Elli Cares app design: a DesignRush awards finalist
The Elli App is a finalist in DesignRush’s June 2026 Best App Design Awards, Health & Wellness. What the jury noticed, and what health apps can borrow.
11 Nov 2026
AI startup branding: how not to look like every AI company
AI startup branding that stands out: the names, colours, imagery and claims that blur together, and an exercise to find what’s specific to you.
Short answer
To stop your AI startup branding looking like every other AI company, stop branding the AI and brand the job it does. Take your name, colour, imagery and claims from your customer’s world and your evidence. Purple gradients, sparkles and “copilot” names only say AI, and buyers already assume that.
An illustrative test. Take five AI startup homepages, cover the logos and try to guess which one reviews contracts, which one reads scans and which one writes sales emails. Same gradient, same glowing orb, same promise to “supercharge your team”. You’re guessing.
Buyers can’t tell them apart either, and they don’t try. In a Semrush survey of US B2B professionals published in July 2026, only 7% said they notice a vendor in an AI answer because they recognise the name. 53% notice it because it closely matches their use case. Fame barely registers. Fit does the work.
Why does AI startup branding all look the same?
Because most AI startups brand the technology, which is the one thing they all share. Many build on similar models, pitch to the same investors and start from the same website templates. When the brief says “show that we’re AI”, everyone reaches for the same symbols.
The sparkle icon now means AI the way a floppy disk means save. Everyone understands it, nobody chose it and it tells a buyer nothing about you. A brand made of shared symbols is a category badge. Handy on a conference lanyard, less so when a buyer is choosing between three of you.
Category symbols versus specific choices. Framework by Dtail Studio.
What makes AI brands blur together?
Four habits, and most AI brands have at least three. Check yours against the list.
Names.
Copilot, Assistant, Mind, Neural, Genius, anything ending in -ly or -ify, and a .ai domain doing the job a name should do.Colour.
Purple to blue gradients, dark mode with a soft glow, one neon accent. It photographs well and looks like everyone.Imagery.
Glowing brains, particle clouds, robot hands, abstract orbs and the sparkle. None of it shows a customer, a task or a result.Claims.
“AI-powered”, “10x faster”, “the future of work”, “your intelligent assistant”. These describe a category, so they fit every competitor’s homepage too.
None of these is wrong on its own. Purple is a perfectly nice colour, and some of our favourite designers would fight you over it. The trouble is the full set. It says you’re in the category and nothing about why you’re the one to pick. Our post on how typography shapes brand identity covers one of the quieter ways out.
Should you put “AI” in your name and your claims?
Rarely in the name, and late in the claims. Lead with what the product does for someone. Mention AI when you explain how.
There’s evidence this matters. In experiments with more than 1,000 US adults published in the Journal of Hospitality Marketing & Management, researchers at Washington State University and Temple University found that describing a product as using artificial intelligence lowered emotional trust and purchase intention. The effect was stronger for high-risk purchases such as medical devices and financial services. Wider sentiment points the same way. Pew Research Center found in 2025 that 50% of US adults are more concerned than excited about AI in daily life, up from 37% in 2021.
Our view: a name with AI in it will date the way names with “e-” and “cyber” did. Name the company for what it does, or for something ownable, and let the product page say what model is under the bonnet. Buyers in health or finance want to know what happens to their data and who checks the output long before they care about the model.
How do you find what’s specific to your AI startup? A summary of this section.
How do you find what’s specific to your AI startup?
Look at your customers, not at other AI companies. This exercise takes a founder and one colleague about an hour.
Write the recommendation sentence.
How would a happy customer describe you to a peer, in one line? Use their words from calls, not yours.List what exists only because of who you serve.
A data source, a regulation, a workflow, a file type. A legal AI in London deals with clause libraries. A radiology AI in Berlin deals with scans and sign-off. Those are your raw materials.Find the object.
Pick one real thing from your customer’s day, like an invoice, a meter reading or a contract clause, and start the visuals there instead of with an orb.Choose one claim you can prove.
“Recovered GBP 12,000 of missed invoices in month one” (illustrative) beats “AI-powered finance”. Use EUR or USD if that’s where your customer is.Run the swap test.
Put a competitor’s name on your homepage. If nothing reads wrong, keep digging.
Whatever survives step five is your brand. Our post on how your brand already exists goes further on finding it rather than inventing it.
What does a specific brand look like in practice?
One built from what the product does. Cubee isn’t an AI company, but it’s a tech brand in a category with plenty of clichés of its own: energy management. The technology was strong and the visual presence was fragmented. The brand became a geometric symbol system based on energy flow and connection, with a bold logomark that balances precision and warmth, built to work across decks, digital and physical. Cubee now has 2,000+ installations. “Dtail helped us launch with a brand and product that stood out instantly,” says David Vondracek, Founder and CEO. The Cubee case study shows the system, and our guide on how to brand a clean energy startup applies the same thinking to that sector.
How we’d approach it
We’d start with your customers’ language, then build the visuals from it.
Audit the category.
Line up your competitors’ names, colours, imagery and claims, and mark everything you share with them.Run the exercise with your team.
Recommendation sentence, raw materials, object, proof and swap test.Write the positioning and claims first.
One sentence a buyer would repeat, and the proof that backs it up.Design the identity from the raw materials.
A name or naming direction if you need one, then logo, colour, type and imagery only you could use.
Bring three customer call recordings, your current deck and the competitor you lose deals to most. More on how we work.
FAQ
Is a .ai domain a bad idea?
No. It’s a fine address. It just shouldn’t be the most distinctive thing about you.
Do we have to avoid purple?
No. Avoid the purple gradient as your entire identity. A colour used with intent, next to type and imagery that are yours, is fine.
How do we brand when the product changes every month?
Brand the customer problem, which changes slowly. Features move. The job usually doesn’t.
Should an AI startup look technical or friendly?
Whatever your buyer trusts. A compliance team and a design team want different things. Our clarity test helps you check you’re talking to the right one.
Try this today
Screenshot your homepage hero and the heroes of your three closest competitors. Cover the logos and show them to someone who knows your customer. If they can’t pick yours in ten seconds, run the swap test and start with the claim.
If you’d like help finding what’s specific to you, let’s talk.
About the author: Alma Johnson is Content Lead at Dtail Studio and writes about brand strategy, positioning and how founders explain what they do.
Sources: Semrush, How AI shapes B2B buying (Margarita Loktionova, July 2026), Washington State University, Using the term ‘artificial intelligence’ in product descriptions reduces purchase intentions (July 2024), Pew Research Center, How Americans view AI and its impact on people and society (September 2025)
Digital health UX agency: 7 US studios to consider
Choosing a digital health UX agency in the US: 7 studios with real health work, the criteria we used and questions to ask on the first call.
Workflow automation UX design: build the log first
Workflow automation UX design people trust: why users switch automations off, and how status, previews, run history and undo keep them on.
