Design Sprint for a Small Business: Five Days, One Test

The design sprint was built for teams that had a designer, an engineer and a conference room to spare, and the part worth stealing for a five-person business is the calendar, not the craft. Jake Knapp created the method at Google in 2010 and published it in 2016 as Sprint, written with John Zeratsky and Braden Kowitz (thesprintbook.com). GV, the venture arm of Alphabet, describes it as “a five-day process for answering critical business questions through design, prototyping, and testing ideas with customers,” with one job per day: map the problem on Monday, sketch solutions on Tuesday, decide on Wednesday, build a prototype on Thursday, and test it with real people on Friday (gv.com/sprint).

Read that list again as a business owner rather than a product manager. Every one of those five jobs is something you already do, badly, over four months: you talk about the idea, someone sketches it on a napkin, you argue about it, you half-build it, and the market tells you the answer after you have paid for it. The sprint does not add work. It puts the five steps you were going to take anyway into one week, in the right order, and moves the customer’s verdict from the end of the spend to the start of it. That reordering is the whole value, and it costs nothing but five people’s time.

Monday: write down the question, not the idea

GV’s Monday is a day of “structured discussions” that “create a path for the sprint week,” ending with a target: one customer, one moment in their experience, one question the week has to answer (gv.com/sprint). The discipline for a small business is to write the question as a sentence a customer could answer, not as a feature. “Should we add a membership?” is an idea. “Will a customer who comes every six weeks pay $45 a month to come every four?” is a question, and Friday can answer it.

If the room cannot agree on who the customer is, Monday is not wasted; it has found the real problem. The six lists behind the Value Proposition Canvas are the fastest way to settle it, because they force the owner to write the customer’s jobs and pains before anyone draws a product. A small shop can do GV’s full Monday in a morning. Keep the afternoon for the one task owners skip: recruiting Friday’s customers. GV starts recruiting on Tuesday, but when you are also the person who answers the phone, start a day early.

Tuesday: sketch alone, then compare

GV’s Tuesday is for sketching “competing solutions on paper,” and the word that matters is competing (gv.com/sprint). Each person sketches their own version of how the customer would meet the idea, in silence, before anyone sees anyone else’s. In a company of five, that rule is worth more than in a company of five hundred, because in a small company the loudest sketch is usually the owner’s, and the owner is the one person who cannot see the business as a customer does.

The sketch is not art. For a membership it is three pieces of paper: the sign in the window or the line on the website that announces it, the page or form where a customer signs up, and the text message they get on the first of the month. For a new service it is the quote, the confirmation and the invoice. If your front-desk person sketches a sign-up form with two fields and you sketched one with nine, Tuesday has already paid for itself.

Wednesday: one decision, written as a storyboard

By Wednesday morning, GV says, “you and your team will have a stack of solutions,” and since “you can’t prototype and test them all,” the morning is spent choosing and the afternoon turning the winning sketches into a storyboard, “a step-by-step plan for your prototype” (gv.com/sprint). Our advice is that the decision belongs to one person, not a vote. In a small business that person is almost always the owner, which is fine as long as the owner decides after the sketches rather than before them.

If the choice feels paralyzing, remember what kind of decision it is. A prototype that will be shown to five customers on Friday is about as reversible as a business decision gets, and the decision-making framework that sorts choices by whether you can walk them back says to give reversible decisions a date rather than more thought. Wednesday is the date. The storyboard for a small shop is six to ten boxes drawn on one sheet: the customer hears about the offer, reads it, signs up, pays, gets the first reminder, uses it. Each box is one screen or one piece of paper to fake on Thursday.

Thursday: fake it, on purpose

GV’s Thursday instruction is to “adopt a ‘fake it’ philosophy to turn that storyboard into a prototype” (gv.com/sprint). This is the day owners most often get wrong, because the instinct is to build the real thing a little early. Do not. The prototype only has to be real enough for a customer to react to it for twenty minutes on Friday. A membership prototype is a one-page website, a sign-up form that saves to a spreadsheet, and a text message you send by hand. Nobody is charged. Nothing is wired into your booking system. If Friday goes well, you build it properly the following week, and if Friday goes badly you have lost a Thursday.

The tools for faking it are cheap, and the prices below were read on each vendor’s pricing page on November 9, 2026.

Canva’s Pro plan is listed at US$144 a year for one person, with the Free plan at no charge (canva.com/pricing). Figma’s Starter plan is free, and a Professional full seat is $20 a month billed monthly or $16 a month on an annual plan (figma.com/pricing). Carrd’s Pro tier starts at $19 a year (carrd.co). Maze’s Free plan allows one study a month and five seats, with prototype testing and surveys included (maze.co/pricing). For most small businesses the free tiers cover the whole sprint, and a paper prototype shown across a counter costs nothing at all. Pay for Figma only if the idea is genuinely an app or an online booking flow that customers have to click through.

Friday: five customers, one at a time

GV’s Friday is to “test it with real live humans” (gv.com/sprint), and the number five is not arbitrary. Jakob Nielsen’s 2000 article on usability testing found that a first study with five participants uncovers about 85 percent of a design’s usability problems, and that three studies of five users each improve a product more than one study of fifteen (Nielsen Norman Group). Nielsen was writing about software interfaces, and his article is explicit that fifteen users are needed to find every problem; the small-business reading is that five real customers reacting to a prototype will show you the big mistakes, and a second five after you fix them will show you most of the rest.

Run the five as conversations, not surveys. One person asks, one person takes notes, and the customer gets the prototype with as little explanation as you would give a stranger walking in off the street. The questions are the ones you wrote on Monday, and the answer you are listening for is not “that’s nice” but “how much?” and “when can I start?” A customer who asks to sign up on Friday afternoon, when there is nothing to sign up for, is the signal. A customer who says it is a good idea and changes the subject is the other signal. Pay for their time; a gift card is normal, and the five conversations are the cheapest market research you will buy this year.

Then decide, that afternoon, in writing, which of three things happens next: build it, change it and run a second Friday, or drop it. The 30-day product launch checklist starts where a successful Friday ends, and the one-question product-market fit survey is the check to run ninety days after launch, once real customers have real experience to report.

A hypothetical week at a three-chair barbershop

Consider a made-up barbershop with three chairs and four staff, used only to show the shape of the week; it is not a case study. The owner has talked about a monthly membership for a year. Monday, the four of them agree the customer is the man who comes every five or six weeks and the question is whether he would pay $45 a month for unlimited cuts. Tuesday, each of them sketches the window sign, the sign-up card and the monthly text; the newest barber’s sign-up card has three lines and everyone else’s has eight. Wednesday, the owner picks the three-line card and storyboards six boxes. Thursday, one barber builds a one-page site on Carrd, another designs the window sign in Canva, and the owner writes the welcome text in his own voice. Friday, five regulars who were booked on Monday see the sign, read the page, and are asked what they would do. Two ask to sign up now. One says $45 is more than he spends. Two ask whether the membership includes a beard trim, which nobody on Tuesday had thought of. The owner now knows the price is near the line, that the beard trim is the feature, and that two of five regulars are ready, and he has spent five mornings and about $20 to learn it.

Where AI fits in the five days

An AI assistant earns its place on three of the five days and should stay out of the other two. On Monday it can turn an hour of messy discussion, typed as notes, into a clean list of candidate questions for the team to choose from. On Thursday it can draft the website copy, the welcome text and the sign-up confirmation from the storyboard in minutes, in the owner’s voice if you paste in a few of your own emails first, and the owner edits before anything is shown to a customer. On Friday evening it can sort the five sets of notes into what customers asked, what they objected to and what they ignored, which is the part nobody wants to do at six o’clock. Claude Pro is $20 a month billed monthly or $17 a month with an annual subscription (claude.com/pricing), and Google AI Pro is $19.99 a month (one.google.com); both have free tiers that will handle a single sprint.

Keep it out of Tuesday and out of the Friday conversations. Tuesday’s sketches are valuable because four different people drew them from four different seats in the business, and an assistant asked to sketch a membership will produce the average of every membership on the internet, which is exactly what you do not want. And Friday is five customers talking to a person they know; a recorded script read off a screen would cost you the one thing the week was for.

BusinessPrompter, a sister venture of this site, publishes an Innovation Sprint Facilitator prompt in its Innovation and Growth category, available to Pro members as a single prompt and as a five-step chain, with the summary line “Run innovation sprints to rapidly prototype and test ideas.” Its About panel describes the audience as product development teams and startup founders, so an owner using it for a barbershop membership should expect to translate some of the vocabulary; the five-step shape maps onto the five days above. The BusinessPrompter.com catalog has free prompts alongside the Pro ones.

Our view: the sprint is a hiring decision in disguise

Most advice about testing ideas points a small business toward a consultant, an agency or a freelancer who will come in, run a workshop and leave a deck. Our view is that the design sprint’s real argument is the opposite one. The four people who already work in your business know more about your customers than anyone you could hire for a week, and the method’s only job is to get their knowledge onto paper, into one decision and in front of five customers before the owner spends real money. A business that runs two or three of these a year does not need more people to have more ideas tested. It needs its existing people to spend five mornings a quarter doing it on purpose.

So pick the idea that has been in the conversation longest, and put five customer names on next Friday’s calendar before you do anything else. Once Friday is booked, Monday to Thursday tend to take care of themselves.

Frequently asked questions

How long does a design sprint take for a small business?

The method as GV describes it runs five working days, one job per day: map, sketch, decide, prototype, test (gv.com/sprint). A small business can usually do each day’s work in a morning if the question is narrow and the prototype is paper or a one-page website, so the realistic cost is five mornings, not a full week off the floor.

How many customers do you need to test a prototype?

Five, tested one at a time. Jakob Nielsen’s research found that a first round of five participants uncovers about 85 percent of a design’s usability problems, and that several small rounds beat one large one (Nielsen Norman Group). Recruit real customers who match the person the idea is for, and book them on Monday so Friday is not spent chasing people.

Do you need a designer to run a design sprint?

No. The Thursday prototype only has to be real enough for a customer to react to for twenty minutes. A paper sketch, a Canva mock-up on the free plan or a one-page site on Carrd Pro at $19 a year is enough for most service and retail ideas (carrd.co). Pay for a design tool such as Figma only when the idea is an app or a booking flow customers must click through.