This is a pattern composite, not a real client. The arc is assembled from operator threads, the channel-economics research below, and the strategic patterns I see across consulting work. Specific numbers are illustrative, drawn from the firms I’ve watched make this kind of bet.
I’ll call him Daniel. Mid-forties, founder of a 16-person services firm, mid-market clients, deal sizes between twenty and thirty thousand a year. Three years of revenue growth, paid-channel-led. By Q1 2025, his cost per qualified lead had climbed every quarter for six straight quarters. He was spending close to eight thousand pounds to acquire a customer worth twenty-five thousand a year. The CFO modelled the trend forward and showed him the line where the unit economics broke entirely, somewhere in Q4. Daniel sat down with his head of growth and his CFO and asked the question founders commonly avoid: what if the channel mix we’ve used for five years has structurally stopped working, what would we do instead.
What was actually happening to the unit economics?
Daniel’s CAC had compounded rather than drifted. Paid search and LinkedIn together had moved from a blended £1,300 per qualified lead in 2021 to nearer £4,500 by Q1 2025, with conversion rates softening at the same time. MSP Success documented the same arc across B2B services in 2025, with cost per qualified lead more than doubling inside two years for many firms. The firm was running harder to stand still.
The pattern was structural rather than firm-specific, driven by AI-flooded content saturation, increased advertiser competition for the same intent keywords, and buyers who’d learned to filter out the surface noise. The honest read on the numbers was uncomfortable. At the old conversion rates, the firm could keep going. At the new ones, doubling spend wouldn’t double pipeline, it would inflate CAC further. Daniel’s CFO ran a sensitivity analysis showing that even with a fifteen per cent reduction in cost per lead, the unit economics would be marginal by year-end. The standard playbook offered two options: spend more to maintain volume, or diversify into more saturated channels and hope volume offset cost. He took a third option, the contrarian one.
What was the reframe that changed the question?
The team accepted that paid acquisition was structurally breaking rather than temporarily expensive. Once they accepted that, the question changed. It stopped being “how do we lower CAC in the channels we use” and became “which channels are now relatively undervalued versus where the market is putting money”. Where everyone is, value compresses. Where almost nobody is, value opens up.
eMarketer’s 2025 work on AI content saturation made the underlying mechanic explicit. Generative AI has flooded the channels that were already crowded, which means buyer attention has moved to narrower, more opinionated, more human voices. The reframe was permission to ask a different set of questions.
Where are the buyers we want actually paying attention now? What would it look like to be the firm those buyers know by name in eighteen months? What can we afford to stop doing if we commit to being known narrowly rather than reaching everyone? The team spent two weeks answering those questions before they touched a single channel. The work that followed was downstream of that reframe, not parallel to it.
What did the four-move bet actually look like?
The team executed four moves over twelve months, deliberately as a package. One, narrowed the ICP from “mid-market services firms” to one specific named segment of roughly 1,200 firms in the UK. Two, founder-led narrow content, fortnightly, opinionated, written by Daniel personally. Three, a small outbound function of three people working a tight list of 400 named accounts. Four, sustained participation in two specific buyer communities.
Each move had a discipline behind it the standard playbook would have softened. The narrow content often took controversial positions inside the niche. Haus Advisors’ framing, that positioning is strategic decisions and not website copy, sat behind the ICP work. The outbound team targeted a reply rate above two per cent rather than chasing volume. The community presence was treated as a long-term commitment, not a promotional channel.
The fourth move was the one services firms commonly cut first. Communities don’t pay back in a quarter. Daniel held it because the relative-value logic told him the firms that showed up, contributed, and didn’t sell would be the ones the community remembered. The outbound numbers backed the choice in a different way. GenSales’ 2025 data put cold-call conversion at around two per cent versus cold email at half a per cent, the four-to-one ratio that justified investing in real human outreach over sequenced automation. The whole package only worked because each move reinforced the others. Narrow content gave outbound something to point to. Outbound surfaced names the communities knew. The communities surfaced invitations the founder couldn’t have bought.
Why did the early months feel like a mistake?
Months one to four looked worse on every dashboard the firm had. Inbound dropped because paid spend dropped. The volume metrics the team had spent five years optimising went the wrong way. Two account managers asked whether they should brush up their CVs. The CFO noticed that revenue from new logos had visibly slowed. The discomfort was real.
Daniel held the line on the bet partly through conviction, partly because the fallback, returning to the old model, would have crystallised the unit economics they were running from. The discomfort was also the reason firms commonly can’t take this bet, which is precisely why the relative-value arbitrage was available.
The leading indicators turned in months five to nine. Reply rates from outbound were strong, well above the two per cent target on the right segments. The narrow content started getting shared inside specific buyer groups. The two niche communities began inviting Daniel to speak, first as a panellist, then as a workshop lead, then as a co-author on a community-led research piece. Pipeline started filling with prospects who came in already knowing what the firm did and why. Sales cycles shortened because the qualifying conversation had largely happened before the first meeting. Price sensitivity dropped because the prospects had self-selected on fit rather than landed via a paid funnel. The shape of the pipeline was structurally different.
What did the result look like at eighteen months?
By month eighteen, new deals were closing at higher average values, around £35,000 against the previous £25,000, with shorter sales cycles and lower acquisition cost per customer than the firm had seen at its old peak. Daniel’s firm had become the obvious answer in a niche where, eighteen months earlier, they’d been one of forty competing for the same attention. Growth had become structural rather than coincidental.
Goldman Sachs’ 2026 small-business research framed the broader pattern. Ninety-three per cent of SME owners report AI is having a positive business impact, but only 14 per cent have it embedded operationally, and the buyers who valued narrow expert positioning were the ones moving fastest. The bet had been on accepting the old channels had broken, not on finding a clever new one. The firms that take that bet win the relative-value arbitrage that opens up because the rest of the market is still spending into saturation.
The firms that don’t, often because the four-month discomfort is too high to absorb, end up where Daniel’s CFO had modelled they would: spending more to acquire less, optimising channels that the underlying mechanics no longer support. I’ve watched a few firms make this kind of bet, and the pattern after twelve months is consistent. The ones who hold the line get a clearer story to tell, a tighter book of clients, and a CAC line on the spreadsheet that finally moves the right way. The maths makes it inevitable. The discomfort makes it rare.
