The Startup Multiverse: Why Five Years Reveals the Winning Branch
GitHub Co-Founder Tom Preston-Werner on the 5-year rule, pivots, and finding the winning branch in the startup multiverse.
Posted by David Thyresson • 5 min read
🗓️
In the five years I’ve worked with Tom Preston-Werner, I’ve learned many things. One of them is that he loves time-travel stories. The kind with branching timelines, alternate realities, and the tension of choosing which path becomes the “real” future.
It turns out you can see startups the same way. Each idea creates a new branch in the founder’s multiverse. Some timelines fade. Some reinvent themselves. And sometimes—just sometimes—one emerges as the winning branch.
And for Tom, the point where you know which branch survives is five years in.
Give something five years, if you can. If you have an idea and think, this is going to change the world, within five years you’ll know.
Why Tom Shared This Story
Tom didn’t share this idea as a neat theory. He framed it as a teaching moment for the founder community around one of the hardest choices in entrepreneurship: when to pivot or quit.
We thought it would be interesting to tell a story of how do you decide when to quit or pivot or change your mind about something? Because in the startup game, we like to think that we’ll succeed. And how do you do that? The general advice is, well, you just have to not quit. And then you’ll succeed. Is that true? For me, that definitely hasn’t been true.
The Chatterbug Example
His case study was Chatterbug, a language-learning platform he co-founded after GitHub:
- Great product – It worked, taking learners from zero to C1/C2 fluency.
- Clear market need – It addressed gaps left by Duolingo and others.
- Strong team – GitHub co-founders with proven track record.
- Adequate time – They gave it five full years of effort.
- Still failed commercially – The economics of user acquisition and scale didn’t work.
The sobering lesson: sometimes even with all the “right” ingredients, a branch of the multiverse fails.
The Five-Year Filter
That’s why Tom emphasizes the five-year mark. It’s the checkpoint in the multiverse where you step back and ask: has this branch stabilized, or is it collapsing?
You don’t want to spend your whole life doing the one thing that doesn’t work. That’s not a good algorithm for startup success.
The first two years are too soon—branches are still forming. Years three and four start to show patterns. By year five, the verdict is clear.
Infinite Branches, Finite Outcomes
Tom’s framework comes from what he calls the Infinite Ideas Theory:
There’s no sense in banging your head against the wall for twenty years on an idea that won’t take off. There are infinite ideas, and plenty of them are good enough to actually work.
Even within a single startup, infinite micro-futures unfold—product pivots, experiments, features, marketing strategies. Most are disposable branches. That’s not failure—it’s natural pruning.
The work of the founder isn’t to keep every branch alive. It’s to recognize the one that can grow into the canonical storyline.
Field Note: Famous Winning Branches
Some of the most iconic startups only became what they are today by pruning dead branches and finding the one that stuck:
- Slack – Began as an online game studio (Tiny Speck) before pivoting into an internal chat tool. The game faded; the chat branch won.
- Twitter – Originally a podcasting platform (Odeo) until Apple entered the space. The team pivoted to a micro-messaging app. That branch became Twitter.
- Instagram – Started as Burbn, a cluttered location-based app. The founders noticed people loved just one feature—photo sharing. That branch became the entire story.
- YouTube – Launched as a video dating site (“Tune In, Hook Up”). The dating angle fizzled, but general video sharing exploded.
- Shopify – Born out of a snowboard e-commerce store. The product wasn’t the boards—it was the underlying store platform. That branch grew into Shopify.
Lesson: The startup multiverse of pivots isn’t wasted effort. It’s exploration. Each branch teaches something. But only one becomes the canonical timeline.
Sanity in the Multiverse
The hardest part is letting go. In time-travel stories, the danger isn’t branching—it’s clinging too long to timelines that are already collapsing.
As Tom puts it:
For your own mental sanity, there comes a time to move on. It’s not within the first year, or even the second. But if by the third, fourth, and especially the fifth year things aren’t happening, that’s the signal.
Persistence is important. But endless persistence can be toxic. Five years is long enough to know if the branch is alive—or if it’s time to step into a different future.
Closing Loop
Tom’s Chatterbug story isn’t about failure. It’s about wisdom. It’s about giving founders permission to pivot, reinvent, or quit—without shame—because in the startup multiverse, not every branch is meant to survive.
The beauty of this perspective is that you don’t need every branch to win. You just need one.
By year five, you’ll know which storyline is becoming the real plot—the version of your startup worth telling. And if none has stabilized? Close the book, and begin another.
As Tom reminds us:
There’s no sense in killing yourself to make an idea work that isn’t timed right, or that you’re not the right person to execute.
Because in his favorite time-travel stories—as in startups—only one timeline becomes canon.
The Permission to Quit
Founders are often told “never give up.” But Tom’s five-year rule reframes the story: quitting or pivoting isn’t weakness, it’s strategy.
Giving up on a failing branch isn’t the end of the journey. It’s how you find the one timeline that works.
Persistence matters. But clarity matters more.