Pillar 03

Acquisition Channel Development

Every channel you depend on today is getting more expensive, and the one that replaces it is currently cheap because nobody has found it yet. This is the programme that goes looking — with a budget cap and kill criteria written down before a dollar moves.

Get a free acquisition audit Book a strategy call

The problem

Channel decay is predictable. Being unprepared for it is a choice.

Every acquisition channel follows the same arc. Early entrants get absurd returns. Word spreads. Competition bids the price up. The returns compress until the channel is merely acceptable, then marginal, then negative — and the businesses still on it are the ones who never built anything else.

Most companies respond by testing occasionally, when results get bad enough to force it. That is the worst possible time: budget is tight, patience is short, and a test run under pressure gets read too early and killed for the wrong reason.

The alternative is to always have three or four small bets running, funded from a separate line, judged on their own timeline. That is what channel development means here.

Signs you need this

Over 70% of your new customers come from a single channel
Cost per acquisition rises every year and you absorb it
The last new channel you tried was more than a year ago
Tests get cancelled early because "it wasn't working yet"
A platform policy change could take out most of your pipeline
You have no idea which channels you have already ruled out, or why

The method

Every test is written down before it is funded

A test with no pre-agreed success threshold is not a test. It is a spend with a story attached afterwards.

Map the opportunity set

Every channel your buyers plausibly touch: platforms, marketplaces, communities, partnerships, affiliates, offline, sponsorships, creators. Scored on reachability, cost and effort to test.

Write the hypothesis

One page. What we believe, why, the audience, the offer, the budget cap, how long it runs, what "working" means numerically, and the conditions under which we kill it.

Build the measurement

Decided before launch — geo holdout, matched-market comparison or clean last-touch, depending on the channel. Some channels cannot be measured cleanly, and we say so up front.

Run to the cap

The test runs its full window on its full budget unless a kill criterion trips. No reading results on day three. No quiet budget increases because it "feels promising".

Judge honestly

Against the threshold agreed at the start, not a revised one. Roughly one in four tests works. That ratio is the programme functioning correctly, not failing.

Scale or shelve

Winners get a scale-up playbook and move into the main programme. Losers get a written record of what was tried and why it failed, so it is not repeated in two years.

Portfolio thinking

Three to five bets running at all times

The programme is deliberately structured like a portfolio rather than a sequence of one-off experiments. Several small positions run concurrently, sized so that any single failure is affordable and uninteresting.

We typically recommend ring-fencing 10–15% of acquisition budget for channel development. Below that the tests are too small to read. Above it you are gambling with money that should be buying customers today.

Every quarter we sit down with you and your finance team, review what ran, what it cost, what it returned, and what gets funded next.

Example quarter Illustrative

Creator partnerships Beat threshold — moving to scale-up Scale

Niche marketplace listings Marginal, extending one more cycle Extend

Programmatic audio Killed at week 4 — CPA 3x threshold Killed

Community sponsorships Week 3 of 8, holdout running Live

Sample portfolio for illustration — not client data.

Deliverables

An institutional memory for acquisition

The most under-rated output of this programme is the record of what did not work.

Most companies lose that knowledge every time someone leaves. Two years later the same idea comes back around, gets tried again with the same result, and nobody remembers the first attempt. The channel ledger stops that.

Talk about a programme

Opportunity map. Every plausible channel, scored and prioritised.
Test briefs. One page per bet: hypothesis, cap, threshold, kill criteria.
Measurement plan. How each test will be read, agreed before launch.
Channel ledger. A permanent record of everything tried and its result.
Scale-up playbooks. For every channel that clears the bar.
Quarterly review. Run with you and your finance team.

Questions

Channel Development, specifically

How much budget does this need?

We suggest 10–15% of acquisition spend, ring-fenced. The important part is that it is a separate line. Development work funded from whatever is left over at month end never survives a bad month — which is precisely when you most need a second channel.

What if most of the tests fail?

Most will. Roughly one in four clears the threshold, and that is the expected shape of the programme rather than a bad quarter. The economics work because failures are capped at a known amount and winners are uncapped.

Can we just test channels ourselves?

Yes, and some teams do it well. The two things that usually break it in-house are discipline and measurement: tests get read early because someone is impatient, and results get attributed to the wrong cause because no holdout was built. Both are process problems, not talent problems.

How long before we know if the programme is working?

One full quarter for the first read, two before the pattern is meaningful. Anyone promising a validated new channel in six weeks is either running tests too small to trust or reading them too early.

Do you use our existing agency or team to run the winners?

Wherever possible, yes. The scale-up playbook is written to be handed over. If your team can run the channel once it is proven, that is cheaper for you and keeps us focused on finding the next one.