If your sales rep can't find the part number, your catalog doesn't exist

If your sales rep can’t find the supplier’s part number in under a minute, your catalog doesn’t exist. It can be digitized. It can be beautifully organized. But if it lives in a PDF that only the person who saved it can open, for the sales floor it’s dead data. The problem isn’t the file format. It’s that nobody can search it. What can’t be searched doesn’t get used. Behind this there are two real numbers and a 15-minute test to find out whether yours is alive.

The scene that repeats every week

A rep is on the phone with a customer. The customer asks whether you carry a specific part number, or something equivalent from another manufacturer. The rep can’t say “hold on.” So they open the shared folder, find the supplier’s catalog, download the PDF, and now comes the hard part. Searching inside 300 scanned pages with a search box that can’t tell “ref.” from “reference”. In the end they ask the one person who always knows where everything is. That person drops their own work to do someone else’s.

That’s the real cost of a catalog you can’t search. It’s not the typing hours. It’s that the answer arrives late, or never. And every week, hours of the wrong person’s time go into finding what was already in the building.

Why the obvious fix fails

When the catalog gets heavy, there are two usual answers. The first is typing it all in by hand. Nobody finishes, because while someone types, three new catalogs arrive. The second is buying a search tool and pointing it at the PDF as is. It works for two weeks. Then the variants show up. The same brand spelled four ways. Part numbers with stray spaces. Abbreviations only the person who typed them understands. The tool starts clean, and three months later it finds nothing.

Both fail for the same reason. They attack the symptom, which is search, and not the cause, which is that the source data isn’t ready.

What we would do today

Order matters more than the tool. Before anyone mentions AI, the work is this.

First, eliminate. Most accumulated catalogs never get consulted again. Ask each team which sources they actually use. Archive the rest, no ceremony.

Then, standardize. One way to write every brand, every part number, every unit. If the same brand appears four different ways, no search will work, no matter how good the tool is.

Next, simplify. One place to search, not five folders. The person selling shouldn’t need to know which catalog each part came from.

And only at the end, automate. Load new catalogs without anyone typing. Classify, infer equivalents, suggest alternatives. AI comes in when the rest is healthy, not as a patch over broken data. That’s the logic of the Zero Friction Method. Eliminate before you automate.

The two numbers

From a real catalog project, two production numbers. More than 112,000 product references extracted and searchable from supplier catalogs. And a 388-page industrial catalog, digitized in a single run, with nobody typing any of it.

The second number doesn’t matter because it’s big. It matters for what it means. Nobody stopped selling to make it happen. Data work can’t compete with the day’s work. If cleaning the catalog means pausing operations, that cleanup never happens, and the PDF wins another season. That’s why manual entry is the red line. Not for elegance. It’s because we already know how the alternative, typing it all in, ends.

For the sales floor the result is simple. The part number turns up during the call, not three days later. And without interrupting the person who knows everything, who can finally get on with their own job.

The 15-minute test

Pick 20 part numbers from a single source, ones you actually sell. Hand them to someone who isn’t the person who always knows where everything is. Time it.

• 18 or more found within minutes. Your data is alive. The problem is somewhere else, and this catalog is not your priority.
• Between 10 and 17. The data exists, but access depends on one person. That’s real risk, not just slowness. The day that person is out, the search stops.
• Fewer than 10. The catalog isn’t an asset, it’s an archive. Write down where each search breaks. Is the data missing. Is it there but unreadable. Or is it there, readable, and nobody knows it exists.

That note, in writing, beats any audit. In fifteen minutes you’ll have a map of the problem built on your data, not on a report’s.

If the test goes badly, you don’t need an AI project to start fixing it. Start with the data, like we wrote in data before AI. And if you want to see how we handle the supplier catalog end to end, it’s described in our catalog management capability.

If you want to audit your catalog with this same test, start with the checklist. And if you want to talk it through, write to us or get to know us on About.