How SalesEvolution turned twenty years of duplicated data and a legacy ERP into a single, reliable customer picture at a hospitality wholesaler — in three weeks, not years.
The client is a 60-person hospitality wholesaler supplying commercial kitchen equipment and consumables to restaurants, catering companies and institutional kitchens across the region. The company has run the same ERP system for close to twenty years — installed when it was a third of its current size, and never seriously reconsidered since.
Technically the ERP held everything: every customer, every order since the system went live. In practice, almost nobody in sales opened it if they could avoid it. Two decades of different people entering orders meant the same restaurant could appear on three separate customer records, described three different ways. Some of the decision-makers listed as primary contacts had left those companies years earlier. And products carried different codes and descriptions depending on who had entered them.
People coped as best they could: their own spreadsheets, sticky notes, knowledge held in their heads. Nobody had a clear view of what a customer actually ordered, so nobody noticed when a regular buyer quietly stopped ordering a product family — sometimes for close to a year. Whenever management raised the idea of introducing modern sales or AI tools, IT and the team gave the same answer: our data is too messy for that. It became the justification for years of nothing changing.
SalesEvolution’s starting point was that messy data is not a reason to wait — it was the first problem to solve, and exactly the kind of problem AI is good at. BIZTAILORS connected to the existing ERP through the exports it could already produce, pulling customer and order history into a single interface without the company having to touch — let alone replace — the system finance and inventory still depend on.
From there an AI-driven cleansing process did the thankless work nobody had ever had time for: matching and merging duplicate customer records, flagging contacts that were obviously stale, and standardising product data that had drifted into a dozen different formats over the years — using the same categorisation logic we apply to making technical catalogues searchable, only turned on the company’s own customer and product data. At the end of the process, every real customer had exactly one current record.
Once that was in place, AI-driven account management started doing something nobody could do by hand: noticing when a long-standing customer’s ordering pattern quietly broke — a product family reordered every six weeks that had not moved in four months — and putting it in front of a colleague as a signal, instead of letting it go unnoticed for another year. New incoming leads ran through the same AI enrichment process as the historical cleanup, so the chaos could not quietly rebuild itself.
“For years we told ourselves we had to sort our data out before we could do anything like this. It turned out sorting the data out was the project — not something that had to happen first.”
Sales Director, from the client team
How an early-stage startup built on SalesEvolution from day one, turning salespeople with no experience into confident hunters in two months.
How SalesEvolution turned a 15,000-part catalogue and years of accumulated expertise into a source engineers — and AI search engines — actually find.
See how SalesEvolution would apply this same approach to your pipeline. Start with a free 30-minute strategy consultation.
Book a strategy consult →