Typography as a brand asset: where your typeface actually counts

Your typeface is a brand asset until the channel overwrites it

Typography as a brand asset works, and it only works on the surfaces you own. On your own webshop, on your packaging and in your imagery, your typeface carries recognition. On a marketplace page it does not. The bol Retail Media guidelines for Brand Pages say it plainly: changing the typeface is currently not supported. The same sentence appears for text colour.

This is the question you ask before you buy a visual identity: what is left of this brand on the narrowest surface it lands on?

The answer is remarkably consistent. Your typeface survives everywhere your brand arrives as an image, and disappears everywhere it arrives as data. Your packshot, your packaging, your banner artwork and your A-plus modules are images. Your product title, your brand page and your bullet points are fields in someone else's database. Same visual identity, two completely different regimes.

What makes typography a stronger brand asset than a brand colour?

Letters are shapes. Shape beats colour, and that has been measured.

In the largest study of distinctive brand assets to date, the Ehrenberg-Bass Institute scored over a thousand assets on recognition across 21 product categories, four countries and nine years of data. They measured two things: how often category buyers link an asset to the correct brand, and how much of that association points to your brand rather than a competitor's. Shape-based assets such as logos, packaging shapes and symbolic imagery come out on top on both. Colour comes out at the bottom.

A typeface belongs in that first category. It is not a colour you share with the rest of your shelf, it is a drawing. The serif on your g, the height of your lowercase, the angle at which your a closes: those are shape assets someone can recognise without your name next to them.

There is a practical argument underneath as well. Owning a colour takes decades or a trademark, because the colours in your category were not chosen at random: they belong to the category, and your competitors are already sitting in them. The space for typefaces is far larger, and in e-commerce that space is barely used. With almost every brand that comes to us we inherit the default typeface of the webshop theme, because no decision was ever made about it.

That is the opening. If your competitors all run on that same system typeface and you do not, you are recognisable without anyone being able to explain why. That is exactly what a brand asset is supposed to do, and it is the same discussion as in branding is not a logo: the logo is not your brand, it is one of the carriers.

What a one-time typeface licence changes about the maths

The business model underneath custom typography has shifted, at least at one supplier that is aiming at exactly that.

Branding agency Koto launched its own type foundry in July 2026, CCType, with a variable display serif as its first release, with adjustable axes and OpenType alternates. The model is a one-time purchase instead of recurring licences, explicitly intended to make carefully drawn type reachable for freelancers, small studios and brands directly. CCO Jowey Roden describes the reason as a typeface that has to hold up everywhere: on a website, in motion, in campaigns, on products, on billboards.

That reason is precisely the problem e-commerce brands have. The number of surfaces your brand appears on grows every year, and fonts that worked in one context collapse in another. A typeface that looks right on an A2 poster can be unreadable on a thumbnail in a search result. A variable typeface with a weight axis and an optical axis solves that, because you adjust it per size without switching families.

The wider movement points the same way. The PRINT Type Report 2026 states that brands fall flat without a distinct voice and that clients are asking for custom fonts, arguing for flexible identity systems rather than static brands. That is not a trend forecast, it is a reaction to something you can see happening yourself: the more content gets generated automatically, the more everything starts to look alike.

One supplier does not make a market. What has changed is that the ladder between free system typeface and full custom has more rungs than it did four years ago, and that the middle rungs are now within reach for a brand with a realistic budget.

What exactly are you buying when you buy a typeface?

Not a file, but permission. And that permission is cut up in a way many owners only discover when something goes wrong.

A type licence usually splits into desktop (your designers may work with it), web (the typeface loads in your webshop), app (your mobile application) and sometimes broadcast or embedding in a product. Each part is billed separately. The web licence is the part that catches e-commerce brands out, because at many foundries it is tied to a maximum number of pageviews per month. Grow past that ceiling and you move up a tier.

That makes the choice less a matter of taste and more a calculation about where you will be in three years.

Two things you always check before signing. First: may you embed the typeface in a PDF you send to a printer or a marketplace? Some licences allow that only in limited form, and a packaging printer needs an embeddable file. Second: what happens when you stop paying? On a subscription model the typeface drops out of your webshop as soon as you stop, and your entire visual identity falls back on a substitute.

We run into this with brands that had their visual identity made four years ago and can no longer trace which licence was bought. Record that in your brand file, along with the rest of your brand identity.

Where does your typeface survive on a marketplace, and where does it not?

The dividing line is simple: your typeface survives where your brand arrives as an image and dies where it arrives as a text field.

That sounds obvious and it is not, because most brand books are written from the webshop outwards. There you control everything. On bol and Amazon you fill in a form. The bol Brand Page guidelines leave little room:

• You cannot change the typeface, and you cannot change the text colour.
• You do choose the background colour freely, as long as it contrasts enough with that fixed text colour.
• You supply the banner without a logo and without text.
• The logo goes in separately, as a PNG with a transparent background, at least 256 pixels high.
• One Brand Page per seller.

The bottom three rows are the rows where it hurts. The moment an AI assistant summarises your product, nothing of your visual identity is left except the words you put in your feed. There your writing carries the work your typeface does elsewhere, and that is one of the reasons brand positioning for AI agents is a different craft from designing a visual identity.

The practical consequence for your design brief: do not have your visual identity designed for the webshop and then squeezed down for everything else. Design for the narrowest surface. What holds up there holds up everywhere. Finish the brand book first and only then look at where it has to land, and you find out too late that half of your carriers are unavailable. That is also the pattern behind why brands fail on marketplaces.

How do you pick a typeface that still works on a thumbnail?

Judge a typeface at the size buyers actually see it, not at the size you present it.

An identity presentation shows the typeface large on a white page. Everything looks good there. Your buyer sees your brand name on a packshot in a row of search results, on a phone, surrounded by competitors, on the train. That is the real test, and you can run it this afternoon by looking up your own product on bol or Amazon and looking at your thumbnail as it sits there.

Where it goes wrong:

A low x-height is the first thing to go. A typeface with a tall lowercase stays readable when scaled down, because the body of the letter gets more room within the same line height. Low x-heights look elegant at large sizes and vanish at small ones.

Counters fill in on board. The holes in your a, e and o are the first to close up when scaled down and when printed on uncoated material. Check that specifically on a printed proof, not on your screen.

Two weights give you no hierarchy. You need at least three that clearly differ, because on packaging you quickly sit at four levels of information: brand name, product name, variant and content size.

The fourth is less visible and costs the most work afterwards: figures. Price, content size, batch number, nutritional values. Check whether the typeface has tabular figures that line up neatly, otherwise your nutrition table sits crooked and you only see it on the press proof. And finally, look at how the typeface behaves on screen: without an optical axis you use the same drawing on a thumbnail as on a billboard, and that is exactly where typefaces collapse.

Then run two proofs before you choose. Put your brand name on a blank packshot and view it at the size it sits at in search results. Print your packaging design at a third of actual size and put it a metre away. Whatever you cannot read at either can be used in your webshop, but nowhere else. What a buyer actually looks at on packaging is covered in where the buyer looks.

What the Bob's Red Mill rebrand shows about modernising

A new typeface can be technically superior and still destroy value.

Bob's Red Mill was rebranded in 2026 by Turner Duckworth with its own typeface, colour families per product category and wrap-around storytelling on the packaging. As an applicable system it is textbook work: consistent, shelf-friendly and easy to extend to new products. But the brand also removed founder Bob from the logo, two years after his death, and that drew heavy pushback from buyers who read it as erasing the man who was the brand.

The lesson is not in the taste debate. The lesson is that the system and the anchor are two different things. The typeface, the colour system and the packaging structure are the system: you may renew those, and they deserve renewal the moment they work against you. The anchor is the asset buyers recognise you by, and for one brand that is a face, for another a silhouette, and for a third exactly that dated typeface you wanted to replace.

So before you start, list every visual asset your brand has and note whether a buyer would recognise you by it without your name. Everything that scores yes is anchor and comes along. The rest is system and can go. That order is the difference between modernising and starting over, and it is worked out further in rebranding without losing your recognition.

With brands that replace their typeface there is one more trap. If your old typeface sat on your packaging for years, your recognition probably lives in the combination of that typeface with your colour and your format, not in the typeface alone. So replace one at a time. Buyers forgive a new typeface on a familiar pack. They get lost with a new pack in a new colour with a new typeface.

How we approach typography in a brand identity

We start at the surfaces and work back to the typeface, not the other way around.

At Castagnola that was the whole job. The brand sold stainless steel bottles, thermos cups, placemats and kitchen textiles on Bol.com, on Amazon and on its own Shopify webshop, and those three channels did not look like the same brand. The packaging did not follow the style of the listings, the listings did not follow the style of the webshop. That is not carelessness from whoever worked on it last. That is what happens when a visual identity is made for the channel where you control everything, and then has to land on two channels where you do not.

So in a brand identity project we put the channel list down first. Own webshop, packaging, packshot, A-plus content, marketplace brand page, retail media, feed. Per channel we note what you decide and what the channel decides. That list then determines the role typography gets: does the typeface carry your recognition, or is it support while your anchor sits in a shape or an image?

For brands that sell heavily on marketplaces the answer is usually the second. At ELVOU it was the other way round: that supplements brand was deliberately built on its own webshop rather than on marketplaces, so it keeps control over every surface a buyer sees. On a brand like that a distinctive typeface is one of the cheapest ways to become recognisable, because there is no form in between taking it away again.

That choice then determines how we build the packaging design and the listings, so that one system stands instead of three separate designs that happen to resemble each other. Across six years of Oase Creative and 200+ reviews that is the pattern that comes back most often, and the most expensive one to repair afterwards.

What you sort out in three mornings

Three actions that cost no design budget and tell you whether your typography is doing work.

Make the channel list. Write down everywhere your brand appeared in the past year. Own webshop, packaging, packshot, A-plus content, brand page on bol or Amazon, retail media, feed to comparison sites. Next to each channel, note whether your typeface gets in there. For most brands half of them say no, and that is the first time anyone writes that down.

Run the thumbnail proof. Look up your three best-selling products in the bol or Amazon search results, on your phone, and put them next to your two largest competitors. Can you still see a difference at that size? If not, your typography carries no recognition on that channel, however good the identity is otherwise.

Find your licence. Work out which typeface you use, which licence sits under it, who pays for it and where the ceiling is. If nobody in your company can answer that within fifteen minutes, you have an exposure open that you only discover on the day your webshop falls back on Arial.

If your channel list comes out roughly half and half, which is what happens with most brands running both their own webshop and a marketplace, pick the anchor that survives on the narrower half and use the typeface to finish the wider half. Want to fill in that channel list for your own brand? Get in touch and we will walk through it in half an hour.