Pipelines grow on patterns, not volume.

Why one PE-backed company walked away from “good pipeline” and rebuilt growth quality from the ground up

The moment I realised the pipeline was fundamentally misleading happened during a forecast slide.

Two deals. Same forecast call. Completely different futures.

Deal A:

• £120k ACV
• eight months in pipeline
• six custom feature requests
• forecast at 90% to close

Deal B:

• £35k ACV
• ideal customer profile match
• active trigger event
• immediate buying urgency

I asked the CEO a simple question:

“Which one are you chasing?”

The answer was obvious.

Deal A.

Larger contract. Larger forecast impact. Larger apparent upside.

Then I asked the question that changed the room:

“Which one looks most like the customers who actually stay?”

Silence.

That silence exposed the real issue.

The company was optimising for deal size instead of customer quality.

It was scaling pipeline volume instead of replicating the buying patterns of its best long-term customers.

The consequences were already visible underneath the reporting:

• high acquisition effort
• weak retention
• unstable revenue quality
• increasing commercial complexity
• long implementation drag
• poor expansion economics

Walking away from the wrong pipeline created immediate pain.

• the CRO threatened to resign
• three sales reps left
• the board called an emergency meeting
• Q1 missed by 23%

That is why many organisations cling to bad-fit pipeline.

The short-term pain is loud.

The long-term value is quiet.

But twelve months later, the economics looked completely different:

• close rate improved from 34% to 52%
• sales cycle reduced from 87 days to 31 days
• retention increased from 61% to 89%
• average deal size increased by 33%
• CAC reduced by 29%
• NPS improved from 31 to 68

The core shift was simple.

The company stopped chasing every available deal and started cloning the customers that actually compounded value.

That required a different commercial discipline.

Instead of asking:

“How big is the opportunity?”

The organisation started asking:

“Does this customer behave like the customers we most want more of?”

That question changed:

• qualification
• pipeline composition
• forecasting
• onboarding
• retention
• account expansion

Many PE-backed businesses struggle because the pipeline becomes a mixture of fundamentally different customer types.

The reporting still looks healthy.

The economics underneath do not.

The strongest growth systems are usually built around pattern recognition.

Not every customer should enter the pipeline.

The highest-performing companies understand exactly which buying patterns create durable revenue — and they systematically eliminate everything else before it distorts the operating model.

Pattern recognition starts upstream of any sales conversation. In ten seconds on the homepage, right-fit buyers either recognise themselves or they don't. Wrong-fit buyers either self-select out or they enter the system anyway.

The Shift90 X-Ray reads a B2B website the way a buyer reads it; in the first ten seconds, before they decide whether to keep reading. Thirty seconds to run. A 30-point readout that tells you whether buyers can recognise their situation, see why you are different, and find a reason to act now.

Run the X-Ray on your homepage.