When growth flattens, the instinct is to add: more channels, more content, more spend, a new agency. Activity rises, the line does not, and everyone concludes the market got harder. Sometimes it did. More often, the engine has a single broken joint, and every added dollar of traffic is being poured through the same crack.
The mechanism is one page, drawn honestly. Map the actual funnel. Not the strategy-deck version, the real one: every channel where attention arrives, every handoff (ad to page, page to enquiry, enquiry to call, call to proposal, proposal to close, close to repeat), and the true number at each step, this quarter, from the systems. If you cannot produce a real number for a step, mark it in red; an unmeasured joint and a broken one are indistinguishable from the outside, and fixing measurement there is itself the first move.
How it actually works #
With numbers on the page, the shape of the problem changes. Plateaued engines almost never leak evenly. One transition is drastically worse than the pattern around it. Enquiries that never get a reply inside a day. A proposal stage where deals go quiet. A landing page converting at a fraction of what the traffic quality suggests. That is the broken joint, and it has a property the "do more marketing" conversation never has: it is specific, bounded, and fixable this month.
Fix it before optimising anything else, because the joint sets the value of every upstream improvement. Doubling traffic into a broken proposal stage doubles the waste; repairing the stage first means every existing channel instantly performs better with zero new spend. This is why the architecture conversation has to precede the tactics conversation. Agencies optimise the step they were hired for, but nobody inside a channel can see the whole spine.
Running it on Monday #
One page, real numbers, one red pen. If you cannot draw it, that is the diagnosis. If you can, the broken joint will be looking back at you by the end of the afternoon.
Stop feeding the engine. Open it.