Revenue growth systems: connecting marketing to the number

There is a specific meeting that recurs in most growing companies. Marketing presents traffic, engagement and MQLs. Sales presents pipeline, close rate and revenue. The numbers do not reconcile, everyone explains why their own numbers are the right ones, and no decision gets made.

That meeting is a symptom. The cause is that nobody owns the model connecting the two.

Map before you optimise

The first deliverable is a map of the revenue model end to end, with the current real number at each stage. Visitors, enquiries, qualified conversations, proposals, closed deals, retained accounts, expansion revenue.

Two things fall out of this immediately. First, most organisations cannot populate every stage, and the gaps are where the problems live. Second, the binding constraint becomes obvious — and it is usually not the stage receiving the most attention.

The constraint is rarely where the budget is

A business converting two percent of qualified conversations does not have a traffic problem. Doubling traffic doubles the waste.

A business with strong conversion and thirty percent annual churn does not have an acquisition problem either. It has a bucket with a hole in it, and acquisition spend is water.

Identifying the constraint honestly is uncomfortable because it usually means telling whoever owns the well-performing stage that more of their work will not help. It is also the single highest-value output of the exercise.

Expansion revenue is the ignored half

Most growth programmes are acquisition programmes with a retention section appended. That inverts the economics: selling more to an existing customer is generally substantially cheaper than winning a new one.

Building it deliberately means knowing which customers are candidates for which additional service, having a defined trigger for the conversation, and measuring expansion revenue as its own line rather than folding it into total revenue.

It is unglamorous and it is where the margin is.

Forecasting without pretending

A forecast that presents a single confident number for a multi-touch B2B pipeline is arithmetic that will not survive contact with the CRM.

A useful forecast states its assumptions — conversion rates by stage, average cycle length, seasonality — and presents a range with the confidence attached. When it misses, you can identify which assumption was wrong, which is what makes the next one better.

Publishing the assumptions also makes the forecast auditable by the people who will be held to it, which materially improves the quality of the inputs.

Attribution, stated honestly

No attribution model captures a multi-touch B2B journey cleanly. Zero-click AI exposure, offline referral, brand recall and dark social all produce revenue with no attributable source.

The correct response is not a more elaborate model. It is stating the limit in the report: this is what we can attribute, this is what we cannot, and here is the proportion of revenue arriving through unattributable paths.

Reports that state their limits are trusted more, not less — and they survive the meeting where someone finally checks.

Common mistakes

• Optimising the stage with the most attention rather than the stage that is binding.
• Reporting revenue without separating new, retained and expansion.
• Presenting a forecast without assumptions or confidence bounds.
• Implying attribution precision the data does not support.
• Treating retention as customer success's problem rather than part of the growth model.
• Buying a growth system when a single service would fix the actual constraint.