Your next 30 days: the concentration plan

Enough principle. Here is the plan you can run this month, the concrete sequence that turns channel selection from a guess into a search and ends with you concentrated on one channel that fits.

Week one: let your economics narrow the field

Before you test anything, score the menu against your ACV and your constraints. Write down your average contract value and your time-to-value, and use them to eliminate whole channel classes: a low ACV with fast time-to-value rules out high-touch outbound, a high ACV rules out pure self-serve. Then run the matrix. Set your optimisation priority, learning, volume or cost. List your real constraints of time, money, audience and legal. Score the survivors on targeting, cost, input time, output time, control and scale. You should finish the week with a shortlist of three or four candidates, each of which could plausibly clear a three-to-one lifetime-value-to-acquisition-cost ratio at your ACV. Point your agents at the nineteen-channel brainstorm to make sure you have not let bias delete a candidate you never considered.

Weeks two and three: cheap parallel tests with a kill bar

Run your three or four candidates as cheap parallel tests, and set the kill bar before you start, not after, so the results cannot be rationalised. Decide in advance what a passing signal looks like: a cost-per-qualified-lead ceiling, a minimum conversion rate, a reachability threshold. Let your agents stand up the tests, drive the creative and instrument the reads while you watch the signals come in. Kill ruthlessly against the pre-set bar. The tests are small and fast by design; you are buying a learning signal, not building a programme, so do not let any single test bloat into a months-long project.

Week four onward: concentrate and instrument

Name your winner, the single channel that cleared the bar fastest, and concentrate everything on it. Now instrument the two numbers that will run the rest of the story. Track your lifetime-value-to-acquisition-cost ratio so you know the channel actually works rather than merely produces leads. And track your marginal acquisition cost over time so you can read the saturation signal honestly when it eventually comes. Then spend your traction half going deep, experiment volume, segment coverage, sustained optimisation, and hold the line against the second channel.

The discipline that compounds

Re-run the whole search only when the chosen channel proves it has genuinely plateaued: marginal cost rising after real depth, not boredom, not a competitor's shiny tactic, not a flat month. When that day comes, you will have a working machine funding the next search and a process you already trust. That is what building the growth machine looks like, one channel run ten times deep, then the search again, for as long as the business grows.