top of page



Analytics is taking on corporations like a storm. Industry after industry and company after company are adopting analytics and seeing tremendous benefits in the form of streamlined operations and lower costs. But one industry seems to be lagging behind many others, and that’s grocery retail. This should come as a major surprise as they might be the one industry that can benefit enormously from the use of analytics. 

Net margins for the grocery retail industry tend to be quite low, typically ranging from 1 to 3 percent, depending on items. Their business is based on volume, and every little thing they can do to improve their margins multiplies across the thousands of items sold in their stores. This should be a reason enough to adopt analytics. 

In fact, many grocery retailers have been facing rising costs and shrinking margins over the past years, and if that’s not enough, grocery retail is going to be permanently disrupted by coronavirus. It certainly seems like a no-brainer to act. And, they're blessed with tons of valuable data such as procurement information, inventory movements, how much is distributed to each store, inventory within each store, products sold down to the smallest detail, what time the products are sold etc. The list goes on. This untapped arsenal of data creates an enormous opportunity to create growth based on analytics. Yet, so many are struggling to create value from their data and the reason is not so clear. A global study finds 85% of grocery retailers lack capabilities, technology and expertise to use insights to monetize their data and drive customer experience

Those numbers are staggering in the age of analytics. However, some are doing well and currently Walmart is probably the retailer that has managed to put analytics to work better than anyone else. They use big data analytics to make their pharmacies more efficient, improve store checkout, manage the supply chain with 500 million forecasts a week, to optimize product assortment, and to personalize the shopping experience. 

This is certainly not a big-do-well and small-do-bad type of scenario. In fact, LIDL, the gigantic german discounter, dropped SAP after struggling for 7 years to implement the system and lost a staggering EUR 500 million in doing so


The solution is hidden in the problem, like so many times. With the vast amount of data, and unparalleled opportunities to tap into data-driven growth opportunities, many retailers try to do it all at once, biting off more than they can chew, even LIDL. There’s a tendency (that must die) within the IT industry to bring holistic solutions that are supposed to solve all the problems organizations face, including bringing analytics to life. That’s a formula for failure, and even Microsoft themselves are taking a step backwards by chopping down their systems so companies can simply cherry-pick what suits them, instead of taking everything at once with all the time and cost associated with it. 

The formula for success in analytics is to start small. This is a gradual and never ending process. You can’t really finish with analytics, as there’ll always be something new. By starting small, organizations can master one thing and add to it as they become accustomed to using it. The keyword here is “using it. Remember, the number one reason why organizations fail at analytics is that they fail to get employees to use it. And when they plan on doing everything at the same time, the new way of working will be too much for employees to get accustomed to. Therefore, start small and get accustomed to using it before more is added. 


It can seem a daunting task to start, as the pitfalls are easy to miss (read LIDL). Many grocery retailers still don’t know the in-store inventory and the store managers order more inventory based on their experience and what they observe in the store, rather than using fact-based data. Sales numbers are updated overnight, and if they haven’t been able to observe everything, they won’t know until the morning after what to order from the warehouse. As orders are done manually by store managers, mistakes in quantity are common, resulting in overstocked stores or stockouts; both being lost money for the retailer. Additionally, as analytics are not at work, knowing when a busy weekend might be on the horizon is something someone must remember. And how much to order of what products for that particular busy weekend, is something someone is trying to see in their precious Excel document since the same time last year.

The best way to start is to adopt advanced demand forecasting techniques. Yes, you read that right. To forecast your sales properly is The Holy Grail of all analytics and should come first. Most of your other insights are worthless if you can’t predict your sales. Therefore, the starting point is advanced demand forecasting. This is where data scientists use advanced times series models and sophisticated machine learning algorithms, and apply them on historical data to forecast the future. Forecasting has improved tremendously the past 10 years and is finally becoming a reliable tool for organizations. The biggest forecaster in the world today is probably Walmart, but most retailers lag far behind, and have little or no idea how much they will sell next week, let alone next month, and rely heavily on experienced store managers. 

Most will start searching for a sophisticated forecasting software to do the task, but that’s where most fail. Grocery retailers will be forecasting every single product in every single store, which can be millions and millions of forecasts. But there’s no such thing as a one-size-fits-all approach. Each data set might need different forecasting models and different algorithms, and there’s still no software that does that. Just ask LIDL, they lost EUR 500 million trying. And if software was able to do it all, then Walmart would probably use it, but instead they have an army of well trained and dedicated data scientists to accurately forecast their sales. 

We at SUMO Analytics advise our customers to start small and start with forecasting, and as an analytical agency we create sophisticated demand forecasts for all our retail clients. As they get accustomed to using these valuable insights, adding different predictive analytical models turns out to be far more simple than doing it all at once. 

But time might be running out for some retailers. Many are jumping on the analytical train and those who don’t will be left behind. Imagine if all retailers, except one, have all adopted sophisticated advanced forecasting techniques and are literally seeing clearer into the future than ever before. The one that hasn’t will eventually have no chance of catching up. In fact, a research by McKinsey revealed that there's an enormous benefit of being among the first adopters of analytics, but at the same time it can be fatal to be one of the last as it will become harder and harder to catch up. 

The time for analytics is now. Tomorrow it might be too late.

Interested in knowing how advanced demand forecasting can help your organisation


bottom of page