What "on-brand product content" actually means (and why retailers keep getting it wrong)

Most retail brand books open with the same five adjectives. Premium. Approachable. Trusted. Innovative. Customer-first.

I've read maybe forty of these. The same five words always appear, in a slightly different order, paired with a moodboard of dark-mode UIs and lifestyle photography. The first thing that's wrong with how the industry talks about brand voice is that the language used to describe it is itself indistinguishable.

If your brand voice is a list of adjectives, you don't have a brand voice. You have a Pinterest board.

When that brand book gets translated into PDPs, the result is predictable. The product description says "premium quality" four times in seven sentences. It uses every adjective in the book and somehow still reads exactly like every other PDP in the category. The customer doesn't feel a brand. They feel a thesaurus.

There's a better way to think about brand voice for retail product content. It's not a feeling. It's a configuration.

Tone is not adjectives. It's decisions.

A useful definition of brand voice for PDPs is: the consistent decisions a brand makes when describing a product.

Specifically — who the buyer is, and is not. Which features matter to that buyer. Which features are noise. How the product fits into their day. What vocabulary the buyer uses, and what vocabulary they reject. How much expertise the brand assumes.

A retailer of professional kitchen equipment writes for a buyer who already knows what induction-compatible means. A homewares retailer selling the exact same pan has to translate the spec into a benefit. Same product. Two completely different PDPs. Both on-brand for their respective brand.

That's the part adjective-based guidelines miss. They tell you what the brand should sound like in the abstract. They never specify what the brand should say about a product when the buyer is a different kind of person.

The supplier-copy problem nobody admits

The single biggest reason a retailer's PDPs are off-brand is also the most operationally obvious. The descriptions are written by the manufacturer.

A manufacturer writes for every retailer that will sell the product. By definition that copy can't be on-brand for any specific retailer. It's on-brand for the manufacturer.

When the supplier description lands in your PIM and ships to your live site untouched, your store is publishing the manufacturer's voice. Not yours.

This is true even when the description is technically correct. Even when it's well-written. Even when it's grammatically clean. The voice is wrong because the voice doesn't belong to you.

Most retailers know this. Most also publish supplier copy on the majority of their catalog because the alternative — rewriting tens of thousands of PDPs — has historically been impossible at scale.

That's the gap automation closes. Not by replacing the writing, but by making it cheap to apply your brand voice configuration to every product in the catalog instead of just the hero ones.

Persona is not a buyer profile slide

Most brand books include a persona slide. Sarah, 34, lives in a major city, two kids, household income €80K, values convenience and quality, shops mostly on mobile.

You've seen this slide.

Sarah cannot help you write a PDP. She's too generic. She has nothing to do with the actual product.

A useful PDP-level persona looks more like this. A parent buying art supplies for two kids aged 6 and 9, looking for something that won't stain the dining table, will survive a school year, and won't make them feel guilty about spending more than €20.

The difference is that the second persona contains the decisions the buyer is making at the moment they're reading the PDP. Will it stain. Will it last. Is it worth the price. The PDP can speak directly to those decisions. The PDP cannot speak to "values convenience and quality."

Brand voice config that works for retail product content needs personas at this level of resolution. Not one Sarah. One persona per buying mode. Parents buying for two young kids. Parents buying a gift for someone else's kids. Older kids buying their own supplies. Teachers buying for a classroom. Same product. Four different PDPs.

This sounds like more work upfront. It is. It's also the work that makes the rest of the system actually generate on-brand content instead of smart-sounding filler.

Three components of a real brand voice configuration

If you're configuring a brand voice for the purpose of generating product content at scale, three components matter operationally.

Vocabulary boundaries. A list of phrases your brand says, and a list it does not. Game-changer. Unleash. Elevate. Redefine. Never. Built for. Designed around. Made for. Yes. The list doesn't need to be long. It needs to be specific enough that a system can be checked against it.

A persona library, indexed by product context. Three to seven personas, each tied to the products they buy and the decisions they care about. Not lifestyle bios. Buying-decision shortlists.

Attribute prioritisation rules. For each product category, the order attributes appear in the PDP. A camping retailer puts weight first for tents. Packed-size first for sleeping bags. Water-resistance first for jackets. Most retailers default to "whatever order the supplier feed delivered." That's the difference between an expert PDP and a data sheet.

A configuration with these three components produces consistently on-brand PDPs at any scale.

A brand book with five adjectives doesn't.

The test for whether your PDPs sound like you

Open three product pages on your own site. Cover the logo. Cover the URL bar. Read the first paragraph of each.

If you can't tell which retailer it is, your customer can't either.

If the description sounds like a press release, you're publishing manufacturer voice. If every product is "premium" and "innovative" and "designed for the modern customer," you're publishing brand-book voice — which sounds the same as every other retailer's brand-book voice.

The test is brutal because the PDP is an actual customer touchpoint, not a brand exercise. The customer doesn't read your brand book. They read three sentences and decide whether to trust your store.

Brand voice as a config, not a project

The reason this matters operationally: "rewrite the catalog in our brand voice" is the kind of project that gets approved every two years, scoped at one quarter, runs for six, ships at sixty percent, and quietly disappears off the OKR list.

It doesn't work as a project. It works as a configuration that gets applied automatically to every new product as the catalog grows.

That's the model EANScan is built around. You configure your vocabulary boundaries, your persona library, and your attribute prioritisation rules — once. Every EAN that enters the system gets a PDP generated against that configuration. New products inherit the brand voice the day they're added. Old products can be regenerated on demand.

You stop having a "brand voice debt" backlog. You have a brand voice that's just true of the catalog, all the time.

Where to start

If your PDPs currently sound generic — and statistically, they do — the work is sequential.

Audit five to ten of your own PDPs against the cover-the-logo test. Honestly. Most retailers fail. Replace the adjective list in your brand guidelines with vocabulary boundaries and a persona library indexed by buying decision. Pick a single category and pilot a non-generic rewrite. Measure conversion delta against the unchanged version of the same SKUs. Once the delta is real, scale the configuration across the rest of the catalog using a system that can apply it consistently.

Book a demo if you want to see what your products look like through this configuration. We run a sample of your real EANs before the call, against a brand voice config we infer from your existing site. You see your products in your voice. Not a generic example.