Finding leaks where volume drops fastest
A funnel leak is any step where users drop out faster than expected, and the biggest leak is rarely where teams intuitively look. Awareness-to-consideration might leak 95%, landing-page-to-form 80%, and form-to-purchase only 40%. The 95% awareness drop looks catastrophic until you realize it's expected (most people who see an ad never intend to buy). The 40% form-to-purchase drop, though smaller in percentage, often represents a fixable barrier: a confusing checkout flow, unexpected shipping cost reveal, or restrictive return policy. The leak that matters most is the one where conversion is plausible but friction is unnecessary.
Diagnostic vs strategic leaks
Diagnostic leaks are easy to fix: a broken link, a missing form field, a slow page that times out. These show up as sudden drops at a single step. Strategic leaks are harder: users reach your form but your offer simply doesn't match their search intent, or your price is perceived as high relative to competitors. Diagnostic leaks affect a small, reversible percentage of users. Strategic leaks affect large cohorts consistently and often require repositioning or pricing changes. Isolating which type of leak you're facing determines whether the fix is engineering hours or broader strategy work.
Measuring leak severity across cohorts
A leak affecting everyone equally is different from one affecting only paid-search users or only first-time visitors. Segmenting your funnel by traffic source, device, geography, and user cohort reveals that the biggest leak for one segment might be a non-issue for another. Email users might leak hard at the consideration step because your nurture content is stale. Mobile users might leak at checkout because your form isn't mobile-optimized. Breaking the funnel into cohorts transforms a single global number into actionable problems, each with a clear owner.