Personal care
63% and 71% product sales uplift in a salon trial.
63% / 71%
The problem
Specialised beauty practitioners have eaten into the traditional salon book, so salons increasingly rely on product sales to defend revenue. The mechanics are stacked against them: limited retail space, customer reluctance against perceived high prices, e-commerce undercutting on selection and price, and stylists who are trained for craft rather than confident product promotion.
The brief was simple to state and hard to execute: turn every stylist into a natural salesperson, without compromising creativity or the client experience.
What we built
Working with a UK salon group, we trained the TUBR engine on three signals: point-of-sale history (what clients have bought, and what they might buy next), booking data (services, dates, times, and the clients themselves), and product knowledge (catalogue attributes alongside historic sales).
From those signals the engine produced two outputs for each upcoming appointment: a "likely to buy" prediction, and a tailored recommendation of which products to suggest. The recommendations were delivered to stylists ahead of each appointment so they could prepare, rather than guess on the floor.
63%
product sales uplift, line 1
71%
product sales uplift, line 2
8 weeks
trial duration
“I'm really enjoying the product. The messages remind me to keep products top of mind while I'm with a client. This is going to be so useful for stylists.”
Stylist, NovaLux
Why it worked
The salon didn't need a new sales process. It needed the right prompt at the right moment, grounded in what the engine could see about each client's history and the products that would actually fit them. Predictive intelligence delivered into the existing workflow — not another dashboard to log into.