Buskara

Catalogue

Your catalogue is written for the supplier, not for the person buying

References, codes and technical names straight out of the supplier's file. What to do when the customer writes in a different language to ask for exactly the same product.

Daniel Silva Buskara
7 min read

In short

  • Most catalogues inherit the vocabulary of the supplier's file, which was written for logistics and not for selling.
  • The customer's vocabulary already exists in the building: in the site's searches, in support emails and in reviews.
  • There are three ways to bring the two closer, with very different costs: synonyms, custom fields and rewriting the product pages.
  • Rewriting everything is almost always the wrong call. Rewriting the hundred products that keep the store running is almost always the right one.

Open any product page in your store and read the title out loud. If it sounds like a warehouse label, it's nobody's fault: it's the normal path a catalogue takes. The file arrives from the supplier with a name, a reference and attributes, you import it so you don't have to write three thousand product pages by hand, and what's in it becomes what the store says.

The problem is that this file was written for another purpose. "Vela Windbreaker 2.5L" identifies an item in an inventory with no ambiguity at all. It isn't a phrase anyone types into a search box. In PrestaShop stores this path is almost universal, and it's one of the reasons the native search disappoints: it looks for the words that are there, and the words that are there are the warehouse's. See search module for PrestaShop.

Two languages for the same product

The mismatch always takes the same shape. On one side, the catalogue names; on the other, the customer describes.

The catalogue says The customer writes
Vela Windbreaker 2.5L light jacket for the rain
Nimbus Road 14 running shoes for tarmac
Sneaker Court Leather White white leather trainers
MX-3 task chair with lumbar support office chair for my back
E27 opal pendant luminaire ceiling light for the dining room

None of these searches finds anything by matching words, and every one of them has an obvious answer in the catalogue. Notice the pattern: the customer writes a category, a use and a visible characteristic. The catalogue writes a brand, a model and a technical specification.

This isn't a small-store problem. In the 2026 revision of the Baymard Institute search benchmark, which assesses more than 170 sites and apps, 43% have trouble with use-case searches ("for running in the rain"), 39% with feature searches and 20% even with product type searches. Those are exactly the three right-hand columns of the table above.

Where the customer's vocabulary is

The good part is that you don't have to guess or commission a study. Your customers' words are already in the building, in four places:

  • The searches on your own site, above all the ones that returned nothing. It's the most direct source there is: the request is written down, and it failed.
  • Customer support emails and conversations. "That lamp you've got in the window" is the real vocabulary.
  • Product reviews. Someone writing a review describes the product by what it's good for, not by its spec sheet.
  • The external search terms that already bring traffic. They tell you how people arrived, which is a colder version of the same information.

An afternoon reading these four sources with a sheet of paper beside you usually produces fifty pairs of words, and those fifty pairs cover most of what's missing.

Three ways to bring the two languages closer

Approach Effort Effect on internal search Effect on Google When it ages
Synonyms in the search engine Minutes per term Immediate and complete None When the vocabulary changes
Custom fields (use, occasion, audience) Hours per category High, and it gives you filters as well Indirect, through filter pages When you add categories
Rewriting titles and descriptions Days per hundred product pages High Direct and lasting When the supplier changes the range

All three are compatible and they solve different things. Synonyms solve it today, they take effect without reindexing, and they leave no trace outside the store. With a search that understands the request, a good part of them stops being necessary, but no engine guesses a word that exists nowhere in the catalogue. Custom fields solve navigation, because "for outdoor use" stops being a word buried in the description and becomes a filter in the column. Rewriting is the only one that also works outside the store, and the only one that costs real money.

There's a fourth case that is none of these and is worth ruling out first: the word does exist on the product page, but in a field the search doesn't read or that carries zero weight. There no vocabulary is missing, configuration is, and the place to check is products that don't show up.

Pros and cons of rewriting the catalogue

For Against
Improves internal search, Google and how the product page reads, all at once It's the most expensive of the three, and it's manual
It stays in the store: it doesn't depend on any tool's configuration A badly configured import wipes it all on the next sync
Your own text sets the store apart from dozens of competitors working from the same file It needs someone in charge, otherwise quality swings from page to page
It cuts returns, because expectations end up closer to the product The payback takes weeks to show up, unlike synonyms

The line about the import is the one that usually hurts. Before you rewrite anything at all, check that the sync with the supplier is configured not to overwrite the fields written by hand. We've seen two weeks of work disappear on a Sunday night.

What to do this week

We don't recommend starting with the whole catalogue. We recommend starting with the part of it that keeps the store running:

  1. Pull the list of products that make 80% of sales. It's usually fewer than two hundred.
  2. Cross it with the list of terms searched with no results. The intersection is your working list.
  3. For those, write a title that says category, use and the characteristic that sets the product apart, in that order. The brand and the model are still there, at the end.
  4. For the rest, fix it with synonyms and move on.
  5. Book a review in three months, with the same two reports.

The difference between this and a rewrite project is that this one starts on Monday and shows results the following week.

Frequently asked questions

Why doesn't my catalogue find products that exist?

Almost always because the text on the product page was imported from the supplier and uses inventory vocabulary, while the customer writes usage vocabulary. The product exists, it's in stock, but none of the words typed into the search appear in the indexed text.

Synonyms or rewriting the product pages: which do I do first?

Synonyms, because they fix things the same day and carry no risk. Rewriting comes next, and only for the products that justify the work, typically the ones that make most of the sales and show up in failed searches. One caution: synonyms are not for typos, which are handled somewhere else and by another mechanism, as explained in why typo correction can't be silent.

How do I find the words my customers use?

In the site's searches with no results, in customer support emails, in product reviews and in the terms that already bring traffic from search engines. All four sources are free and already in the building.

Does rewriting product titles hurt Google rankings?

No, as long as the page address doesn't change and the title still contains the name the product is searched for by. Titles that describe category and use usually improve rankings, because they get closer to what people write in the search engine.

What stops the supplier from wiping the text I wrote?

The import configuration. Before rewriting product pages, check that the catalogue sync is ignoring the title and description fields, or that it keeps them when they've been edited by hand.

Catalogue Search

Was this article useful?

Daniel Silva

Writes about search and product discovery on the Buskara blog.

We only write when we have something to say

Get new articles on search, conversion, e-commerce and what we learn from real data.

No data sharing. Cancel whenever you like.