How to Prepare Product Data in Four Steps

Before You Start

Preparation begins by addressing the right part of the catalogue rather than all of it.

Isolate the products that generate revenue

Start with the twenty products behind most of your sales. Once that group is done, most of the effect has been captured and the task remains finishable. Teams that begin with the whole catalogue usually abandon it in week three.

Measure where you are

How many products have which fields empty? Without that number, progress cannot be tracked either. A simple table suffices: product name, populated field count, missing fields.

Fix the field list

Decide up front which fields are mandatory: name, brand, dimensions, material, colour, usage, origin, price, stock. Without a fixed list, completion varies product by product.

Assign an owner

If it is unclear who completes these fields when a new product is added, the discipline decays within months. Maintenance needs an owner as much as the setup does.

Step 1 · Standardise Naming

The product name is the most-read and most carelessly written field.

Define the pattern

Establish a fixed pattern — brand, product type, distinguishing feature, size — and apply it across the catalogue. Without a pattern, the same product appears differently in different places. Writing the pattern down once and handing it to the team ends the debate for every product that follows.

Remove marketing adjectives

Superb, unrivalled and special appear in no query. The space they occupy should carry a real attribute instead. The name field is for definition, not praise.

Keep internal codes out of the name

A stock code written into the product name reduces readability and confuses customers. Codes belong in their own field.

One product, one name everywhere

Using different names on your site, on marketplaces and in your catalogue creates three separate products as far as systems are concerned.

Step 2 · Complete the Fields

An empty field means invisibility for every search using that attribute.

Dimensions and weight are not estimated

Both are used in the customer’s decision, in the shipping calculation and in customs classification. Writing an approximation creates a problem in three places at once.

Materials and contents in detail

Writing 100 per cent cotton, 180 gram instead of cotton measurably improves query matching. Detail raises the chance of containing the word the customer typed — and lowers return rates, because the buyer knows what they are getting.

State the use case

Customers type their need, not the product name. Details such as fits through narrow stairs, dishwasher safe or suitable for a child’s room correspond to real queries. These sentences are already asked of your sales team daily; the task is moving them onto the page.

Do not skip origin and certification

In export and corporate sales these two fields are direct elimination criteria, and they also ensure customs classification is correct.

Step 3 · Write in Need Language

A description should explain not the product but the problem it solves.

The first sentence states what it is

The opening sentence should cover what the product is, who it suits and what it resolves. Descriptions that begin with a story are neither read nor matched.

Bring customer questions inside

The five questions your sales team hears most should form the backbone of the description. Compiling that list takes one meeting and produces the most reliable content plan you will find.

Provide comparison information

Which model is right for me is the most common sticking point. Stating the difference between two models plainly speeds the sale and lowers returns.

Coverage matters more than length

A long but empty description performs worse than a short, dense one. The measure is the number of questions answered, not the word count.

Step 4 · Maintain the Order

The difficulty is not the setup but the table still being clean six months later.

A checklist for new products

Preventing publication until mandatory fields are complete is the simplest rule that makes the discipline permanent. Businesses without it repeat the same clean-up within six months.

Price and stock from one source

These two fields age fastest. Every manually updated channel eventually diverges from the others and it becomes unclear which is correct.

Scan quarterly

Measure the empty-field count every quarter. A rising number shows the maintenance rule is not working, and early intervention makes it easy to correct.

Manage removed products

Deleting a discontinued product’s page outright leaves both users and systems in a void. Redirection to a suitable alternative is required.

A Solid Digital Foundation

The most neglected part of product data discipline is what happens to removed and changed pages.

Permanent addresses produce permanent value

Deleting a product page invalidates every link pointing to it. Discontinued products should be redirected to a suitable alternative and error pages monitored; the criteria are explained in the Google Search Central documentation. A broken link returns the value you accumulated.

Seasonal products should not be deleted

Where a page for a recurring seasonal product is closed and reopened each year, the accumulation resets. Keeping the page and updating availability is the better route.

Filter pages need checking

Colour and size filters can generate hundreds of near-identical pages without anyone noticing. How those pages are managed directly affects the readability of the catalogue.

The same preparation works at four doors

Work done here pays in the assistant channel, in search, on marketplaces and in customs declarations at once. We set out that connection in our guide to the period and carry out the setup within digital consulting.

Frequently Asked Questions