Oleg Spirin, VARUS CDO: "Corezoid became the digital core of VARUS"

Oleg Spirin, VARUS CDO

– Oleg, how did your work at VARUS begin?

– I became Deputy Marketing Director of VARUS in 2017. I was immediately assigned to the project “Loyalty Program”. Also I became responsible for the digital customer service and data analytics projects from scratch.

I immediately decided that I came to VARUS not for a cool coffee machine and not for smoothies, but to do real projects. And the projects impressed me. I saw a great field for applying the knowledge that I had previously received at PrivatBank.

In VARUS we started the loyalty program with 700,000 issued cards, of which only about 400,000 were registered (these are the customers whose phone number we knew). At the beginning, a customer communication system was not working. VARUS used to send SMS messages, some messages to Viber. E-mail was not used as a channel at all, and the call center could only be called to complain.

– What did the VARUS IT landscape look like when you joined the company?

– The main software for the Loyalty Program was Manzana Loyalty (outsourced) and custom version of Microsoft Dynamics CRM. Leaving PrivatBank and using the technologies that were then in the bank, I realized that the retail world is still very far from perfect and full of new challenges. But also I understood that there are always some limitations. Therefore, together with the team I began to use what we had.

– What did you do in the first months of your work at VARUS?

– I studied the funnel of customer registration in the loyalty program. I didn’t like the funnel very much. It looked more like a leaky bucket. Only 10% of customers reached the end of the funnel. We began to seriously work on each stage of the funnel. For example, we refused paper questionnaires. Together with the team, we went to the Boss two or three times to sell the idea of ​​rejecting the paper questionnaire, and in the end we sold it.

Later, we began to segment customers: “new”, “active”, “estimated churn rate” and “churn rate”. I am a big proponent of trigger communications based on certain actions. From the moment when the client came to the company, and until the moment when he left, we set up sending notifications as a reaction to events that arise in the relationship between the client and VARUS.

As for the customer service, we first of all rebuilt the call center. Back then only every second customer managed to get through. We changed the contractor, put KPIs and introduced SLA. The project team from scratch built the correct interaction of the call-center with customers: from digitizing the incoming requests to closing the tickets. Then we digitized our work in social networks and Google Maps. By the way, Google maps turned out to be a cool source of information! It had everything: reviews, ratings, store ratings, photos and analytics on how many people visit the store.

Having digitized all customer requests, we began to promote a customer-oriented culture in the company and introduced such a metric as NPS. The first measurement of NPS showed us a 30% indicator. By the way, NPS is now at the level of 55%. As for analytics, we learned how to match customers with their receipts and got the opportunity to build models and predictive analytics.

– Who can be considered to be the “Dream Client” for VARUS?

– A “dream client” is a person whose purchase frequency and average bill are out scale. And it is advisable that you buy only from us, and not from competitors. Recent research says that a customer visits an average of 2.6 different stores per month. The loyalty program is precisely aimed at making VARUS your favorite and only store.

– What KPIs do you have in the “Loyalty Program”?

– For a digitized customer, the average check and frequency of purchases should be higher than for an un-digitized one.

– Tell us about the work of VARUS with messengers?

– Initially, it was believed that the bot in the messenger would become for us a free communication channel. We thought: “Oh! Bot. We will save on sms.” But then we understood that the bot is a way to quickly “digitize” the client. Literally in one click. You take the client’s phone number, register in the bot - and that’s it, you can start communicating. When we understood this, we mastered all the messengers that are available in the country.

– What can VARUS bots do?

– A bot is, in fact, a virtual loyalty card. This card can be controlled directly in the messenger: block, unlock, configure communication rules. We have created a section with promotions and personal offers. Also in the bot there is an assistant who helps to find out the price of the goods, complain about the lack of goods or the wrong price, evaluate the supermarket.

There is a “Shopping list” section that is synchronized with our “Promotions” section. When you look at the “Promotions” section, you can immediately add selected products to your shopping list. We send all checks after client transactions in VARUS to the bot. We set up polls and advertising “tails” in the receipt that we send to bot. In the “tail” we recommend the buyer personalized offers in relation to the receipt.

Also we connected the online chat so that the client has the opportunity to receive online support without dialing our call center. We plan to gradually launch similar mechanics in all instant messengers. For now, we started with Apple Business Chat in iMessage.

– What are your main achievements in 2 years?

– Now we have about 2 million cards issued: these are plastic + virtual cards. From 1.2 million cards, we know the phone number and can communicate with the client. In fact, we have grown 3 times in 2 years.

– What is the place and role of Corezoid in the company?

– Corezoid has become for me the main tool that allowed me to build communication with the client. This process began with a bot factory. All our bots are fully built on Corezoid. We also actively use the UCB (Unified Client Base), Communication Calendar, Sender Communication Engine, Sync API, GitCall.

Today, Corezoid works as the “digital core” of VARUS. With the help of Corezoid we get data from different sources, process it according to the logic laid down in Corezoid and use it in business processes.

When we started making the first bot on Corezoid, we were faced with the fact that we had almost no APIs. By the end of the first year of work with Middleware, we had already developed about 50 APIs, which allowed us to receive information on employees, on a client, on a product, on a store. We got our Manzana contractors to do the API. Today, VARUS has about 200 APIs.

We recently finished testing an expert system as part of the ECB based on the Deepmemo Rules Engine. We came to understand that the 3 entities that we worked with before - the client, the supplier, the employee - are, in fact, one entity: the client. He may be our employee, may be a supplier. The ECB just gives the opportunity to “digitize” this entity and work with it.

The communication calendar began to gradually replace the old CRM. Today, almost 100% of our communications go through the Communication Calendar. All our communications are now digitized, occurring according to unified rules and standards. We monitor the reaction of clients to communication, targeted actions, conversion rates and all the movement along the funnel.

– In which channels do you now communicate with customers?

– In all the messengers: Viber, Telegram, Facebook, Apple iMessage (Apple Business Chat), SMS, E-mail, IVR.

– How did you set up the interaction between IT and business analysts in those departments who work with Corezoid?

– At first there was some kind of natural denial and rejection. But after 3-4 months, everyone saw the benefits, and we started to move fast this way. Now I believe that we have already built a good interaction between IT and business. Moreover, there is an understanding that it is necessary to enable the business to do almost one-click APIs for their needs. SyncAPI allows you to create from some process in Corezoid not just a webhook, but an API to which you can connect and work in synchronous mode.

In 2019, a new IT director, Andrei Bobrov, came to us. We quickly found a common language. Andrei created a strategy, where he announced that there should be close interaction between business and IT. IT must provide the foundation and uninterrupted supply of data, and the business must become a consumer of this data.

– How many processes do you have on Corezoid right now?

– 520.

– Today in the media they often write the words “digital retailer”, “open retailer”. What do you think of the work of VARUS in the digital world? How is a “digital” retailer different from an “analog” one?

– There is a lot of talk about “digital retailers”. But in reality, companies take only small narrow steps: they release mobile applications, set up self-service checkouts. As for me, conceptually nothing new is happening. It seems to me that a “digital retailer” should completely change the whole experience of a customer’s purchase. I think you need to expand offline purchases with online tools. For example, the topic of self-delivery from the store on the principle of MacDrive, when you go home from work and order a basket or simply click the “Repeat previous order” button. Then you drive up to a specially designated place, you take the bag and leave. This is a new shopping experience.

We have a project in the plans: “Skipline”, which is an opportunity to “skip” the line. Imagine that the office manager wants to have lunch, subscribes to the bot “Lunch every day”, pays. Then he just comes and picks up a beautifully packed lunch bag. Such a VARUS-cafe.

In the near future, grocery stores will remain almost the only offline place where people will come. Here in this direction it is possible and necessary to make a new customer experience!

– Please recommend books that have impressed you recently?

– “Crystal Horizon” by Messner Reinhold and “Creating a data-driven organisation” by Carl Anderson. The first is about overcoming yourself, the second is about how to build a company culture where decision making is based on data analytics.