Bootstrapped startup lessons from Finder's Jeremy Cabral

Most founders treat a bootstrapped startup as the fallback plan, the thing you do until someone writes a cheque. Finder is a useful counterexample. It grew out of a set of experiments, one of which was Credit Card Finder, ran on a $2,000 a month budget, and didn't take outside capital until December 2021. Along the way it became a global business with more than $100 million in revenue.

Jeremy Cabral, one of Finder's three co-founders, joined in early 2009. On Now or Never he walked us through the playbooks behind that growth: SEO, earned media, empathy and systems. Here's what founders bootstrapping a startup today can take from it.

Start with a simple rule: invest a dollar, make two

Finder's founders came from performance marketing, and their operating rule was as basic as it gets: invest a dollar and make two.

"We had a $2,000 a month budget and we would just spend every single dollar very carefully"

That budget grew to $4,000, then $10,000 a month, and eventually millions. The point isn't the size of the starting number. It's that every dollar had to prove itself before the next one went in, which forces you to understand exactly which channels return money.

The path wasn't smooth. When Jeremy joined, the site had been penalised by Google (he fixed it within a couple of weeks), and penalties kept recurring. At one point the whole site lost its traffic for three months. That was when the team drew a line in the sand: no more gaming the algorithm. From then on they would do it the hard way, but in a way that could scale.

How to bootstrap a startup with earned media

If you can't buy attention, you have to earn it. Earned media became one of the key pillars of Finder's growth, and Jeremy says they never really stopped.

Capture a moment

Around the iPhone 5 launch, Finder was about to release mobile plan comparison. Jeremy walked past the Apple store three days out, noticed there was no queue yet, and started one himself. The thinking behind it is the useful bit:

"if I was a journalist, this kind of thing's happening, like what's the story that's gonna be written, how do we be the story?"

Finder bought the first iPhone 5, got the photo, and was referenced by more than 300 publications globally. The site went from around 1,100 visits a month to 124,000 visits in 72 hours. Later, the team went further and started manufacturing moments of their own, like the viral competition built around Raygun's Olympic breakdance.

Become the source journalists rely on

Stunts are one-offs. The repeatable version was research. Jeremy noticed that news is full of statistics, and newspapers need a steady supply of new things to report.

"to have news, you have to do something new. So therefore, the variable is manufacturing something new repeatedly."

Finder built a survey function and ran research that raised the importance of the problem it solved: how much people could save if they compared. Then Jeremy offered journalists exclusive data for stories they were already planning. Over time they started calling him, sometimes at 6pm on a Friday or 11pm on a Friday for a 5am studio slot. He'd stop anything to take a journalist's call, because responsiveness was the variable.

Empathy is a growth strategy, not a soft skill

Every one of these plays starts from the same habit. Jeremy runs a simulation of how the other person will see what he's doing.

"As a Google engineer, what would I think? As a journalist, what would I think?"

His earliest example is from 2009. Hoping for a link from Lifehacker, he sent the editor a carton of beer with custom labels, timed by courier to land on a Friday afternoon, pointing to a page announcing Credit Card Finder's launch. The editor declined to run the article Jeremy pitched but agreed to a coffee, and on the way to that coffee the launch story went live anyway. Five or six years later, Jeremy hired him as Finder's editor in chief.

The principle is a simple overlap: what does the other person need, and what do you need? Find the win for both and things move. For journalists, that means understanding they're time poor, short on resources and need accurate stories every week. Make their job easier and they'll come back.

Systems: find the variable and outperform on it

Jeremy's view of systems is refreshingly unglamorous. Most founders have a good idea. The hard part is traction, and the thing that often gets in the way is overthinking the launch. He'd rather launch incrementally, learn from the data and iterate than wait for perfect.

From there, every growth channel is a set of variables. For SEO it's content, links and technical foundations. For sales it's outreach per day, booked calls and close rate. Everyone has roughly the same flywheel, so the edge comes from going further than anyone else on one input.

"you need to outperform in one of the variables to take market share."

Links are his example. Buying them is the easy shortcut, and Google says you shouldn't. The best links come from publications you can't pay, like the AFR, which is exactly why earning them is worth the effort. Pick an input where a competitor would never go to your extremes, and the results follow.

The same logic shaped how he thinks about agencies. In his experience, the typical SEO agency model puts the most junior person on your account to protect margin. He prefers augmenting an in-house team with specialists, and being very clear about what you're trying to do.

Set the ambition, then obsess over inputs

When Finder was making $20,000 to $30,000 a month, the team set a goal of $100,000 a month, which felt absurd. Later it was a million visitors a month. Jeremy bought an inscribed samurai sword to mark the commitment, and a few years later they used it to cut the cake celebrating a million visits. Goals like that are only partly logical, he admits:

"numbers can get you most of the way, maybe 70%, but then you really gotta back yourself and have good intuition."

Does the bootstrapped startup playbook still work with AI?

Jeremy thinks it does, and possibly more so. AI answers still rely on citations and trusted sources, and journalism, with its editorial standards, sits near the top of that trust pile. Earned coverage feeds both search and AI.

He also argues the blueprint applies well beyond comparison sites. Every startup needs proof: design partners who validate the product, case studies, customer logos, ratings and "as seen on" badges. His laddering technique still works too. Coverage in Lifehacker and Gizmodo gave him something to mention to an SMH tech journalist at the same events, and from there it snowballs. Sometimes growth is just being in the room with the person who holds the keys.

We'd add one thing from the UntilNow side. AI makes it cheap to build a product that looks like everyone else's, so the trust signals Jeremy describes, and the brand that carries them, become the real moat. A clear brand strategy gives journalists, partners and customers a story worth repeating. Our guides to startup branding on a limited budget and the minimum viable brand are a good place to start if you're bootstrapping.

The Finder playbook in five lines

Finder's story shows that a bootstrapped startup can compete with far better funded rivals if it's smarter about where each dollar goes. For the full story, including Jeremy's thinking on co-founders, values and life after Finder, listen to the episode with Jeremy Cabral on Now or Never.