The "missing traffic" trap: why most conversion fixes fail when the real issue is distribution

Consider a team that spends six weeks running A/B tests on their signup page. They tested headline copy, button colour, form length, social proof placement. They got statistically significant results — the winning variant converted at 4.2% versus 3.1% on the control. Textbook execution.

Signups didn’t meaningfully increase. The reason: 310 people visited the signup page that month. A 35% conversion rate improvement on 310 visitors is 4 extra signups.

This is the missing traffic trap. The funnel tool showed them exactly where people were dropping off. It didn’t show them that the funnel’s input was running on a trickle. They optimised the drain while the tank was nearly empty.

Conversion rate optimisation is legible. You instrument a page, run a test, get a number. It feels like science. The tools — Optimizely, VWO, Google Optimize (while it existed), PostHog’s experiment feature — are designed to produce clean before/after comparisons. The workflow is satisfying.

Distribution work is not legible. Building an SEO content programme takes 6-12 months to compound. Organic social has no reliable feedback loop for months. Partnership channels involve negotiation, relationship-building, and long lead times. None of this fits into a two-week sprint.

Product teams are measured on what’s measurable in a sprint cycle. Conversion rate is measurable in a sprint cycle. Traffic growth isn’t. So the team that should be working on distribution keeps running A/B tests, because the A/B tests produce closure.

The other factor: funnel analytics tools are almost universally scoped to sessions that entered the funnel. Mixpanel, Amplitude, and PostHog all default to showing you behaviour inside your product or website. They are excellent at answering “where do people drop off?” and structurally silent on “what fraction of our potential audience is even aware we exist?” That’s a question about distribution, and none of these tools answer it.

There’s a specific scale threshold below which conversion optimisation is almost always the wrong investment. The exact number depends on your market, but as a working heuristic: if fewer than 1,000 qualified visitors hit your key conversion page per month, distribution is the constraint, not conversion rate.

Here’s why the maths forces this conclusion. Say your signup page converts at 3% and gets 800 visitors/month. That’s 24 signups. A heroic CRO effort — sustained testing, designer time, copy rewrites — might get you to 5%. That’s 40 signups. A 67% improvement in conversion rate gives you 16 extra signups.

Now say you instead spend that same effort getting your traffic to 2,400 visitors/month (3x growth from a new channel or improved SEO). At the original 3% conversion rate, that’s 72 signups — 3x the outcome from the same effort investment.

Conversion rate improvements are bounded at 100% (you cannot convert more than everyone). Traffic improvements have no ceiling. At low absolute volumes, traffic wins every time.

The calculus shifts above ~10,000 monthly visitors. At that scale, a 1% absolute improvement in conversion rate is worth serious investment. But getting there from 800 visitors/month via CRO is not a coherent strategy.

This argument has limits worth naming directly.

If your product is in a paid acquisition channel — Google Ads, Meta, LinkedIn — and you’re spending real money on traffic, conversion rate becomes critical immediately. Spending $5,000/month on ads to a page converting at 1.5% is a capital efficiency problem. Here, fixing conversion before scaling spend is correct.

Similarly, if you have high traffic but it’s the wrong traffic — visitors with no intent to buy, from misaligned channels, or with a severe product-market mismatch signal in the behaviour data — then more traffic of the same quality won’t help. The diagnostic is whether your existing converting visitors are succeeding with the product. If they are, the problem is distribution. If they’re not, you have a product problem that traffic and conversion fixes both miss.

The useful test: look at the customers you do have. Are they finding value? Are they returning? Are they telling others? If yes, you have a distribution problem dressed as a conversion problem. If no, you have an earlier problem that more traffic will only expose faster.

Before any conversion sprint, run this audit first. Pull 90 days of data from whatever analytics you have — Google Analytics 4, Plausible, Fathom, PostHog — and answer four questions:

1. How many unique visitors hit the conversion page per month? If it’s under 1,000, stop reading CRO content and start reading distribution content. 2. Which traffic sources send visitors who convert at all? Channels with zero converting visitors in 90 days are a strategy problem, not a CRO problem. 3. What’s the ratio of people who could have found you (search volume for relevant queries, audience size on relevant channels) to people who actually did? If that ratio is 1,000:1 or worse, you have headroom in distribution that dwarfs any conversion gain. 4. If you doubled traffic from your best-converting source tomorrow, would the business meaningfully change? If yes: build that source. If no: you may have a product or pricing problem that sits upstream of both traffic and conversion.

The conversion funnel is real and worth optimising — after the funnel has enough input to make the optimisation matter. Most teams start there because the tools make it easy and the work looks like progress. The actual leverage is in the step before: getting people to the funnel in the first place.