How Salomon fixed its data problem (and what you can learn from it)
A few years ago, Salomon had a problem a lot of growing companies run into: nobody could agree on what the numbers actually said. Different teams kept their own copies of the data in spreadsheets, so a marketer and a store manager might look at "sales this week" and see two different figures. That made every decision slower, right as the business needed to move faster; Salomon (the outdoor brand behind those trail-running shoes) was opening new stores around the world and growing its online business fast.
"People were constantly asking: what is the truth? Which number is right? What report, what tool do I even use?" – Thomas Vigneron, Salomon's data director
Thomas Vigneron, Salomon's data director, led the fix. Here's what they did, why it worked, and how you could do something similar.
What they did
They started watching what customers actually did, in real time. Instead of guessing or waiting for a report, Salomon put real-time tracking on its website and app using Piano Analytics, so they could see what visitors were doing as it happened.
They put all their data in one place. Sales figures, loyalty data, marketing numbers, in-store data, website data: all of it now flows into a single system instead of living in separate spreadsheets. That gave the whole company one shared, trusted set of numbers to work from.
They gave everyone role-specific dashboards. One view for leadership, one for regional teams, and one for people running individual stores. Store managers can now check how their store is doing in real time on their phone, instead of waiting until the next morning.
They started using the data to personalize marketing. Salomon grouped customers by what they seemed likely to want, then used those groups to run smarter, more targeted ads on Facebook, Google, TikTok, and Meta instead of showing everyone the same thing.
They're now testing AI to go even further. Vigneron sums up the ambition simply: "I think the holy grail for everyone is personalization at scale, one-to-one." The problem is that no person can pull that off by hand: deciding the right message, offer, channel, and timing for every single customer is, in his words, "endless, a nightmare, something you can't achieve by hand." So Salomon is testing an approach where a person sets the goal and the guardrails (say, more repeat purchases from their most loyal customers) and lets an AI model work out the details for each customer automatically. A separate experiment lets staff ask plain-language questions about sales or stock ("what were yesterday's best sellers?") without needing to know how to pull a report. Another is exploring letting software decide the details of a marketing message automatically, based on rules a person sets.
Why it matters
Vigneron is direct about why any of this was worth doing: "This isn't just about reporting. It's a growth enabler for D2C." The payoff showed up in real numbers too: teams save up to four hours a week that used to go into hunting down and double-checking figures, more than a hundred employees have been trained to pull their own reports instead of waiting on a specialist, and the store app is now one of the most-used tools inside the company. Early tests of the personalized ads also showed a small but real bump in how many people bought after seeing them.
4 tips if you want to do the same
Fix the foundation before you fix the reports. If teams keep arguing over whose number is right, the problem usually isn't the dashboard, it's the data feeding it. Start there.
Get leadership using it first. Salomon rolled dashboards out to executives before anyone else. Once leaders trusted the numbers and used them daily, it was much easier to get everyone else on board.
Put real-time information in the hands of the people making decisions. A report that arrives the next day is a history lesson. A number you can check right now is a decision you can act on today.
Start small with AI, and only where there's a real problem to solve. Salomon's rule: no AI project moves forward without someone senior asking for it to solve a specific, real issue.
Where Piano fit in, and how you can start
To see what customers were doing on the site and app in real time, Salomon used Piano Analytics – without needing their own data team to set it up.
It's private by design, built to meet strict privacy laws like GDPR from the ground up, so better insight can flow without asking more customers' data than you need. And it works well with whatever you already use: you can send data into a BI tool or your ad platforms, the same way Salomon layered on Snowflake and Power BI over time.
If your team argues over whose numbers are right, or store managers still wait until the next morning to find out how yesterday went, that first step is usually the fix. Get it right, and everything you build on top, from dashboards to personalized marketing, gets easier and more accurate.
If you want to hear more presentations like this on AI, data and personalization from teams putting it into practice, join us at Piano Academy 2026 in Amsterdam, October 6–7.
Register here to save your spot.
