What Is a Pivot? Types of Startup Pivots With Examples

"Pivot" is one of the most overused words in startup life. It gets applied to a new logo, a layoff, a panicked change of market, and a carefully reasoned shift in strategy. Those are very different things. Only one of them is a pivot in the sense the word was meant to carry.

In the Lean Startup tradition, a pivot is a structured course correction. You've run your experiments, the evidence says the current approach isn't working, and you change one fundamental part of the plan while keeping what you learned. This article defines the term precisely, separates it from its look-alikes, walks through the main kinds of pivot with real examples, and covers the part most guides skip: how to decide, and how to tell people.

The definition, in Eric Ries's words

The official Lean Startup principles page puts the trigger plainly. When measurement shows the company isn't moving its business model forward, it says, it is time to pivot or make a structural course correction "to test a new fundamental hypothesis about the product, strategy and engine of growth."

Three ideas are packed into that sentence.

• Structural. A pivot changes something fundamental, not a button color or a price by ten percent.
• Hypothesis. The new direction is still a guess. A pivot doesn't promise success. It starts a new round of testing.
• Grounded in learning. In his 2009 post Pivot, don't jump to a new vision, Ries writes that successful startups "change directions but stay grounded in what they've learned," keeping one foot in the past and placing one in a new possible future.

That last point is what separates a pivot from a restart. If you throw away everything you know about the customer, the technology and the market, you haven't pivoted. You've started a different company.

For the wider method this idea lives inside, see the Lean Startup method, which covers the build-measure-learn loop and gives the short version of the pivot-or-persevere decision. This page goes deeper on the pivot itself.

Pivot, persevere, iterate, or quit

Founders often argue about whether to "pivot" when they really mean one of four different moves. Naming them helps.

This table is our own framing, not a quotation. The useful boundary is between iterating and pivoting. If you can fix the problem without changing the answer to "who is this for and what problem does it solve," it's an iteration. If the fix requires a different answer to that question, it's closer to a pivot.

Quitting also deserves a straight word. Pivoting is sometimes treated as a polite way of not admitting failure. It isn't. A pivot needs a new hypothesis you believe in, and enough time or money to test it. If you have neither, the honest call may be to stop.

The types of pivot

Pivots can be sorted by which part of the business you change. The sorting below is a descriptive taxonomy in our own words, built around the questions every startup must answer: who is the customer, what problem are we solving, what do we actually sell, and how do we reach and charge people.

Ries himself gave concrete names to some of these. In the 2009 post cited above, he describes three, each with an example from his own telling:

• Segment pivot. Take the existing product and use it to solve a similar problem for a different set of customers. His example is PBworks, which found unexpected demand from enterprise buyers.
• Customer problem pivot. Solve a different problem for the same customer. He points to Starbucks moving from selling beans and espresso machines toward brewing drinks in its own stores for the same coffee enthusiasts.
• Feature pivot. Select a single feature of the current product and reorient the whole company around it. His example is PayPal, where customers gravitated to email payments rather than the complex PDA-based security product.

Those three cover a lot of ground. A fuller working list, again our own, looks like this:

If you've seen a "ten pivots" list elsewhere and attributed to Ries's book, treat it as a useful classroom taxonomy rather than something we can quote from a page here. The Lean Startup site states the general idea of a structural course correction but doesn't print a list, and the table above shouldn't be read as his.

Real examples worth studying

Most famous pivot stories are retold so often that the details drift. Two examples stand on firmer ground because they come from the founders or from contemporary reporting.

Instagram: from Burbn to one thing done well

Instagram began as Burbn, a location-based app with check-ins, plans, points and photos. In a November 2010 TechCrunch piece, Kevin Systrom explained that they'd built a full iPhone version of Burbn, but it "felt cluttered, and overrun with features." So they cut almost everything except photos, comments and likes. What remained was Instagram.

Read that through the taxonomy above. It's a zoom-in pivot: one feature became the entire product. It also shows the "keep what you learned" rule at work. Instead of starting over with a blank page, the team narrowed what it had already built to a single job. The TechCrunch report adds that they chose this over competing head-on with check-in apps, and that their investors backed the change.

The Ries examples

Ries's own cases from the 2009 post are the cleanest teaching material, because he describes the logic rather than the legend. PBworks changed who it served. Starbucks, in his telling, changed the problem. PayPal changed the unit of the product. In each, the starting point was evidence about what customers actually did, not a fresh vision from nowhere.

Key Facts: Startup Pivots

• The Lean Startup principles page calls for a pivot, or structural course correction, when measurement shows the business model isn't advancing, to "test a new fundamental hypothesis" (source).
• Ries says successful startups change direction but stay grounded in what they've learned (source).
• His 2009 post names three pivot types: segment (PBworks), customer problem (Starbucks), and feature (PayPal) (source).
• Instagram's founders cut Burbn down to photos, comments and likes after finding it cluttered (TechCrunch, 2010).
• A pivot needs a new hypothesis you can test. Without one, the choice is persevering or stopping.

Signals that a pivot may be due

No single number tells you to pivot. What matters is a pattern: you've made real changes, tested honestly, and the core hypothesis still isn't holding. Common signals, framed as questions rather than thresholds:

• Flat learning. Several rounds of experiments, and the key behavior you care about doesn't move. The Lean Startup site frames this as the business model's drivers not advancing.
• Customers who like it but don't act. Compliments and polite interest, but no repeat use, no commitment, no payment. That's often a sign you've solved a problem people agree exists but won't pay to fix.
• A small group that behaves very differently. If a niche you didn't plan for uses the product heavily, that is a segment or zoom-in pivot waiting to be tested. Ries's PBworks example fits this pattern.
• Unit economics that don't close. Where acquiring a customer costs more than they'll ever return, and no plausible tweak fixes it, the growth or revenue model may be the thing to change. See what a startup's traction should show for how that connects to fit.
• A feature that outshines the product. If one piece gets all the use and the rest gathers dust, Instagram's story is the obvious precedent.

Be careful with the opposite error. A weak result from a thin MVP can come from poor execution rather than a bad idea, which would send you into the wrong pivot. The minimum viable product article covers how thin is too thin. And a useful early test of fit is a direct question to users, which the Sean Ellis test formalises.

What to keep when you pivot

A pivot that discards everything is a restart. Before you change direction, list what carries over.

1. What you learned about customers. Who responded, who didn't, and what language they used. Even a failed segment teaches you what the next one needs. The customer discovery interviews you already ran remain useful.
2. Your technology and assets. Code, data, integrations, brand and relationships often transfer. Instagram, for instance, cut features rather than the work behind the ones it kept.
3. The team's skills and trust. People who've worked through a hard stretch together are an asset. Don't spend that casually.
4. The evidence on what failed. Write down which hypothesis you tested, what you did, and what happened. It stops you repeating the experiment under a new name six months later.
5. Your relationships. Early customers, advisors and investors often have useful opinions about the new direction, and some will become the first users of it.

How to run the decision

Treat the pivot as a hypothesis, just like the product.

1. Write the old hypothesis and the evidence against it. Be specific about what you expected and what happened.
2. Define the new hypothesis in one sentence. For example: "Independent clinics, not hospital groups, will pay for appointment reminders if setup takes under a day." Your riskiest assumption in the new direction is the first thing to test.
3. Decide the test and the pass mark in advance. Setting criteria before you see results makes the next pivot-or-persevere call less emotional.
4. Set a time and money limit. A pivot without a deadline can become a drift.
5. Run the test, then decide again. Some pivots work. Many need a second adjustment. Some end in a shutdown. All of those outcomes are legitimate.

Before you act, check whether you've actually exhausted the cheaper options, such as a change in pricing, a different message, or a better onboarding flow. Those are iterations, and they can resolve what looked like a pivot-sized problem.

How to communicate a pivot

This section is general guidance rather than research findings, but it follows from what a pivot is.

To the team. People need the reason before they need the plan. Show the evidence, name the hypothesis that failed, and say plainly what stays the same. Be honest about uncertainty: the new direction is a bet, not a promise. Expect questions about roles and job security, and answer them directly. Vague reassurance does more damage than an uncomfortable fact.

To investors. Lead with the learning. A good update says what you believed, what you tested, what you found, what you're changing, and what you'll measure next. Investors have seen many pivots. What they're looking for is evidence that the change is reasoned and that you still have a path to testing it. Bring the new hypothesis, the pass mark, and the runway you have to test it. Giving investors a heads-up before the all-hands is usually better than letting them hear it secondhand.

To customers. If existing customers are affected, tell them early, explain what changes for them, and honor commitments you've made. If you're abandoning a segment, a clear exit path protects your reputation.

Keep the story consistent across all three audiences. Different emphasis is fine. Different facts are not.

Pivots in bigger companies

Established companies change strategy too, but the stakes differ: more people, more legacy commitments, and slower feedback. If you're leading a larger organisation, the article on strategic pivot timing covers how leaders decide when to move. This page is about the earliest stage, where the company is still searching for a business model that works.

Related reading

• The Lean Startup method
• Product-market fit explained
• The Sean Ellis test
• Riskiest assumption test
• Growth vs profitability
• Strategic pivot timing for leaders