data and predictive analytics

September 7, 2017

Avoiding waste and markdowns in the fresh food sector  Tweet This

If you have been shopping in a supermarket in the early evening you’ve probably seen the assistant with one of those sticker applicator guns. You might even have seen a few eager bargain hunters follow this person around as they start marking down products about to hit their sell by dates. While a few lucky customers are getting a great bargain, this end-of-day discounting of fresh goods is a big headache for retailers.

On the one hand, while having too many fresh food items on the shelves at least means you haven’t run out of stock and disappointed customers, discounting or throwing away excess stock can really bite into profits.

Reducing markdowns and write-offs

So how do we find the sweet spot where you neither lose potential revenue through unmet demand, nor reduce margins through markdowns and write-offs?

The answer lies in big data and machine learning.

Traditional methods of predicting sales based on past data are only a starting point because they result in conservative forecasts and cannot factor in missed sales due to items being out of stock.

With the new scientific field of predictive analytics, a computer algorithm is able to gather millions of data sets and build forecasts with a high level of granularity. This is because the algorithm is not just relying on historical data but real-time factors as well like the weather.

Do you want to know how well a special promotional price on tomatoes will work on a rainy day in Manchester?  The machine can tell you.

Well, actually, that’s not quite true. The machine will give you a forecast which, initially, might be a little off the mark, but that is the beauty of machine learning. As you feed back in all the new data you receive, the algorithm learns from its mistakes and gets incrementally more accurate.

Give it time

Just as humans need time to acquire and assimilate data, so too do computer algorithms.

When Evo Pricing is working with new clients we explain that this is a long-term process, not a magic button that produces results overnight. Once we have set up all the parameters and loaded the data, we find that we can get measurable improvements after approximately four weeks.

These improvements can be for KPIs in as many areas as you choose to track. In terms of replenishment we can go so deep as to tell you at what time during the day you need staff available to restock apples on the shelves, or when you need to reorder lettuces based on your suppliers’ lead times.

And, while it’s not possible to get stock levels of fresh produce 100% perfect, you can still use predictive analytics to decide the optimum end-of-day markdown for each individual product.

When I was a student, I used to haggle with the assistant with the price ticket gun and, sometimes, just to get rid of me, he would reduce something by 90%. Nowadays, that assistant would give me short shrift because he would have a list telling him the exact time to reduce each particular item, and the exact price to be used.  No more global discount one hour before closing time.

Supplier side innovation

Apart from internal measures to optimize revenue or margins on fresh food, there are also innovations to look at on the supplier side.

What if a large city supermarket could have a farm right on its doorstep, producing low-cost fresh fruit and veg year round, free of pesticides?

That’s exactly what, Bowery Farms, a startup near New York, is doing. In a massive warehouse just outside the city, they are pioneering mass vertical farming methods – producing food indoors in vertically stacked layers under scientifically-controlled environments.

So, quick lead times, no shortages due to bad weather, and pesticide-free products.

Combine that with machine learning and predictive analytics and you have a package that will put a smile on the face of your shareholders.

Are you a retailer worried about a high level of fresh food markdowns and waste? Talk to Evo Pricing about the kind of data-driven insights that are already helping companies around the world increase profitability by hundreds of millions of dollars.

To contact us go to

About the author

Martin Luxton is a writer and content strategist who specializes in explaining how technology affects business and everyday life.

Big Data and Predictive Analytics are here to stay and we have only just begun tapping into their enormous potential.

Hey! Was this page helpful?

We’re always looking to make our docs better, please let us know if you have any
suggestions or advice about what’s working and what’s not!

Send Feedback