Sign up

Growing collaboration of AI and human intuition

Evo leads the way into the future for smarter data-driven decision making

London, Turin – October 1, 2018

Behind just about every important decision made in retail sales, there’s a growing influence of artificial intelligence and predictive analytics. From some of the smallest choices, such as where to put store items and how high they should be, to the bigger ones, like which store should stock which item and how many of each, data models are secretly underlying more and more of the decisions. Behind every perfectly arranged store display window, there is an ever-increasing chance that an AI may be helping determine the layout.

Evo is one of the pioneers of this trend. CEO Fabrizio Fantini tells how the company was born from a PhD thesis at Harvard: “In 2013, I picked up the arguments of my doctoral thesis which essentially proved the use of intelligent algorithms could help a company to make better predictions and then make better decisions to get more profits. Since then we have developed two offices, in London and Turin, we have established international academic collaborations, while initially we were a spin-off from the University of Turin. Today we work with universities in France, Spain, England and the United States so we can say that we have become truly global.”

Last year, Evo’s systems managed goods worth over 10 billion euros, from over 2k retail stores, changing over 1 million prices, and physically moving it over 15 million times, generating an impact of over 100 million euros in profit for customers. Evo utilizes 6 different products, Evo Replenish, Evo Pricing, Evo Promo, Evo Forecast, Impact Tracker and Market Tracker, to bring these results to clients.

“The world has changed since we started our business and the choice to open a branch in Turin, Italy, proved to be successful, despite some of the criticality in our beautiful country,” says Fantini. “Four years ago, you could count our team on one hand. Now we have clients in 6 countries and sell in 60+, with typical revenues from $100m to double-digit billion $.”

Earlier this year, Accenture released a study on artificial intelligence and work. This study interviewed 1,200 top managers and workers from a dozen countries, including Italy. The documents show that by 2022, artificial intelligence will improve business revenues by 38 percent and increase employment by 10 percent. However, there’s one condition: companies will have to update their current business models and promote training for employees and consultants.

“In the coming years, the companies that will grow most and hire more people are those who will be able to innovate with artificial intelligence before their competitors” assures Marco Palminiello, partner of Evo.

Demonstrating the successes of their data pricing model, in April of 2018, Evo won the award for best Digital Supply Chain Start-Up at the It4Fashion Awards in Florence. This prestigious competition is dedicated to highly innovative start-ups that use technology to revolutionize the fashion industry.

So how is it possible that artificial intelligence can understand exactly where to put things and to manage stock allocations? It starts with structured data. Thousands of data points collected across months and years of retail sales. The AI then extracts information from the data and give suggestions, so that nothing is left to chance.

“We transform data into predictive signals of market trends and we use them to help companies make more systematic and profitable decisions,” says Fantini. “Variables such as historical sales, geographical area, climate and social media, and even product features and web images are used to predict the sales potential and therefore optimize prices, promotions, inventory allocation to stores and planning. Thus, providing a software tailored to companies, who then benefit from the extremely rapid and low-cost of the implementation time, thanks to artificial intelligence and machine learning.”

However, people aren’t irrelevant in this data-driven model. In fact, the opposite is true. What Evo does, is put all of this data together with the help of store managers and employees. The data provides the raw materials, but it’s only with the help of human intuition are the results made into real world actionable decisions. It’s the combination of powerful AI data processing, and the human understanding and application of the data that leads to the best business results.

“This is a one of kind work, based on effective measures of product popularity for which the human contribution is essential.” ensures Giuseppe Craparotta, senior data scientist at Evo. “Many retailers are surprised when they discover that their daily contact with the public and their product is the decisive factor: the algorithm’s soul.”

Indeed, this surprising approach is continuing to attract attention around the world. In June of 2018, Evo was honored by Intesa Sanpaolo Innovation Center and invited to attend the Decoded Future Fashion Summit in London, as one of the most promising fashion tech startups. Evo was able to meet, pitch and participate in various session with big brands, retailers, technology players, innovators and investors of international importance, further demonstrating the value of the Evo data tech and boosting the brand to new heights.

As for the future, Evo is constantly innovating and pushing forward. The next groundbreaking project is an app aimed at providing advanced sales expectations and prices for a product before it is released, all from just a single photo of the product. The shopkeeper simply has to take a picture of a new product model sketch and the app will provide recommendations for how many pieces it will sell, what the potential pieces of that item are and what it would be, what could be an ideal selling price. The app is currently in the final stages of development and is scheduled to be released within a year’s time.

Download PDF: English

For press-related queries please contact press@evopricing.com

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