Revenue Diagnostics

What is a Revenue Diagnostic?

A Revenue Diagnostic is a structured pull-apart of how your revenue actually gets made and what it actually costs to make. Three to six weeks of work. The output is a written read on customer acquisition cost by channel, lifetime value by cohort, the attribution model and where it's lying to you, and the margins after every real cost is counted: fulfillment, support, refunds, platform fees, the operating overhead allocated honestly.

It's not a marketing audit. A marketing audit looks at campaigns and channels through a marketing lens, asking whether the creative was good, the targeting right, the bids efficient. A Revenue Diagnostic looks at the same channels through a P&L lens. The question isn't "is this campaign working?" It's "is this dollar of spend earning more than a dollar back, after everything is counted, on a cohort that lasts?"

The output is a written report and a working meeting. Not a deck. Anywhere we make a claim we cite the data behind it. Anywhere we flag a problem we put a dollar range on what it's costing you and what the fix is worth.

When does a company need one?

Four states tend to drive a Revenue Diagnostic.

CAC has crept up and nobody can explain it. The dollars going out per new customer are 30% or 50% higher than they were 18 months ago, but the channel mix looks the same on paper. Something has shifted under the reporting layer and the reporting layer isn't surfacing it.

The CFO and CMO disagree about which channel actually pays. The CMO is reading platform-reported attribution. The CFO is reading cash in versus cash out. Both can be right about their own number and both can be wrong about the underlying truth. Until somebody reconciles the two, channel-mix decisions are being made on contested data.

The unit economics on a product line or segment feel wrong but nobody has pulled it apart. The aggregate looks fine. The CEO senses that one product or one segment is dragging the blended number down. A Revenue Diagnostic with cohorts cut by product or segment will tell you whether the intuition is real and how much it's worth.

The attribution model just changed, or is about to. iOS 14, the GA4 migration, an MMM rollout, a switch from last-touch to multi-touch. The model change implies a channel-mix change, but nobody has tested whether the new model is showing reality or just a different fiction.

If two or more of those describe your company right now, the cost of operating without a diagnostic is usually higher than the cost of the diagnostic.

What do you actually look at?

Six inputs, roughly in order of how much they usually matter.

Channel-level spend, conversions, and attributed revenue across every paid platform: Meta, Google, LinkedIn, TikTok, programmatic, retail media, whatever you run. Pulled from the platforms directly, not just from the reporting layer.

CRM and pipeline data if there's a sales motion. Lead source, lead-to-close rate, sales cycle, deal size by segment. Most B2B companies don't actually know their CAC by channel because the platform numbers stop at MQL and the CRM numbers start at SQL, and nobody has tied the two halves of the funnel together honestly.

Order and transaction data with customer IDs. Used to build real cohort LTV, not the LTV the BI tool reports out of the box (which is usually wrong in interesting ways), but a cohort-by-cohort accounting that ages forward as the cohorts age.

Cost of goods, fulfillment costs, support costs, refund and return rates. The variable costs the marketing team usually doesn't see. Without these, "margin" is fiction.

The attribution model and what it's doing under the hood. Last-touch, position-based, MMM, blended. We read the methodology, run a few sanity checks against incrementality where we can, and flag where the model is making the company look better or worse than it is.

Operating overhead allocated to revenue lines. Not perfect cost accounting — proportional, with assumptions stated. Enough to answer whether a given segment or product is actually paying for the resources it consumes.

How is this different from what our analytics team already does?

Most internal analytics teams produce reports. A Revenue Diagnostic produces judgments. Different shape of work.

Reports answer "what is the number?" Judgments answer "what does the number mean, what is it worth in dollars, and what should we do about it?" Internal teams are usually staffed and incentivized for the first. They produce dashboards, weekly reads, channel reports, lifecycle metrics. That work is important and shouldn't go away.

But the act of pulling all of it together, reading it as one coherent picture, comparing it to how a hundred other companies in the same revenue band look, and writing an opinionated read on where the money is leaking is a different muscle. Most internal teams don't have the bandwidth or the cross-company reference set to do it well, and the ones that do are usually too close to the work to be honest about it.

The deliverable also looks different. An analytics team's natural output is a dashboard or a deck of charts. A Revenue Diagnostic is prose with charts in support of the prose. You can read it in 45 minutes. A board member can read it in 45 minutes. The decisions it implies are unambiguous.

What does the output actually look like?

Three things land at the end.

A written report. Twenty to forty pages depending on scope. Numbered findings, each with a dollar range on what it's costing you and a one- or two-paragraph reasoning trail to the recommended fix. No glossy chart design. No executive-summary deck.

A walk-through meeting with the operating principals — CEO, CFO, CMO, whoever owns the decisions the report implies. Two hours. We answer questions, defend specific findings, and push back if we think a fix is being misread.

A prioritized fix list, ranked by dollars-at-stake divided by effort-to-implement. The top item is usually the obvious one in hindsight. The top three are usually worth a multiple of what the diagnostic cost.

Who is this not a fit for?

Three disqualifiers.

Companies whose finance data is too broken to read. If COGS isn't tracked, refund data isn't categorized, customer IDs don't tie across systems, and the CRM is a free text field, we can clean some of it as part of the engagement. Past a certain point, though, we'd be doing six weeks of data engineering before the actual diagnostic could start. Better to fix the underlying data first and commission the diagnostic on top of clean inputs.

Companies that won't act on the answer. The most expensive failure mode is commissioning a diagnostic for a political reason, getting back a clear-eyed read, and then ignoring it because the answer doesn't match the position someone needed it to support. We screen for this on the first call.

Companies under $2M in revenue. Below that threshold, there usually isn't enough channel diversity or cohort depth for a structured diagnostic to find something the founder doesn't already know. The right move at that size is direct work with the operator on the one or two channels actually carrying the business, not a multi-week audit.

What happens after the diagnostic?

Whatever you want.

We don't make the Revenue Diagnostic conditional on engaging us further, and we don't price it as a loss leader for a bigger deal. The deliverable stands on its own. About half of diagnostics end with the inside team executing on the prioritized list themselves; we're available for advisory check-ins if the team wants them, and otherwise we step out.

About a third end in a fractional engagement to actually rebuild the systems the diagnostic identified as broken. That happens when the fix list is large enough that an inside team can't carry it on top of their existing load, or when the fixes need someone who has shipped them before in a comparable company.

The remainder are split. Some companies use the diagnostic as a hiring brief. The prioritized list becomes the JD for the next senior hire, and the recommendations get owned by that person on day one. A few end in the diagnosis that the current team is doing the right things and the revenue problem is somewhere outside marketing. Those reads are uncommon but they happen, and we say so when we see it.