What is a brand sprint for a startup?

A brand sprint is a short workshop, often a single week, where a startup's founders agree on what the company is and how it should look before design work starts. The useful outputs are decisions: a plain sentence about what the product does, a list of companies the brand must not resemble, and what the first screen must show. A list of values on its own is not enough for a designer to build from.

How long should a brand sprint take?

One working week is enough for most seed-stage startups. Spend Monday on the plain sentence, Tuesday choosing three companies to avoid looking like, Wednesday turning those into directions, Thursday testing a rough hero and Friday writing a one page handoff. Longer sprints tend to produce more adjectives, not more decisions.

Who should be in a startup brand sprint?

Keep it to three or four people who will still be at the company in a year. The founder who sells must attend every day, because he has heard buyer objections firsthand. Add whoever runs marketing and, if possible, the designer who will build the site.

What should come out of a brand sprint?

A single page a designer can read in five minutes. It should hold the plain sentence, the name if settled, three refusals written as not this, so instead that, the one thing the first screen must show, and what is still open. Logo, color and type files come later, built from that page.

Do I need a brand sprint before redesigning my website?

If your team cannot agree in one sentence what the product does and for whom, yes, because the designer will otherwise guess. If you already have that sentence, a one or two day version focused on the three companies to avoid is often enough. Either way, finish the sprint before the designer's first day.

How do I stop my startup brand looking like every competitor?

Name three companies your buyer has seen this year and write down exactly what you refuse from each: the color, the image style, the headline shape or the kind of proof. Then decide what you will show instead, taken from your buyer's working world. Test it by pasting each company's logo on your draft hero to see if it still fits.

Brand Sprint for Startups: A One-Week Version

A brand sprint for startups is a short, focused stretch, usually one week, where the founders and whoever runs marketing agree on what the company is and how it should look, so a designer can start building. The version that works ends with three decisions, not a list of values: one plain sentence about what the product does, three companies your buyer already knows that you refuse to look like, and the one thing your first screen must show. Run it Monday to Friday and hand the designer a single page.

This matters because a stranger decides what kind of company you are in about four seconds, and he decides it from the surface. You know your product and your buyer better than anyone who will ever land on your site. A sprint is the week you write that knowledge down in a form a designer can use. Most sprints write down adjectives instead, and adjectives are the one thing every company in your category already agrees on.

Monday, 9:40: Innovative, Trustworthy, Fast

Picture a seven person startup. This is an example, not a client. It sells an AI agent to outpatient physical therapy clinics. When an insurer denies a claim, the agent reads the denial letter, the visit notes and the insurer's own policy, then drafts an appeal letter that quotes the policy section the denial ignored. The clinic's billing manager checks it and sends it.

It is Monday morning in a borrowed meeting room. The two founders are there, plus the first marketer, who started three weeks ago. The whiteboard is headed OUR VALUES. Under it, in green marker, someone has already written Innovative, Trustworthy and Fast. Below those sits a fourth word, Human, with a question mark after it. Twenty three yellow sticky notes from a personality exercise cover the left edge of the board. The marketer's Figma file, called Brand refs, holds 31 screenshots of sites the team likes.

A freelance designer starts on the website next Monday. She needs a direction by Friday at 5pm. By lunch on Monday the team has a fifth word, Approachable, and an argument about whether Fast sounds cheap.

Why the adjective list gives the designer nothing to build

Here is the claim most brand sprint advice gets wrong. The exercises that fill a whiteboard with values feel productive because everyone agrees. That agreement is the problem. Trustworthy is on the values list of the practice software the clinic already runs. It is on the list of the billing firm the clinic fired last year. It is on the list of every AI startup that pitched the clinic at its last conference. A word that all three would sign is a word that tells a designer nothing about you.

So the designer does the only reasonable thing. She opens the 31 screenshots, finds what they share, and gives you the average. Navy for trust. A soft gradient for innovation. A photo of a smiling therapist for human. The first round comes back looking like a company the buyer has already met, and the revision rounds turn into a debate about what Trustworthy meant.

You can watch this happen from the outside. The founder sends three competitor links into the team chat with no comment. The marketer adds eleven more screenshots to Brand refs. Someone asks for the version that is the same but warmer. The launch date moves from the first of the month to the fifteenth, and the old site stays up a little longer.

The refuse list: three companies your buyer already knows

The useful output of a brand sprint is a short list of three real companies you refuse to look like, each with the exact thing you refuse. A refusal costs something, so people argue about it, and the argument is where the decisions get made. Nobody argues about Innovative.

The rule for picking them: each one must be a company your buyer has seen this year. Not a company you admire, and not one you dislike. In a real sprint you write the actual company names on the board. For the appeals startup they are easy to pick, and here they stay as descriptions:

The practice software every clinic runs. Navy and white, a stock photo of a therapist bending a patient's knee, the word solutions in the headline. The refusal: no stock clinic photos and no navy. So instead the hero shows a real denial letter and the drafted appeal side by side.
The billing firm many clinics hired and then left. A page of guarantees about recovering more money, with green arrows on dollar charts. The refusal: no recovery promises and no money graphics. So instead the proof is the quoted policy line, highlighted inside the letter.
The AI note taker that pitched the clinic last month. A dark page, a violet glow, a chat bubble as the hero. The refusal: no dark mode hero, no glow, no chat box. So instead the page is light and reads like paper, because the product's output is a letter.

Look at what came out. Three sentences that each start with no and end with so instead. A designer can build from every one of them on Monday. Notice too that none of them needed the word trust. The highlighted policy line does that job better than the word ever could.

A one week brand sprint, day by day

Keep the group to three or four people who will still be at the company in a year. The founder who sells should be in the room every day, because he has heard the buyer's objections in person and nobody else has.

Monday: write the plain sentence. What the product reads, what it writes, and for whom, with no adjectives allowed. For the example: it reads denial letters and drafts appeals for physical therapy clinics. Then list every company your buyer sees in a normal working week, including the ones you think are not competitors.
Tuesday: choose the three. One incumbent the buyer uses every day, one vendor the buyer has been burned by, and one from your own category. Screenshot each homepage. Circle the exact parts you refuse: the color, the kind of image, the shape of the headline, the type of proof.
Wednesday: turn each refusal into a direction. Write every one as not this, so instead that. The so instead half has to come from the buyer's world. For a clinic that means letters, policy numbers and claim dates, not abstract shapes.
Thursday: test it. Sketch one rough hero from the three directions, even in a slide tool. Then paste each refused company's logo onto it. If any of the three logos looks at home there, rewrite that refusal and sketch again.
Friday: write the handoff page and erase the board. One page, described below. The adjectives stay in a photo on someone's phone. They do not go to the designer.

Turn every adjective into something on the screen

You do not have to throw the values exercise away. Use it as raw material for one more step on Wednesday afternoon. Take each word from the board and write next to it what a stranger would have to see on the page to believe it. If nobody can name a thing on the screen, cross the word out.

For the appeals startup it goes like this. Trustworthy becomes the insurer's policy section, quoted word for word inside the sample appeal. Fast becomes a line under the letter: drafted 4 minutes after the denial arrived, shown as an example. Human becomes the billing manager's name in the Approve and send button, because a person signs every letter. Innovative has nothing to point at, so it goes.

This is the same move the designer would have to make anyway, guessing. Doing it in the room means the guess is made by the people who know the buyer. It also gives the first marketer something she can reuse in every ad and email for the next year.

What lands on the designer's desk on Friday

The handoff is one page, not a deck. A designer should be able to read it in five minutes and start sketching the same afternoon. It holds six things:

The plain sentence from Monday.
The name, if it is settled. If it is not, settle it before the designer starts. The guide to naming an AI startup covers that week on its own.
The three refusals, each written as not this, so instead that, with the circled screenshots attached.
The one thing the first screen must show. For the example, a denial letter and its appeal.
The adjective conversions from Wednesday, with the crossed out words left visible so nobody adds them back.
What is still open, so the designer knows where she can push.

Color, type and logo files are the designer's job, built from this page. If you want to know which of those files a seed company actually needs, the post on what a seed-stage brand identity should include lists them. For what a direction looks like once it is built, the Chariot case study shows a frontier AI lab's brand direction, landing page and site, taken from kickoff to public launch in seven weeks, with an illustration and animation in each section that shows what the model does before the text explains it.

What a week of adjectives costs

The cost is not the week itself. It is what the week sends into the next two months. A designer working from adjectives needs more rounds, because each round is a guess at what the words meant. Those rounds are paid in weeks, and at seed you count runway in weeks.

Then the site goes live and starts doing its real job, which is being read by strangers who do not know you. Across all sectors the median gap from seed to Series A is around 616 days, and roughly 70 percent of seed companies never raise one. Whatever site you ship after this sprint is likely the one an investor, a hire and a clinic owner all open before they ever speak to you. If it looks like the practice software with a violet glow added, they file you next to companies they already decided about.

Sometimes the sprint shows you need more than a new look. If the plain sentence on Monday is different from what your current site claims, read rebrand or refresh before you brief anyone, because a new look on an old claim only makes the old claim look deliberate.

Bring the Friday page, not the whiteboard

A good Friday page means fewer guesses and fewer revision rounds, and it is the part only your team can write. The building is where outside help pays off. Studio Maydit designs and builds websites for AI founders in Framer, Webflow or custom code, usually as a fixed scope of three to four weeks, which is enough to put a sprint's direction live before it goes stale. If you are comparing studios for that step, our list of design agencies for post-raise rebrands sets out who does what. Or bring your refuse list to a call: book 45 minutes with us here.