X5 Group: looking for startups that bring significant money

Following the results of the first year of operation of the Retail Innovation Tech Alliance (RITA), 9 startups will conduct a pilot with the X5 Group. Denis Levchenko, Head of Scouting at X5 Group, and Natalia Martemyanova, Scouting Project Manager at X5 Group, spoke about this with Eugenia Barchenko, Co-Founder and Chief Operating Officer of GoTech Innovation. Denis and Natalia highlighted what was key in the selection of these companies, and also gave advice to the participants in the next RITA selections.

What were the most important criteria for selecting startups within the RITA framework during the first year of operation?

The main thing we pay attention to is the correspondence of the proposed solution to the main areas and interests of X5 Group. We are talking about two key business drivers. The first is the operational efficiency of the company, reducing the cost of implementing the main business processes that can be optimized through the implementation of technological solutions. First of all, this is offline retail: staff work, acceptance and display of goods, cash services, interaction with the customer. Further on, ensuring the work of retail: lighting, heating, air conditioning, network development, maintenance of buildings and structures. The next area is supply chain: all solutions related to logistics, transport and warehousing. Another segment is back office work. A separate area that may be inside the previous ones is food tech: for example, projects that allow you to reduce losses from food spoilage. Finally, decisions for the company’s online divisions can influence the optimization of business processes.

The second business driver is the company’s revenue. These are all solutions and technologies that are associated with improving the shopping experience when interacting with X5 Group, affecting such indicators as customer satisfaction, NPS, frequency of visits, average bill, attracting new traffic, etc.

In addition to these criteria, the experience of the team is important: whether or not there were previously pilot projects and implementations, whether the team knows how to operate with figures and facts, what economic effect it managed to achieve earlier, how does it assess the total cost of its solution. Of course, responsibility is important.

How has the pandemic affected the search for new solutions? Have you adjusted your needs?

The pandemic further increased the focus on cost-effectiveness amid reducing revenue growth — an important factor that we needed to support, including through technology adoption.

On the other hand, new needs have emerged. For example, due to innovative formulas, we managed to reduce the purchase cost of sanitizers three times. In addition, we tested various air purifiers, thermal chambers, and began to work intensively on self-service technologies. Some projects are of a short-term nature, but such areas as self-service systems have moved into the category of steadily developing ones.

In general, regardless of the pandemic factor, we are always determined to achieve the business value that technologies bring by influencing the company’s financial performance.

Is it important for you that a startup that potentially claims to cooperate with X5 is at the stage of a working business? Are you considering earlier stages?

As a rule, projects at the stage of technology or product maturity are more preferable for us, since the main task is to quickly achieve economic efficiency. Ready solutions can be quickly tested and launched into business. But at the same time, we do not disregard technologies that are at an early stage. We are ready to deal with them for the future if we cannot find a similar mature solution.

Were there startups that, within the framework of RITA, returned with a repeated application and were selected during the year?

Yes, there were such examples. This mainly happens when the level of maturity of the project rises or the startup presents a better evidence base — whether it was a pilot or an implementation and is ready to show in numbers what it has achieved.

Of course, if we see that the technology is very interesting and relevant for us at the moment, then we ask for additional information. But in those applications that are filled in poorly, it is difficult to understand that this particular decision may be a priority for us. Therefore, the better the meaning of the project is conveyed in the application, the more likely it is that we will pay attention to it.

What are the main mistakes you can note due to which solutions had to be rejected?

The solution does not correspond to the priorities that I mentioned above, or the startup does not operate with specific facts and figures but enjoys its product and describes it in every possible way in order to catch us somehow.

It often happens that a company focuses specifically on a product or its development but does not always understand that from the point of view of business value for X5 Group it is difficult to measure it in money. For example, a solution to reduce losses from product spoilage, which is measured in billions, will immediately attract us, and we will deal with it in detail so as not to miss it. And if we are offered a chat bot that trains employees, then we will not see the benefits in it, there are a huge number of such solutions.

Our main KPI is additional EBITDA. We invest the company’s money and, as a venture fund, we must return many times more than what we invested.

Then, tell us, which applications can catch you right away?

It’s great when a startup understands our business processes and correctly correlates with them. As a rule, these are companies that have had projects with other retailers. They are confident that the proposed processes are working. Special attention is given to applications that meet current business needs.

It is useful if the company is ready to quickly test hypotheses, rather than enter into a long and complex contractual relationship.

There are examples when modern technology is proposed, but a lack of understanding of business processes puts an end to such projects. In particular, proposals to place robots in sales areas that will advise customers, check the availability of goods on the shelves or issue loyalty cards. Everything seems to be cool, but the question arises: how will such a robot move around a shopping center, where there are a large number of customers, carts and boxes?

What aspects of piloting should startups working with X5 Group especially consider?

You need to be prepared to integrate with X5 Group IT systems during the pilot. Another point: you should pay attention to the quality of onsite real testing. Technology often breaks down against the realities of life — the very business processes. A startup may initially ignore this or rely on our project teams. But, if we see that, on his part, he is actively engaged in this, then this will only be a plus.

For example, we tested a solution for labeling products in stores. And the startup team itself went to the trading floors, communicated with the staff, and as a result highlighted those moments that were incomplete. The more we receive such information, the more efficiently we can implement the solution and more accurately assess the business effect in real figures. If according to the result there is a delta “was / is” and it is significant, then we make a decision on scaling. And if a good product that could bring an effect was poorly implemented, then the actual numbers will not show the proper change.

Over the past year, how have startups benefited from working with X5 Group as part of RITA? What was their benefit? And what’s yours?

In total, over the past year, we made a decision to hold 9 pilots with companies that have passed the RITA selection. The most advanced stage is the Chattermill project so far — a service for processing customer feedback using artificial intelligence and machine learning tools. The service is recommended by the Pyaterochka retail chain as a priority.

Another project that is being prepared for launch at Pyaterochka is Legium — remote signing of documents with sub-tenants.

Now pilots with two companies are on their way: RetailNext, which came from a foreign funnel, and Tokinomo.

RetailNext can help us to competently analyze experiments and hypotheses, to find out how they affect RTO. Thus, we will have the opportunity to increase this indicator.

Tokinomo is a robot displaying a product on a shelf. Now we are testing its functionality in several supermarkets in Moscow. If the technical aspects are successfully tested, then we will start a business pilot.

Overall, in our group, the largest numbers can be achieved through economies of scale. It is enough for technology to bring an increase in turnover by 1%, when a large multiplier is applied (X5 Group revenue), we will get an excellent result.

The next two projects — Yoloco and Retouch4me — were tested, our divisions liked them, but we cannot achieve significant economic effects with them. Departments have kept them as tools that can be deployed when needed.

The Road Gatherer startup has received feedback on the revision from our transport department. Plus, we gave information on the results of the pilot to KAMAZ, so that they increase the efficiency of their systems. The startup itself has potential; it hits the right business driver — cutting fuel costs. The X5 Group has more than 4,000 trucks, the costs are significant, and a 5% reduction can be a very good effect.

All of these startups received feedback from us. If the project enters the implementation stage, then annually it has the opportunity to receive from X5 Group on average tens of millions of rubles a year in revenue.

What do you generally mean by piloting — what is the process and the result?

When we receive an application, we try to understand which business process the project is suitable for, after which we communicate with business units and departments within X5 Group, which can act as customers of this solution. Further, together with the startup, we develop specific business hypotheses. For example, if we implement the solution in such a number of stores, we will be able to reduce losses from theft by 1%, or increase RTO by 2%, or reduce the time that cashiers spend on customer service by 5 seconds.

The architectural committee becomes an important stage. Here it is important for us to understand whether the solution can be integrated into the business process from an IT point of view. On the one hand, there is a possibility that we will not be able to upload sensitive information to the outside, on the other hand, a startup may be engaged in some additional protection of its software for running on our servers.

If we believe in the business hypothesis, then we approve the total cost of ownership of the technology with the startup. We build a financial model of a pilot and a potential implementation project, determine the possible economic effect in terms of IRR, payback period, annual additional EBITDA. Priority is given to the project where these indicators are more attractive, while taking into account our belief in the implementation of the case. This is where the final decision on the pilot is made. In the course of the project itself, we check the approved hypothesis on actual figures: we determine the required sample in order to obtain a statistically significant result. If it is achieved, then we most likely will see it on a larger scale. It may also happen that the delivered KPI will not be met. Then we go back to the financial model. If IRR, payback period and annual additional EBITDA are in attractive areas, then we are also ready to implement such a project.