MySales Labs implementation at Eko Market

‘Those who are directly involved in the ordering process, admit that the MySales tech is really effective’ — Alexey Kruglikov

Accurately predicting sales and optimizing inventory is the key to success in the retail segment. The innovative MySales technology supports in this mission the leading grocery chain in Ukraine, Eko Market ( It is currently implemented across all the stores of the chain for one of the most difficult product groups to plan: fresh products. In the near future, it is planned to extend MySales tech to all product groups available in Eko Market stores.

Alexey Kruglikov (Head of Inventory Management) and Sergey Pritykin (CIO) from Eko Market shared with us their impressions on MySales and mentioned the future plans of the retail chain.

Alexey Kruglikov

What was your main motivation behind implementing MySales? What were your expectations before the start of the project, and to what extent were they met?

Our main goal was minimizing the so-called “lost sales”. Due to suboptimal sales forecasting which led to bad decisions in the ordering area, we had problems with assuring the presence of demanded products in our stores. We were searching for a solution to that problem which will not lead to overstocking, being an issue especially for short shelf-life products.

What were your KPIs?

First of all, minimizing forecasting errors which led to wrong ordering decisions. Automating the process is crucial, since many mistakes were driven by human errors, for example negligence due to haste. We wanted the orders to be placed based on fully automated sales forecasts. At the same time it is important for us that the system enables us to analyze in detail the forecast and order recommendation.Secondly, we do our best to isolate our employees from repetitive work and create the environment in which they can grow in areas which can not be automated.

Last but not least, we aimed to improve the quality of service provided to our customers. After all this is the main indicator of the company's success.

Thirdly, we focused on improving the quality of service for our customers. This is the main indicator of the success of our company. 

MySales allowed us to achieve these goals. It definitely saves the time of our employees. Their forecasting technology takes into account a multitude of factors like seasonality, weather, macroeconomic indicators, exchange rates and more. No human can take it into account in the manual forecast. For example, the hotter the summer, the greater the growth in sales of ice creams or water. If the temperature outside rises to over 30 degrees Celsius, sales can increase even tenfold. It is incredibly difficult for a person to calculate this relationship precisely, while the system builds it instantly. We did the pilot implementation on fresh goods which are the most difficult group of products to forecast. MySales did a great job in improving the sales forecasts and optimizing the stock levels in our warehouses.

How would you describe MySales auto-ordering system?

Auto-ordering is based on several key parameters. The main parameter is the sales forecast built by MySales. The second parameter is the safety stock, which is calculated also by MySales. We are responsible for preparing the schedule of delivery dates, based on which the number of days for which we place an order is calculated. The final parameter is the minimum stock level desired to have on hand for display purposes in stores - also called presentation stock.

Could you describe an example of how you work with MySales?

Yes, let's use milk as an example at which we show how MySales auto-ordering works. The milk supplier delivers goods to certain stores. Each delivered package contains a certain amount of products. Delivery frequency is usually three, five or seven times a week, depending on the region and the store. The frequency of milk deliveries is quite high, that is, orders can be formed almost daily. In MySales, we prescribed the minimum inventory for each SKU, determined by the delivery schedule and sales forecast.  We also take into account the overhead cost and safety stock for a certain number of days. After that we subtract the current stock, subtract the goods transit period and as a result we get our net need. This entire process is automated in MySales. Both our central team and stores are able to check the recommended order by system at any time. If necessary, we can always make changes to the order and inform the supplier. Currently, the order is sent to the supplier upon our employee confirmation, but in the near future it is planned to move to a completely automated workflow.

How often do you need to amend forecasts and order recommendations provided by MySales?

Based on the results of the 3 months that we have already worked with the MySales product, we received a very small number of adjustments from stores. The reasons for the edits are mainly related to the fact that at the moment only the store itself can track, for example, the actual availability of goods in the store, so for the time being it is still possible to make manual adjustments.

Which product groups have benefited the most economically from implementing MySales?

The greatest impact was observed on goods with a long shelf life, 1-2 months. There is practically no need for manual intervention.The same can be achieved for goods with a shorter shelf life if we introduce batching and RFID tags. Future looks bright!

How did your employees react to this innovation?

Frankly speaking, there were different reactions. Of course, some team members took the innovation enthusiastically. Typically young people. Certain share of employees was rather skeptical, but we did not encounter any open opposition to it. The main conclusion I would like to share is that those directly involved in the ordering process, consider MySales to be effective and saving their time. It really allows them to dedicate more resources towards more creative tasks requiring human touch. According to our analysis, employees dedicate significantly less time to the ordering process with the quality of ordering improved . For stores with balanced assortment and efficient staff, MySales almost imperceptibly helps to achieve excellent results at the store level.

How did the implementation of the MySales system affect the availability of goods in your chain stores?

With regards to the availability of goods, we can already see improvement, however it is still too early to discuss it more in depth. We have been using the system for only three months. The main point I’d like to note is that the order processing time has been reduced by 2-3 times. I consider it a huge success! We can also see improvement in certain qualitative indicators, such as the display of goods, their appearance, and accuracy of information about discounts and promotions. As I mentioned, MySales forecasting is currently working only on the fresh product segment. We go step by step, but ultimately plan to deploy it across all product categories.

Did you deliberately choose one of the most difficult to predict product groups for testing?

Yes, definitely. We are actually pleasantly surprised with the quality of forecasting. MySales forecast accuracy is much higher than any other forecasting system I ever worked with. One of the common reasons for our forecasting failing in the past was inability to precisely measure the price impact on sales. Unlike systems, humans do not always manage to recalculate forecasts after the price change. It is a huge advantage of MySales that it makes it automatically. We already see a significant reduction in lost sales and excess inventory that led to write-offs of wasted food and poor cash flow.

What are your plans for the future?

We plan to implement MySales forecasting for all groups of products. At the moment it is live at 105 stores in the fresh group products. Additionally, we plan to create a so-called automated work area, which would allow our employees to see all the analytics on: movement of goods, sales, write-offs, returns, receipts, price, etc. Wrong data is sometimes the reason for wrong ordering decisions nowadays. Further, we want to formulate a flexible ordering formula which can be tweaked to match the needs of various product categories. That will allow us to keep the isolated order logic for different products. Even further, we aim to build modules for controlling orders and other processes.


After scaling MySales throughout our store network and adding incremental improvements, the benefit we get from this technology will definitely increase. According to my estimation, we are just now seeing 5% of what can be achieved in terms of total optimization thanks to this tool. I am pretty confident about that. As soon as we make the case in our company that automated forecasts can be trusted, we will keep moving forward. There is still so much room to grow. And if we execute well on that, effects may be incredible!

Sergey Pritykin

Tell us briefly about yourself.

I have been working in the IT field for 15 years of which 12 years in retail: You can say I know it inside out. I structure my work so that 40% of my time goes to managing IT operations and the remaining 60% is reserved for supporting research and creating conditions for new technology related business initiatives.

How did the project with MySales start?

The project began with a demonstration of the product by its creator, Alexei Ivasyuk, to the management of our company. The MySales forecasting system, in our opinion, is very innovative. Our management approved it and supported the implementation.

What are the main stages of MySales implementation?

Before starting the project, we had to develop the logic for management of delivery schedules. It is a rather complicated workflow, given that it is often changed in our retail chain. Next, we needed to do an integration with MySales. There were no difficulties and it went quickly. Week and a half taking into account all the development work. We plan to invest more in this integration, however it already works stably and there are basically no problems with it. There was no need to hire additional staff, we managed to do it with our own resources.

What has already been achieved and what are the further expectations from MySales?

This automation of business processes is aimed at reducing operating costs. Our goal is to improve the quality of work without increasing the number of staff. We want to free our employees from monotonous and routine work so that they can focus on improving sales efficiency and the quality of service provided to our customers. We are constantly working to improve the skills of our staff and strive to retain the most qualified team members.

MySales helped us quickly solve three main problems. Firstly, minimizing human errors. Secondly, improving the staff productivity by expanding their expertise. Finally, reducing the negative business repercussions of human errors. We expect to see decreased write-offs on wasted food due to improvement in the ordering process. It is still early to say, but we already can see that persons involved in the ordering process dedicate much less time to it. We expect to see even better results for the other product groups.

Original source (RU):