Find the Leak

The instinct when growth stalls is to build something new. A new feature, a new onboarding flow, a new pricing page. New things are exciting, they're easy to justify, and they feel like progress.

They're also the slowest way to grow.

A new feature has to be discovered, understood, adopted, and retained before it moves a number, and most never clear the first hurdle. A leak, by contrast, is a place where users who already wanted your product gave up before they got it. Fixing it doesn't require anyone to change their mind. It just stops losing the people you already convinced.

That's the difference in leverage. Adding fights for new intent. Plugging a leak collects on intent you already paid for: the ad spend, the SEO, the word of mouth that got someone to the door. The traffic is there. The want is there. You're just closing the gap between what they came to do and what they managed to do.

The best growth teams internalize this early: you almost always have more to gain from the users you're losing than from the users you don't have yet.

Picture your product as a funnel, the five stages nearly every business shares:

• Acquisition: someone arrives.
• Activation: they reach the first moment of real value.
• Retention: they come back.
• Revenue: they pay.
• Referral: they bring someone else.

Every stage has a hole in it. The question is never whether you're leaking, it's where the biggest hole is, because that's the one worth your only experiment this week.

Here's the trap: the biggest hole is rarely where you're looking. Founders obsess over acquisition because it's visible, measurable, and everyone else talks about it. But if 60% of new signups never reach activation, doubling your traffic just doubles the number of people you disappoint. You've made the leak bigger, not smaller.

Walk the whole stack before you touch anything. Nine times out of ten the largest recoverable loss is sitting in activation or early retention, the unglamorous middle, not at the top where it's easy to see.

You don't need new tracking to find a leak. You need to read what you already have, correctly. Four habits separate a real read from a vanity one:

Follow cohorts, not averages. An average blends your best week with your worst and hides both. Group users by the week they signed up and watch each group move through the funnel. A leak shows up as a cliff, a step where a cohort's line falls off, long before it shows up in a blended number.

Measure the drop between steps, not the size of each step. 10,000 people hit your pricing page and 400 check out. The number that matters isn't 400, it's the 96% who didn't. Leaks live in the transitions, so instrument the transitions.

Find the empty-state moment. The most common activation leak is the first screen a new user sees with nothing in it. They signed up, landed on a blank dashboard, felt the cold, and left. 3,214 new users opened an empty dashboard last week and never came back is not a rare story, it's the default one.

Read the friction signals. Rage clicks, dead clicks, rapid back-navigation, and form abandonment are your users telling you where it hurts, in a language most teams never listen to. A spike in rage clicks on the pricing page is a leak with a neon sign over it.

Spotting a leak is easy. Every product has a dozen. The skill is sizing them, because you can only fix one at a time and you want the one that's costing you the most.

A leak's size isn't how annoying it looks. It's how much recoverable value flows through it:

Leak size ≈ traffic through the step × the drop-off × the value of a conversion

A checkout bug that loses 5% of a 200-person-per-week flow is a rounding error. A confusing empty state that loses 40% of 3,000 new users a week, each worth a shot at a subscription, is the whole quarter. Same product, wildly different priorities, and you'd never know which was which by staring at the screens.

This is also how you avoid the classic mistake of fixing the leak that bothers you instead of the one bleeding the business. Rank by recoverable value and the list reorders itself, often dramatically. The ugly, boring, mid-funnel drop-off usually wins.

A few ways teams waste this step:

• Optimizing the top when the leak is in the middle. More traffic into a leaky activation flow is more waste, faster.
• Trusting averages. A healthy blended retention number can hide a brutal new-user cliff. Always look by cohort.
• Chasing vanity metrics. Pageviews, signups, and total users feel good and move for reasons that have nothing to do with value. Anchor on the step where users reach the thing they came for.
• Instrumenting instead of reading. Teams spend six weeks building a perfect tracking plan to answer a question their existing data already answers. Read what you have first; add instrumentation only when a real decision depends on a number you genuinely don't have.