Data Management at Sklizeno: Making Purchasing, Sales and Management More Effective

Data management at Sklizeno: making purchasing, sales and management more effective

In the Czech Republic, Sklizeno needs no introduction to healthy living enthusiasts and fans of local, organic products. The chain of grocery stores stocks mostly Czech but also imported food, with clearly marked origins and guaranteed quality. The first store opened in Brno in 2011; today, Sklizeno Foods, together with the Náš Grunt chain, which the company acquired, operates 46 grocery stores in total.

We talked to the company’s Sales Director, Jan Langer, about how data can help in the dynamic expansion and management of the network.

For what do you mostly use Keboola Connection?

From my point of view, it all comes down to effectiveness. Above all, I look at two parameters: labour effectiveness and rent effectiveness. As for labour, I first focused on setting monthly targets so that the stores and regional managers can have a continuous idea of how they’re performing. In the next step, we want to link this to remuneration, which will turn a purely monitoring tool into a motivational tool. I look at labour effectiveness and rent effectiveness side by side, which clearly shows me where the rent has been agreed at the wrong level. I thus have to negotiate a better contract or terminate it. I also know where I have low labour effectiveness, in which case I need to be stricter with employees.

Could you be more specific about how data helps you with effectiveness?

One example is sales per employee. We didn’t know this indicator before. Today, for instance, I know that the best store generates CZK 300,000 per employee, while the worst one just CZK 83,000. With this information available, I can look for the causes and find ways to improve the situation.

We’ve also improved our stocktaking procedures, which we couldn’t check at all at the start. Even if deficits of hundreds of thousands were incurred in a store, we had a hard time finding out why. In the end, we discovered that the store manager didn’t do any stocktaking at all, and that goods kept disappearing from the store.

A third example is the number of products in a store. I’ve had this metric uploaded to Keboola, and now I know in which stores where an extra shelf needs to be installed because stores elsewhere have one and generate higher sales.

The  last example has to do with our product range: at first glance, I can now tell which stores sell groceries daily, i.e. fresh food, and which tend to be delis or collection points. Our goal is the first option. Data allows us to do a good job benchmarking the stores and focus on those that need it; we can set the good stores as examples for the shops that are performing less well.

How did you process data before?

We used to run our ERP on WinShop and then converted data from that system to Excel files. At night, we produced reports for the top management from those inputs so that we could come up with conclusions, compare stores with each other, year on year, by segments, etc. But we had less and less time for that, and the reporting frequency gradually dropped to one six-monthly report.

Another drawback was that we kept needing to connect new stores to the system as we acquired other players and consolidated the market. Concerning capacity and financial aspects, the WinShop-based reporting was no longer feasible. What’s more, it always required a lot of effort to integrate this database with other programmes and systems, which we wanted to purchase to manage customer accounts or electronic price tags. Often, the systems and software available looked very promising, but it was a big hassle to create a bridge to connect them with WinShop. So, we decided to switch to Keboola, migrate our entire existing database, and start connecting additional software products to it.

How did the migration go? Was it easy for users take to get used to Keboola?

Originally, the project was managed by our Head of Controlling, who had one IT guy to  help him. Somehow, the project worked. However, we soon realized that there was a lack of business drive, and it wasn’t aimed at achieving financial results. Therefore, the management, specifically Petr Borkovec, as the owner, and I, took over the project from IT and Controlling. We started implementing only projects for which we could clearly calculate the benefits. We decided to allocate income and costs to every task. Before we started anything, we checked the cost-benefit ratio. We leave the IT aspects to Keboola; our IT is actually not involved in it at all.

Where do you see Keboola’s greatest benefit for your management position?

My goals as Sales Director are based on profits, and everything revolves around that. I’ve already mentioned the effectiveness reports that help me manage regional managers and push them to perform better, based on continuously updated data. This also saves me time: on a quarterly basis I give important feedback to the stores, and it only takes me about 30 minutes to put together the source materials and information. Before, it took me a whole day.

In general, I’ve more fact-based arguments up my sleeve so to speak… For example, stores often have an aversion to accepting new products because we’ve loads of them. On the other hand, they’ve an aversion to getting rid of products that they like. Their arguments are based on feelings and their relationships with suppliers. However, I can review that and decide based on their results, since all my statistics and data are prepared based on sales. Before, basically there were arguments, calculations and correlations. Today, I can drive the business forward much more, and focus on looking for solutions.

So, you seem to be happy with the tool; how would you describe your customer experience?

It’s a huge benefit that things are happening very quickly with Keboola. For example, right now, we’re working on improving our purchasing process. We realized that we’re not ordering products for our stores efficiently – each of them places orders individually, which leads to unnecessary losses. And we also felt that in today’s digital world, this had to be calculated. So, we talked to the guys from Keboola, who took us by surprise with their incredibly fast response. Things started moving within three days, and very quickly a pilot prediction model emerged that we’ve now been testing.

In retrospect – is there any advice that you would have liked to have had at the beginning, when you started with Keboola?

I’ve already mentioned one key finding for me – on the client side, Keboola must be managed by the business guys. I would definitely recommend that for others – don’t assign responsibility for Keboola to any other function because the desired financial benefits will be achieved only as part of a business.

Thanks for your time. We wish you and the entire Sklizeno team every success!

‍