Marketing Report Numbers Do Not Mean What They Claim To Mean

Every metric below has a legitimate technical meaning. None of them are fake. What they routinely become, inside a board deck, is a place to hide from a harder question: did the spend cause the outcome, or did the outcome simply happen to occur near the spend. These eight sit closest to board and Operating Partner capital allocation decisions. Each entry states what the metric is presented as proving, what it technically measures, and a test you can run against your own data this week.

Conversions, event-based

Claimed: the number of buyers who took a meaningful action because of marketing.

Actual: any event a team configured as a conversion, counted regardless of whether the visitor would have taken that action through a different channel, or with no marketing exposure at all.

Test: check whether the conversion event fires for visitors arriving through direct traffic, branded search, or referrals from existing customers. If it fires often in that group, the event is not proof marketing caused the action.

Traffic acquisition channel grouping

Claimed: which channel is driving growth.

Actual: a rules-based label assigned to a session based on the last recorded referrer at the moment of arrival, with no visibility into what happened before that visit.

Test: for a sample of conversions credited to one channel, check whether the same user had a session from a different channel in the prior thirty days. If that pattern shows up in more than half the sample, the credited channel is closing a journey another channel started.

MQL and lead score

Claimed: a lead who is ready, or close to ready, to buy.

Actual: a point total for actions such as downloading a PDF, opening an email, or visiting a pricing page, weighted by rules a marketing team wrote and rarely revisits.

Test: compare the close rate of scored MQLs against the close rate of a random sample of unscored leads of the same size. If the two close rates sit close together, the score is not predicting purchase likelihood.

ROAS and blended CAC

Claimed: the return generated per pound of marketing spend, and the true cost of acquiring a customer.

Actual: a ratio that assumes every counted conversion required the spend that touched it last.

Test: recalculate CAC using only non-branded, category-intent conversions for your top channel, and compare it to the reported blended CAC. If the two numbers diverge materially, the blend was hiding the real cost.

Last-touch attribution

Claimed: the channel responsible for the sale.

Actual: whichever channel recorded the final interaction before conversion, regardless of the sequence of decisions that preceded it.

Test: for a sample of conversions, check whether an earlier marketing touch exists more than fourteen days before the credited one. If it consistently does, the credited channel closed a decision it did not create.

Pipeline growth and pipeline velocity

Claimed: the business is generating more qualified demand, faster.

Actual: a count of opportunities entering a CRM stage, which can rise because sales lowered its bar for what counts as an opportunity.

Test: compare opportunity-to-close rate this quarter against the trailing four quarters. If velocity improved while close rate fell, the pipeline is moving faster because the entry bar dropped, not because demand grew.

Pipeline value without probability weighting

Claimed: revenue that is on its way.

Actual: the sum of every open opportunity's listed deal size, regardless of stage or actual likelihood of closing.

Test: recalculate the same pipeline using stage-specific historical close rates instead of full deal value. If the probability-weighted figure is a fraction of the headline number, the headline number was never a revenue forecast.

Branded search volume without conversion

Claimed: growing brand awareness driving demand.

Actual: search volume for the company's own name, which can rise from press coverage, a controversy, or a competitor mention with no purchase intent behind it.

Test: check what share of branded search sessions convert to a pipeline stage within thirty days. If that share is flat or falling while branded search volume rises, the volume increase is not translating into demand.