Sign up

Mark Up

Fantini (Evo) Push is dead, we need to create value from the symbiotic relationship between human intuition and AI

Turin, 29th September 2020

Interview of Luca Moroni

Mark Up podcast

Fabrizio: “Push as a retail concept, but not only that, is dead. It’s not only e-commerce that’s new, but it’s the management atelier itself. There is a new paradigm that is necessary not only to be successful but also to survive”.

Luca: “Good morning everyone, I’m Luca Moroni, journalist of Mark Up and GDO Week. With us today is Fabrizio Fantini, CEO and founder of Evo, a company he founded in 2013 that deals with price algorithms. Reading the biographical notes I was particularly struck by a sentence, that is to create value from the symbiotic relationship between man and technology. So I ask you: what are the advantages we have from the combination of human intuition and artificial intelligence?”

Fabrizio: “This is mainly due to the fact that managers as human beings have different strengths than machines. In particular, classical training has taught managers to “push”, that is, to push their instructions, their planning and then their vision on the market.

So in planning rather than in retail execution the traditional logic was just a push logic, but this is no longer enough because today there is too much difference in consumer preferences: they have become volatile, they are hyper-local, especially they have a lot of choice not only because of e-commerce but also because of cognition, the speed of information dissemination.

Therefore the symbiotic relationship between man and technology allows to overcome the traditional logic that I call “management of the past” where in fact the same commercial campaigns were always repeated and the objectives were always based on the same historical sales curves. This technology helps to overcome it. I’m talking about autonomous technologies that is able to design their own tests and to draw the learnings of customers on the basis of data and then self-adjust themselves. Here we are talking about a world of defection where the man, the manager, however, continues to select strategies, do the planning, draw the storytelling of the brand, the product but then the execution is literally delegated to the machine. This has enormous advantages: greater importance, therefore greater profits, just because the customer helps to find the solution in real time and remains always relevant and at the same time better inventory therefore less management risks and less risks for the survival of the brands”.

Luca: “We are therefore talking about predictive analysis. But how can this data help to make accurate decisions in the retail world?” 

Fabrizio: “I’ll try to explain it to you in simple words. Essentially, in order to have effective, relevant and what we call always autonomous decisions, it is necessary to have a dynamic understanding of the demand and then an optimization of management choices.

So, in reality the help is twofold: on the one hand we help companies to complete the few data they have in hand because we directly observe the behavior of 1 billion consumers in the world out of 10 million products on the web and over 100 billion dollars worth of orders, commercial transactions that is the sum of everything we have seen and treated throughout the years of work.

Thanks to this data we are able to form a daily, dynamic and quick learning image on the 4 dimensions of each commercial transaction: the price, the person, the place (maybe the store) and the product (all its characteristics).

According to this, the systems help retail managers and not only to make accurate decisions tracing their profit targets. Basically eliminating completely the concept of clustering grading because each store becomes a separate entity and therefore with different assortments, customized inventories, targeted discounts.

Overcoming the concept of initial allocation that for example in fashion was typically the meaning of push logic, so in the same days all stores were supplied with a fairly similar amount of product. Here instead, thanks to dynamic production systems, each item can find its own distribution independently, both initially and then during its life cycle.

Last but not least, defining prices and discounts according to specific market conditions for each location, for each historical moment of time, so as to continue to be always relevant. So in fact it is a bit the soul of the execution that we see  as I said before, having considered planning and strategy then the decision on what to expose, what assortment to have, how much to buy, what products to have, in what locations, what prices and discounts to implement, which are really a significant help to improve the profits of companies.”

Luca: “Very interesting at theoretical level but this technology has had some concrete cases. Among them I was impressed by your case with Miroglio group. You claim that the implementation of the intelligent store, let’s call it like that, has allowed their group to reach 1,000,000 Euros profit more per month. How was this result achieved?”

Fabrizio: “With Miroglio we actually demonstrated the effectiveness of a more dynamic management of inventory and price decisions, also using the input of the stores through what we called “borsino”, which was 4 years ago a revolutionary experience that first connected all customer data on the one hand, then everything we knew directly and indirectly from customers and on the other hand price, allocation, discount and replenishment decisions.

This in a very interesting way was demonstrated only by using this system, like a satellite navigator of a car, when the first navigators came out the only way to determine if they actually worked was to use them.

Therefore what we have done together with Miroglio group has been to develop also with the University of Turin algorithms that would allow to change every day the decisions of allocation of the inventory and price and also to the stores to exchange the goods. This was a case study published in the U.S. by Harvard Business School in 2019 and this year we are publishing a new one on an American company called Event Network which has 200 stores, which we helped to improve the margin by 30%, we are talking about very important numbers, through “filling” actions with a focus on pricing.

For example, a month ago we published with Microsoft the Boggi case study that we helped improve supply chain efficiency by 18% by selling 4% more with 12% less inventory after just seven weeks.

These were very important results. In fact, this summer we achieved the result that I am most proud of in my career, which was to make the sales of dying products bounce back in a V trend, by giving them a second life without selling them at a discount, but moving them from one store to another, and then dynamically changing the assortments in the middle of the life cycle, and then unhinging what was the initial architecture.

This is a result in my opinion incredible, because the real optimal discount is 0% so when it comes to optimizing discounts and sales in truth is much better to avoid having to make these discounts. So by being able to sell more dying products we were able to push the discount optimization to its extreme, as low as to zero.

All this with a very pragmatic approach, that is to say by providing tools that in everyday life help to make decisions, not only to implement them, but also to determine what to do every day”.

Luca: “You talked about a few months ago when we were in full Coronavirus emergency. I ask you what Covid has left us, what has it left to the retail world and what do you see in the near future?”

Fabrizio: “Let’s say apart from the heavy human stock that all of us have experienced at a business level, in my opinion, Covid has accelerated some trends that were already existing anyway. Push, as a retail concept but not only, is dead. It’s not only e-commerce that’s new, but it’s the management atelier itself. There is a new paradigm that is necessary not only to succeed but also to survive. So what has happened is that Covid has dramatically accelerated some latent issues and therefore differentiating the winners and losers, thus creating a stress for the change that was very often delayed.

So today there is a need for more integration in market decisions, more speed in the dynamic response, therefore local, what to do at any given time, because it is now clear to all that last year is no longer a guide for this year and this year hopefully not will not be a guide for next year. So today the same autonomous systems can define not only prices, purchases, assortment but in the future they can also help CRM decisions at the level of space in stores and much more.

In the near future I see that organizations that still insist on continuing to make top-down changes coming from the headquarters or classic consulting projects to have great strategic visions without changing everyday operations are destined to extinction. Excuses for not really changing were eliminated by Covid, it was actually a shock in some ways, even in the tragic events, but productive because helping companies to renew themselves after all is a noble goal and for me is the reason why I decided to leave McKinsey and launch myself in this new exciting entrepreneurial adventure. In the future I see greater integration and a radical change in management actions and no longer push but a strategy designed by the manager and then an increasingly integrated and customer driven execution through data and systems”.

Hey! Was this page helpful?

We’re always looking to make our docs better, please let us know if you have any
suggestions or advice about what’s working and what’s not!

Send Feedback