In 2004, Evo’s award-winning team of 32 data scientists created its model of Artificial Customer Intelligence, which uses machine learning to generate optimal results for pricing, promotions, forecasting, and replenishment—all in line with each individual business’s strategies.
Our $250 million client margin impact, from more than 8,000 stores across 35 countries, shows that the science behind Evo’s “secret sauce” is solid.
Our biggest challenges this year have been explaining to businesses what advantages this revolutionary new technology can bring, and translating complex explanations about what we do into language our customers can easily understand. Tweet This
That’s why at this year’s convention we decided to practice what we preach by bringing together both the techie and business sides of our predictive analytics company for a two-day conference in the Italian town of Termoli.
Twelve members of our team were unfortunately unable to attend, owing to visa problems, calendar clashes, or travel restrictions. We look forward, however, to welcoming them to EvoCon 2019.
As for the rest, throughout Thursday and Friday we traveled down to Termoli, a little-known retreat on the Adriatic coast, situated in the remote region of Molise—or, as some call it, “Molisn’t,” because so few people have heard of it.
Termoli's remoteness did nothing to prevent EvoCon 2018 from being our most popular, highly attended conference yet.
Fabrizio Fantini, Evo’s CEO, kicked off the conference with some round-table introductions.
For some of us at Evo who’ve been working remotely for several months, the conference was our first chance to meet face to face.
Evo chairman and retail expert Robert Diamond flew in from London. Evo’s team of data scientists traveled down from Turin. Following a testing experience with a possibly non-existent E-ticket company, our product-quality guru, Tina, and full-stack developer, Nick, finally made it over from Athens.
Fabrizio started by discussing our values and outlining our current trajectory. We’re shedding the skin of our Italian business to become a leader on the international stage.
“Globally we’re a family,” said Fabrizio, “but the nature of the family is changing.”
After this introduction, our systems analyst Davide gave us the rundown of Evo’s system.
As Evo grows and we crunch growing amounts of data, we’re having to match an ever-increasing demand with speed, access, and most importantly, security.
We decided that the best way to do this was to build our own in-house data center.
The same sum of €30,000 would have bought us less than a year of storage on the cloud––an unnecessary barrier to the data that drives our company. So the decision to invest in more than 20 physical and virtual systems of our own was a no brainer.
Tina and Giulia then gave a demo on our pricing and replenishment tools. Evo Replenish addresses a simple—if not strongly felt—pain point: the waste of potential in retail resulting from the misallocation of stock.
To combat this, the machine-learning algorithm analyzes actual and historic data for sales, inventory, and weather and combining this with strategic decisions taken by the company headquarters, calculating the optimal stock levels for individual SKUs at individual stores.
What did EvoCon 2018 do differently to previous events?
While Evocon 2017 looked more at the software side of things, this year’s event tried to connect the dots between the data and the business sides, and really dive down into how to present to our future clients.
We addressed this through a series of hackathons, moderated by our senior scientist Giuseppe Craparotta. These workshops were designed to address real customer questions. So what better way than focusing on real questions?
The hackathons brought together data scientists, business thinkers and, yes, even technical writers to tackle a series of questions beginning with how a customer could increase their average price by 1 percent without changing revenues or promotions.
We made some mistakes, though on the presentational rather than the mathematical side. And like the clients we serve, we have the habit of focusing not on the 95 percent we got right but on the 5 percent where we fell short.
But like the machine learning software that drives our results, we learn only by making mistakes.
At the beginning of the hackathon, we directed our focus towards the technicalities of the software rather than the results we were producing—a mistake we need to learn from, said Evo’s Chairman, Robert Diamond, in his concluding speech.
“Customers buy time,” Robert emphasized. “They don’t want a pricing tool. They don’t want a forecasting tool. Fundamentally they don’t want to know how the tool works. What they want to know is that we’ve produced results.”
Just as we train our machine learning software by feeding back information, Robert’s closing comments served to sharpen our focus for the following presentations.
The devil is in the details, especially where data scientists are concerned. As important as the details are, however, Robert’s feedback reminded us not to lose sight of client relevance when packaging our product in understandable terms.
Evo’s an evolving organization, with a driven team of future leaders in the fields of AI and data science.
Rather than one-off consultancy, we offer ongoing solutions to recurring business problems. As such we’re constantly learning, and constantly improving.
The nature of what we do means our work never stops. But to work most effectively, and bring our clients the best results, we have to cooperate, collaborate, and come together.
Our team-building program was designed to create this cohesion.
Our team mulled over several slow-paced, gourmet Italian lunches and dinners, partied into the night at our private gazebo on the beach, and spent a lazy Sunday cruising around the Tremiti Islands.
For joining the dots and bringing us closer together, this program worked perfectly.
What does the future hold for Evo?
The company aims to grow by hosting new servers in Termoli—which from a security perspective would be perfect, as Termoli is in the middle of nowhere.
From a systems perspective, 2019 will see us move towards a model of offering “Infrastructure as a Service” (IaaS) as standard.
Evo is learning, growing, and, just like the results our software suite produces, incrementally improving.
On the last day, while cruising the Tremiti Islands, a group of us led by Fabrizio swam our way into the claustrophobic Grotta del Bue Marino—Cave of the Sea Lion.
The deeper we went, the darker (and colder) it got until we lost our CEO to the dark.
Rather than panic about Fabrizio’s apparent decision to turn his back on civilization and embrace his new life as a hermit, I saw this as a metaphor—that in Evo we don’t shy away from challenges; we face them head on and see how far we can go.
Our next challenge is how to make our limitless potential the talking point for our customers.
About the author
Alexander Meddings is Evo's content expert on artificial intelligence, machine learning, and related topics.
He is an experienced journalist who covers branding, social media, marketing, and technology, with degrees from the University of Exeter and University of Oxford.