Channel-market fit

Poor distribution — not product — is a leading cause of failure. Engineers over-index on building and under-index on how strangers will ever hear about the thing. A mediocre product with a working channel often beats a beloved product trapped in a private Slack of early adopters.

Distribution is the set of channels that repeatedly put your product in front of people who can buy. GTM strategy picks the motion (PLG vs sales). This page is the first piece: what a channel is, and the arithmetic that says it can or cannot work.

Why one channel beats five half-channels

Distribution follows a power law — one channel usually dominates. Spraying effort across five dilutes learning. Trying "a little sales, a little ads, a little viral" is how you get zero working channels.

PayPal's early growth was a deliberate channel into eBay power sellers plus referral incentives. Hotmail put a signup link in every email. The pattern: pick a wedge audience and a mechanism, then obsess.

You need a theory of product-market fit and a theory of distribution. One without the other stalls — great product, no path; or loud acquisition into a product nobody keeps.

Walking through a math hole

Paid search converts at 2% from click to trial and 10% from trial to paid. Target CAC is $400 (from LTV math). Max cost per click:

Paid customers per click = 0.02 × 0.10 = 0.002
Max CPC                 = $400 × 0.002 = $0.80

If Google auctions clear at $2.50 for your keywords, this channel cannot work until conversion or ACV rises. No amount of "better copy" fixes a 3× math hole. That is channel-market mis-fit.

The same arithmetic applies to outbound (labor cost per meeting × close rate) and to partnerships (rev share + BD time). If the unit economics cannot bear the channel's cost shape, pick another channel.

Common mistakes

• Spraying five channels. You will not learn which one works.
• Blaming copy for a math hole. A $2.50 CPC against an $0.80 ceiling is not a creative problem.
• Distribution without retention. Loud acquisition into a product nobody keeps is wasted CAC.

Check yourself

2% click→trial, 10% trial→paid, $400 target CAC. Max CPC?

$0.80. 0.02 × 0.10 = 0.002 customers/click. $400 × 0.002 = $0.80.

Why is one working channel enough?

Distribution is a power law. One path that reaches the ICP at a bearable CAC can build the company.

Key takeaways

• Channel-market fit is arithmetic: conversion × ACV must clear the channel's cost.
• Most startups get zero channels. One is enough. Five half-channels is none.
• You need a PMF theory and a distribution theory.
• If the auction or the labor cost is 3× your ceiling, change conversion, ACV or channel — not the headline.

Further learning

Next, the channel catalogue.