Keboola in Sklizeno
BLOG / FEBRUARY 2019 / 2 HOURS
Sklizeno: Making purchasing, sales and management more effective
There is no need to introduce this retail chain especially to healthy lifestyle fans and lovers of local healthy products. It is a chain of Czech as well as international shops offering food with clear origin and guaranteed quality. Fist shop was opened in Brno in 2011, today it counts as much as 46 branches of Sklizeno Food (including purchased “Náš Grunt”).
We were talking with the business director of Sklizeno Jan Langer on theme how can data help with their dynamic expansion and global management of whole network.
From my point of view everything is about efficiency. For any case, I look at two parameters: the efficiency of the work and the efficiency of the lease. In terms of work efficiency - it has been mainly about setting monthly targets so retail managers as well as regional managers are able to see how they were standing on a simple dashboard. In the next step, we want to establish a reward system to move from a purely controlling tool to a motivational tool.
For me it is important that I monitor both aspects - the efficiency of the work and the efficiency of the lease. That's why I know precisely where the branches have poorly arranged rent and I have to terminate it or negotiate better conditions. At the same time I see where I have low efficiency of work and I have to push the staff.
Can you tell me more specifically how exactly the data processing help with the efficiency?
There are plenty of examples. One of them is employee revenue indicator. We didn’t know this before. But today I know precisely that the best shop is doing 300 thousand and the worst 83 thousand per head. With this information, I can address the causes and seek improvements.
We also captured the stocktaking that we were not able to control at all. There were shortfalls of hundred thousand in some shops, and we found it very complicated to find out why. Thanks to the Keboola data, we have discovered that the manager does not make inventory checks and the merchandise is missing.
The third example is the number of products in the store. I let Keboola pull this metric, and thanks to it, I know exactly where to add one more shelf to the store as the metrics taken elsewhere prove that the sales are better that way.
The same goes for the assortment: I am now able to tell at first glance which branches sell daily purchases, which sells fresh food and which are more of a delicatessen or a gift shop. Our goal is the first option. Thanks to the data, we can beautifully benchmark, give weaker stores better ones as an example and to focus attention wherever it is needed.
We use WinShop as an ERP. As a basic ERP, it works well, but reporting was a nightmare. Originally, the data from it were flipped into Excel, and then we glued it together over long nights for top management reporting. Basic reports of the type of comparison between stores, year-on-year, by segments, etc. However, we had less and less time for the frequency to be reduced to one report in six months.
It is also very laborious to associate other ERP programs and systems with what we wanted to buy, for example, for managing customer accounts or electronic price tags. Such news sounded great, but the obstacle was the creation of a bridge to WinShop.
That's why we decided to switch to Keboola, flush the entire existing database into it, and start joining other things right there, so it serves as a data transceiver for other systems, such as Quant. We already have this entire data infrastructure on Keboola, so we do not have to involve our internal IT in most projects at all.
In addition, Keboola brings us tremendous flexibility in preparing data for reports, making them quick and editable.
The project was originally taken over by our main controlling person, plus an IT person was involved and somehow it worked. But we soon found out that there was a lack of business engagement, so the projects did not focus on financial results. From IT and controlling, the project was taken over by the management, basically me and the owner Petr Borkovec. From that time we did only projects with quantifiable benefits associated with specific business metrics. Therefore, each project is preceded by cost-benefit analysis and prioritized.
My goals as a business director are based on profit and everything turns around that. I have already mentioned efficiency reports that help me to manage regional managers and push their performance on the basis of continuously updated numbers. I also save time: I go through a feedback once in a quarter with every shop, the preparation of documents takes just about thirty minutes.
In general, I have more factual arguments in my pocket now. For example, stores often have aversion to accept new products because we have a lot of them, and at the same time aversion to remove products they like. Their arguments are based on emotions and on relationships with suppliers. However, I can oppose and decide on the basis of the results, because I have to process the statistics of goods by turnover. All in all, there were some conjectures, calculations, correlations. Today, I can push a lot more to business instead of negotiating to focus everyone's attention on finding solutions.
So, you are satisfied with the tool, right? How do you evaluate your customer experience?
It's a big benefit for me that Keboola is doing things very quickly. I'll give you an example: We're currently working on improving your purchase. We knew we could not order to the stores effectively – each store orders themselves, there are unnecessary losses. At the same time, we felt that it must be possible to calculate it, in today's digital age. We turned to the boys from Keboola and they surprised us with a very quick reaction. It was moving within three days and a pilot predictive model was born very quickly and we are testing it now.
Sklizeno: How We Use Data to Be More Effective
This retail chain needs no introduction, especially for those who appreciate a healthy lifestyle. This chain of Czech (as well as international) shops is known for offering locally sourced food products guaranteed to be of the highest quality.
Sklizeno Food's first shop was opened in Brno in 2011, and it has since grown to a network of 46 branches (including their Náš Grunt locations).
We sat down to have a chat with the business director of Sklizeno, Jan Langer, about how data is enabling their explosive expansion and assisting in the global management of their network.
From my point of view, everything is about efficiency. For any case, I examine two parameters – the efficiency of the work and the efficiency of the lease. Regarding work efficiency, it has been mainly about setting monthly targets so that retail managers, as well as regional managers, can see how they were standing on a simple dashboard. In the next step, we established a reward system to move from a purely controlling tool to a motivational tool.
For me, it’s essential that I monitor both aspects – the efficiency of the work and the efficiency of the lease. That way I can know which branches have poorly negotiated rent so that I can either terminate it or negotiate better conditions. At the same time, I can see which locations have low efficiency of work, so that I can work with management to make the location more productive
I have plenty of examples. First off, employee revenue indicator. We weren't able to know about this before, but thanks to Keboola, I know that our best shop is running at 300 thousand Czech Crowns (around $13,400 USD) per person, and the least efficient is running at 83 thousand Czech Crowns (about $3,700 USD) per person. With this information, we have a clearer idea of where we can improve.
Second, we're able to capture inventory data that we weren't able to get previously. Some shops used to report shortages of 100 thousand Czech Crowns (around $4,450 USD), and it was very difficult to figure out why. Thanks to Keboola data, we determined that the store management wasn't taking inventory often enough and wasn't aware that merchandise was going missing.
Third, we're able to delve more into the number of in-store products. After having Keboola pull this metric, we were able to use store data to know exactly where to add another store shelf.
We've even used Keboola to help us with product assortment. At a glance, we're now able to tell which branches sell more daily purchases, which sell more fresh food, which are more of a delicatessen, and which locations are more of a gift shop. Thanks to the data we have because of Keboola, we're able to benchmark, give weaker stores examples of practices the stronger stores are employing, and focus attention on our goal of having every store deal mostly in daily purchases.
We previously used WinShop as an Enterprise Resource Planning (ERP) system. As a basic ERP, it works alright, but reporting was an absolute nightmare. Initially, we would flip the data we got from it into Excel, and then we would spend very long nights gluing it all together into reports for top management.
These basic reports had all kinds of comparisons, including between stores, year-on-year, by segments, etc. However, since we kept getting less and less time to work on them, we had to drop the frequency of reporting down to one report every six months.
We also found it extremely time-consuming to link other ERP programs and systems with other systems (like customer account management or electronic price tags). Improvements in our system or practices always sounded like a great idea, but creating a bridge to WinShop was always an obstacle.
Once we switched to Keboola, we dumped our entire existing database into it and immediately started linking it to other systems, like Quant. Keboola provided the complete data infrastructure, so we don't have to involve our internal IT team into every project we undertake anymore.
As if that wasn't reason enough, Keboola provides us with incredible flexibility when it comes to data reporting. We're able to create them much more quickly, and they're even editable.
How was the transition process to Keboola? Was it readily accepted in your organization?
Before Keboola, our main controlling person, plus one other IT person, initially oversaw projects, but we later found out that there was a lack of business engagement, so the projects didn't focus on financial results. After a while, the management team, consisting of the owner, Petr Borkovec and I, started overseeing projects, and projects were only approved if they had quantifiable benefits associated with specific business metrics. We started using cost-benefit analysis and prioritization before beginning a project.
As a business director, my goals are based on profit and everything that revolves around that. I've already gone into how helpful efficiency reports have been in assisting regional managers to stay on top of their numbers, which are continually updated.
It also helps me save time. Every quarter I sit down and give every single store feedback, and the preparation of all those documents only takes about 30 minutes.
In general, Keboola makes it possible for us to go into a situation armed with more facts. For example, without data, some stores might be less likely to want to sell additional new products because they feel like they're already selling too many, or not want to remove products that aren't selling very well because they like them. If we have the data to back up business decisions, we don't have to take time allying these fears and reassuring them that this is what's best for their stores. This advantage allows us to spend less time communicating with store management and frees up more time for other essential tasks.
So all things considered, you'd say you're satisfied with Keboola, right? How would you evaluate your customer experience?
Keboola's biggest benefit is that it can get things done quickly. I'll give you an example: We knew we couldn't get orders to the stores efficiently, but if the stores order themselves, there are unnecessary losses. In today's digital age, we knew that there should be some way to calculate this. We turned to the team at Keboola, and they surprised us with an incredibly quick response. We had everything up and moving in three days, and now we're testing a pilot predictive model that we were able to put together very quickly.