Here is a question worth asking in your next planning meeting, in exactly these words: if we turned off every paid channel tomorrow morning, what would still be in the pipeline in six weeks?
The answer is usually a pause, then a hedge, then a change of subject. That pause is the most useful diagnostic in growth marketing, because it is the difference between a business that has built an asset and a business that has been renting attention and calling it traction.
The two look identical until they do not
A business with £40k of monthly paid spend producing steady pipeline and a business with an owned audience producing steady pipeline have the same dashboard. Same leads, same conversion, same forecast. Every board slide treats them as equivalent.
They are not remotely equivalent, and the difference shows up in exactly one circumstance: when the spend stops. One of them has a slow month. The other has nothing, discovers it has nothing, and then discovers that the thing which would have prevented it takes two quarters to build - starting now, from zero, under pressure.
This is the Growth Engine stage, and it is where businesses stall longest. Not because the work is hard, but because nothing forces the issue. Rented demand keeps working. There is no crisis, no falling metric, no quarter where somebody has to explain. The cost of staying here is entirely invisible right up until it is the only thing that matters.
Three checkpoints, and the one everybody skips
Growth Engine is gated by three of the twelve growth checkpoints, in this order:
- Trust and proof - reviews, case studies, and specific results that reduce buying risk without you in the room.
- Owned audience engine - email and organic presence that keeps producing when spend stops.
- Sales follow-up system - a repeatable process for capture, qualification, follow-up, and handoff.
Most teams that attempt this stage attempt the middle one. They start publishing. They build a list. And then they stall, because they skipped the first one and never got to the third.
Proof comes before volume
The reason to put trust and proof first is that it changes what your content has to do. A sceptical stranger arriving at a site with three specific, evidenced customer outcomes is a different reader than one arriving at a site with thirty posts and no evidence. The first can be persuaded by an article. The second has to be persuaded that you exist first, which an article is badly suited to do.
Published proof is also the thing that makes content compound rather than accumulate. A post that can cite your own result is a post that has a reason to exist. A post that can only synthesise what is already on page one is, by construction, a summary of the consensus - and there is an unlimited supply of those now.
Teams skip it because it is uncomfortable. Getting a customer to go on record means asking, and asking means finding out whether the result was as good as you have been saying. That discomfort is doing real work, and routing around it is how a content programme ends up with volume and no credibility.
The follow-up system is where the money actually leaks
The third checkpoint is the one nobody wants to own, and it leaks more pipeline than the other two combined.
Here is the shape of it. A lead arrives. The person who normally handles leads is on holiday, or in a sprint, or was never formally assigned. The lead gets one touch instead of four, or four instead of one, depending on who noticed. Nobody records which. Three weeks later everyone agrees the channel is underperforming.
The channel is fine. The handoff is not. And the leak is invisible because the thing you would need to count - leads that got no second touch - is precisely the thing nobody is counting.
"Repeatable" here means one specific, testable property: every inbound lead gets the same defined treatment whether or not the person who usually handles them is available. If that is not true, demand generation is pouring water into a bucket whose holes you have not measured, and making the tap bigger is the wrong next move.
The test, stated so it cannot be fudged
The checkpoint is not "we have a newsletter" or "we publish regularly" or "SEO is part of the plan". Those are activities. The checkpoint is an outcome, and the honest version is:
Turning off paid spend for a month would reduce your pipeline rather than end it.
Reduce, not end. Nobody is claiming owned demand replaces paid - a good Growth Engine business still runs paid, and runs it better, because it now has a baseline to measure incrementality against. The claim is narrower and harder: something has to survive the switch being flipped.
If you genuinely do not know the answer, that is worth knowing, and you do not have to run the experiment to find out. Look at what share of last quarter's pipeline came from sources that do not have a media cost attached. If that number is under about a fifth, you are renting.
Why this stage stalls for years
Three reasons, and they compound.
The first is that the metric moves slowly. Share of pipeline from owned sources is a quarterly number at best. Paid channel metrics move daily. Given a choice between steering by something responsive and something slow, teams steer by the responsive one and then wonder why the slow one never moved.
The second is that it competes with itself for attention. The work that builds owned demand - proof, publishing, follow-up systems - is the work that gets deprioritised the moment a paid channel needs attention. And a paid channel always needs attention. That is what running one is.
The third is that publishing is now cheap enough to feel like progress. This is the new one, and it is the most dangerous. Agents have made drafting almost free, which means a team can produce the appearance of an owned-audience programme - forty posts a quarter, a full calendar, a dashboard of publish counts - without producing the thing itself.
The rule that protects you from your own throughput
If you take one operational thing from this piece, take this: set your publishing cadence to your review capacity, not your drafting capacity.
Drafting stops being the constraint the moment agents are involved. Review does not - review is a person reading carefully, and that has not got faster. Decide how many pieces someone can genuinely edit each week, with the authority to say "this should not exist", and make that number your ceiling.
What you do with the surplus drafting capacity is better briefs and deeper research per piece, not more pieces. This is the single highest-leverage rule at this stage and reliably the first one abandoned, because the surplus is sitting right there and using it feels like the obvious move.
It is also why content in Spryxa is approval-gated by default and an ignored approval stays a draft indefinitely - there is no timeout that publishes on your behalf. The constraint that protected content quality was never drafting capacity. It was editorial judgement, and scaling the first without scaling the second is how a programme acquires the texture of unreliability, one unremarkable paragraph at a time.
What to do this quarter
- Calculate the share. What percentage of last quarter's pipeline came from sources with no media cost? Write the number down. It is your baseline.
- Get one customer on record. Not a testimonial - a specific outcome with a number in it. One is enough to change what every subsequent piece of content can do.
- Write down your review ceiling and make it the published cap, even if drafting capacity is ten times higher.
- Audit one week of inbound leads against what was supposed to happen to them. Count the ones that got no second touch. That number is usually the cheapest pipeline available to you.
The stage ends when the switch-off answer stops being a pause. The full model, including what comes next and why Scale Readiness is about economics rather than volume, is in the business growth stages guide.
