Founders
Startup pivot: how to know it is time to change course
Pivot a startup when the evidence says the current business will fall short of the size you need, and when your best remaining fixes feel uninspired. Y Combinator’s Dalton Caldwell puts the threshold at months of full-time effort without results, not a few weeks. A pivot changes a core part of the business while keeping what you have learned.
What a pivot is
Eric Ries, who popularized the term in the lean startup method, defines a pivot as a structural course correction made to test a new fundamental hypothesis about the product, the strategy or the growth engine. Rewriting the landing page or adding a feature is iteration. Switching who you sell to, what problem you solve or how you make money is a pivot.
Most pivots keep something. The team, the technology, a feature, a customer relationship or a hard-won insight carries over. That is what separates a pivot from shutting down and starting a new company, and it is why the decision is worth making deliberately instead of drifting into it.
Signs it may be time to pivot
No single metric settles the question, but the same warning signs show up across the advice from experienced founders and investors.
- Lukewarm interest that does not improve. Lenny Rachitsky’s analysis of pivots lists very low retention, a growth plateau and a shortage of strong ideas for driving growth as the clearest pattern. He suggests two to three months of this in the idea stage, or six months or more with a live product, as a point to take it seriously.
- A market that is smaller than you hoped. If everyone agrees the problem is real but the total revenue you could ever collect is small, more effort will not change the arithmetic.
- The right people say the problem is missing. PostHog’s guide to pivoting advises talking to people who actually have the problem rather than friends and family, and pressing for direct answers when replies are vague or overly polite.
- Early users barely use the product and you cannot learn why. PostHog lists this as a signal, together with an inability to answer yes to three questions: are you excited about it, is the work interesting, and are customers desperate for it to succeed?
- You privately suspect it will fail. Caldwell’s Startup School talk lists launching without traction for weeks or months, a plan that cannot start without huge funding, and a private belief that it is not working as the strongest triggers.
If your product is a subscription business, retention and churn numbers are the first place to look. The guide on activation, retention and churn in SaaS explains how to read them.
Bad reasons to pivot
Caldwell is just as direct about the wrong motives. Dodging sales work is a bad reason. Changing direction chronically is another, and so is chasing a hot trend from press coverage. PostHog adds a pitfall that is easy to miss: a pivot that is really your last plan with minor changes will produce the same result.
He also notes that early in a company’s life, switching concepts should feel lightweight and may even be normal, but that pivoting so often you never launch defeats the purpose.
Why founders pivot too late
According to the same Y Combinator talk, founders wait too long for a handful of human reasons: loss aversion after investing time and money, a little traction that Caldwell calls the uncanny valley of product-market fit, mistaking customer politeness for real interest, fear of looking like they gave up, and blaming customers or investors instead of examining the idea.
Every month of runway spent on a business that is stalling is a month you cannot spend on a better strategy. Tracking burn rate and runway turns a vague feeling that things are slow into a date by which a decision has to be made.
Product problem or market problem
Before choosing a new direction, work out which half of the business is broken. PostHog’s newsletter describes Tom Blomfield’s early startup Groupay, where the team stopped building features and went looking for users. Doing so showed that the market, not the product, was the issue. Its rule of thumb: if users are hard to engage or monetize even though the core technology works, suspect the market.
The two diagnoses lead to different changes. A product problem calls for a different solution to the same customer problem. A market problem calls for a different customer or a different problem. Good customer discovery interviews are the fastest way to tell them apart.
Types of pivots
There is no official list. Ries described a small set in a 2009 post, and later writers have extended it. The table below follows the version summarized by Rework, which says its fuller taxonomy is its own framing and not Ries’s.
| Type | What changes |
|---|---|
| Customer segment | Same product, a different group of buyers |
| Customer problem | Same customers, a different problem solved for them |
| Zoom-in (feature) | A single feature becomes the whole product |
| Zoom-out | The whole product becomes a feature of something larger |
| Platform | An application turns into a platform, or the reverse |
| Revenue model | You change how customers pay |
| Growth model | You change how you acquire customers |
| Channel | You change the route to market |
| Technology | You solve the same problem with a different technical approach |
Rework gives PBworks as an example of a segment pivot, where the company found demand from enterprise buyers it had not been targeting, and PayPal as a feature pivot, where users preferred email payments over the complex security product the company started with.
Rachitsky sorts pivots by timing instead. An ideation pivot comes early, before a real product or traction, and is often unrelated to the original idea. A hard pivot comes when a company has a live product and real users, and it keeps a single element and focuses on it. By his count, ideation pivots usually happen within about three months of launch and hard pivots within two years, most of them near the one-year mark.
Business model pivot or product pivot
This kind of change leaves the product largely alone and changes how the company earns money or reaches buyers: the price, the billing method, the sales channel or the type of customer who pays. A product pivot changes what you build. The distinction matters because the first is often cheaper and faster to test. If people use the product but will not pay for it in the current form, look at the model before you rebuild the product. The article on SaaS pricing, billing and freemium covers the usual model choices for software startups.
Either kind of change is really a search for product-market fit, the point where a defined group of customers keeps using the product and pays for it without constant persuasion. Treat each pivot as a test of a new startup hypothesis about the business, run it for a fixed period, and decide to persevere or pivot again based on what customers do, not on what they say.
Two documented examples
Burbn to Instagram
Burbn was a location-based app. According to TechCrunch’s reporting from November 2010, the founders had spent about a year on it and had raised $500,000 before the change. After the iPhone version felt cluttered, they cut nearly everything. Kevin Systrom wrote that they cut everything in Burbn except photos, comments and likes. The company rebuilt it as Instagram in about eight weeks.
TechCrunch reported roughly 100,000 users in the first week, 200,000 a week later and 300,000 a week after that. That makes it a clean zoom-in pivot: the feature people used most became the product.
Glitch to Slack
Tiny Speck, led by Stewart Butterfield, released the browser game Glitch to the public in September 2011. TechCrunch’s 2019 account says the company had raised over $15 million before beta testing, including a $5 million Series A in 2010 and a $10.7 million Series B in April 2011. The game closed in November 2012, about 18 months after launch. The article cites reasons including that the game was too complicated and too dependent on Flash.
While building the game, the engineers had written a real-time messaging tool for their own remote team. When Glitch closed, the company said that tool had uses outside gaming. Butterfield told investors the company would become an enterprise software business. According to TechCrunch, Accel stayed committed while Andreessen Horowitz was reportedly skeptical. Slack launched publicly at the end of 2013, and Social Capital led a $42 million Series C in April 2014.
The two stories share a pattern: the winning product was already inside the old product.
How to run a pivot
- Write down the hypothesis that failed. Name what you believed about the customer, the problem and the way you would grow. A pivot is a new hypothesis, so you need to know which one you are replacing.
- Talk to people before building. PostHog suggests building a rough version in days or weeks rather than months and asking early contacts whether the problem is high priority, whether the solution is easy to use and what is missing.
- Score candidate directions. Caldwell recommends four criteria: market size, founder and market fit, how easy it is to get started, and early customer feedback from real buyers.
- Make the change large enough. PostHog quotes Reid Hoffman’s advice to slash and burn the old business instead of keeping a foot in both.
- Set a decision deadline. PostHog lists open-ended deliberation as a pitfall. Put a date on the choice and keep to it.
- Hold off on hiring. Caldwell advises against adding people while the new direction is still taking shape.
Telling your board and investors
Investors back a team and a plan, so a change of plan needs a reasoned explanation: what you tested, what the data showed, what you intend to test next and how much runway it gets. The Slack story shows that reactions differ even among investors who believe in the team. Practical advice on handling those conversations is in the guide to startup board meetings, and the effect on a later round is covered in Series A explained.
Questions founders ask
How many times can a startup pivot? There is no fixed number. PostHog says it made six pivots before finding product-market fit, and Caldwell warns against pivoting so often that nothing ever launches.
Should I pivot or persevere? Persevere while promising growth options remain and retention is improving. Caldwell’s rule of thumb is that uninspired remaining options are a sign to consider a pivot.
Is a pivot a failure? The two cases above suggest otherwise. Both companies kept their team and their learning.
Sources and limits
- TechCrunch, “The Slack origin story” (2019), for the Glitch and Slack timeline and funding.
- TechCrunch, Burbn and Instagram coverage (November 2010), for funding, feature cuts and early user counts.
- Y Combinator, Startup School week 6 recap on pivoting with Dalton Caldwell (2019).
- PostHog newsletter, “Read this before you pivot”.
- Lenny’s Newsletter, “The art of the pivot, part 2”.
- Rework, “What is a pivot” (taxonomy and examples).
- The Lean Startup website, principles page (definition of a pivot).
Pivot taxonomies differ between authors, and the counts above are those of the cited sources. Funding figures and dates come from press accounts, not company filings. Early stage timing guidance is a rule of thumb from a single analysis, not a rule.