Most early-stage B2B startups acquire their first customers like boats trawling for fish. They drop a huge net into the ocean and try to scoop up everything in their wake, taking on any customer who is willing to buy their product. Trawling feels validating because it can add customers and revenue quickly. But founders frequently realize later on that the resulting “catch” of customers is too diverse to serve effectively with a small team and an immature product. In the best case, customers are only modestly happy with the solution. In the worst case, they churn, leaving founders confused about their product-market fit and a product that’s too general to serve anyone well.
Initial customer definition is one of the hardest things founders have to do. It takes a year or two of learning and experimentation to gain the conviction necessary to define your customer with precision, no matter how much prior experience you may have. It can be emotionally challenging too, because it forces otherwise hungry founders to throw fish back into the ocean and decline to serve customers who could use their product, but aren’t the right ones at the start.
PayPal is an excellent example of a different, and we believe better, approach. As Peter Thiel describes in Zero to One, PayPal’s goal was to invent “internet money.” It’s hard to imagine a broader product with more potential customers and use cases. Yet in its early days, PayPal constrained its initial customer profile to power sellers on eBay, who comprised fewer than 2,000 of the approximately 250 million internet users at the time (or 0.000008% of their eventual market). They could have chosen hundreds of larger customer segments, but they deliberately defined their customer in a seemingly irrationally narrow way. Why did they do this? Because the PayPal team understood that to go big, you have to start narrow and build up a solid cohort of deliriously happy customers. The best way to do that is to serve a narrow niche and then expand your ICP bit by bit. Otherwise, if you’re trying to serve everyone, you’re unlikely to delight anyone.
Irrational narrowness should be a founder’s aspiration early on, because it increases the odds that early customers will be insanely happy and engaged. Founders should prioritize insanely happy customers over “insane” but brittle month-on-month growth. The quantity of new customers is far less important than the quality of their engagement with your product. You want customers who are so deeply engaged that they are willing to jump on sales calls to tell your prospective customers how much they love your product because they want you to succeed.
Trawling is a critical aspect of customer discovery. As long as it is done with the goal of throwing most fish back in the water to find the perfect initial customer, trawling is useful. Once you've identified the right customer, you stop trawling and start spearfishing. You know exactly the kind of fish you are trying to catch and you have to get proficient at catching it again and again with lower cost and effort.
This suggests a reordering of go-to-market priorities for early-stage founders:
Trawl to learn, not to grow. Make your trawling hypothesis-driven, not aimless. Don’t over-optimize for revenue growth from any customer who is willing to pay. Unlike a fishing boat, you don’t get paid for tonnage.
Instead, optimize for acquiring customers whom you believe will be highly engaged. Pay close attention to how your early customers use your product and try to predict the behaviors that correlate with their retention and expansion.
When you see a lot of customer engagement, refine your ICP definition accordingly and search for lookalike customers. Over time, try to narrow your customer definition, even comically so, to increase the odds that all of your early customers will be deeply engaged and satisfied.
Narrowness has an added benefit: it simplifies other components of a go-to-market strategy. Every important go-to-market decision is downstream of customer selection. Once PayPal decided to serve power sellers on eBay exclusively, it became much easier to figure out how to find them, to convince them to try the product, to understand which features they most needed, to price the product, etc. Had the target customer simply been anyone using money on the internet, the roadmap of what to build and how to acquire and sell to users would have been much more difficult to develop.