Start a B2B Marketing Audit at the Revenue Number
Before anyone opens an ad account, write down the number the marketing is supposed to move, who owns it, and how it is split between new business and expansion. It sounds obvious. In most audits it is the first thing nobody can answer in one sentence.
Without that number, a channel review has nothing to be measured against. A cost per lead can only be "too high" relative to what a closed deal is worth and how many you need. The same goes for a content calendar: it can only be "too thin" relative to the market you are trying to open.
This is also where budget arguments get settled honestly. Gartner's 2026 CMO Spend Survey found that 56% of CMOs say their marketing organisation lacks the budget to deliver its 2026 strategy, while budgets held at 7.8% of company revenue, up from 7.7% a year earlier. A B2B marketing audit is the cheapest way to find out whether the shortfall is real or whether money already in the plan is going to the wrong accounts. Do that before asking finance for more.
The order we use for the rest of the audit mirrors the order we build programmes in: objective first, then the customer, then the motion, then measurement and delivery. The first of the six stages on our account-based marketing method is the revenue number, "agreed with sales and finance before anything is built." An audit should start in the same place, for the same reason.
Check the ICP Against the Deals You Won
Most companies have an ideal customer profile written down. Far fewer have checked it against their own closed-won and closed-lost deals in the last year. That check is the highest-value hour in the whole audit.
The reason is in the sales data. Ebsta's 2024 B2B Sales Benchmarks mid-year update found that only 18% of 2024 pipeline matched the selling company's own ICP, and that win rates were 3.1 times higher on opportunities that did. Most pipeline, in other words, was aimed at accounts the sellers themselves had said were not the best fit, and those accounts won far less often.
How to run the ICP check
- Pull four quarters of closed deals, won and lost, with industry, company size, region, deal value and source.
- Mark each one in or out of the written ICP. Do it with someone from sales in the room, because they will know which "in profile" deals were never really winnable.
- Compare win rate, deal size and cycle length for the two groups. If the in-profile deals do not win more often, the profile is wrong, not the pipeline.
- Look at where the out-of-profile pipeline came from. That is usually a campaign or a channel optimised for volume, and it is the first place to cut.
If the profile only exists as a slide, the audit's first finding is that it needs writing properly, and a go-to-market plan is where it gets decided.
Audit the Funnel Definitions and the Handoff
Marketing and sales describing the customer differently is a common failure, and an expensive one. Our go-to-market strategy page lists the result plainly: two definitions of the ICP means two funnels, two messages and a permanent argument about lead quality.
The audit here is mostly reading, not analysis:
- Stage definitions. Are lead, accepted lead, opportunity and closed deal written down, and would two people put the same CRM record in the same stage?
- Handoff rules. What has to be true before marketing passes something to sales, and what happens to everything that does not qualify?
- Follow-up service levels. Is there an agreed time for sales to act on a handoff, and does anyone check it?
- Sales acceptance rate. What share of what marketing hands over does sales agree is worth working? It is the fastest early warning that targeting or qualification has drifted.
When these are missing, every channel metric downstream is unreliable, because nobody agrees what is being counted. Fix the definitions before you trust a conversion rate. Our post on how to build a sales pipeline covers the stage design in more detail.
Audit the Measurement Before Trusting Any Number
This is the part of a B2B marketing audit that most often changes the conclusion. A channel that looks expensive may be the one creating the demand that a cheaper channel is getting credit for, and you cannot tell from a dashboard that has not been checked.
The measurement checks in a B2B marketing audit
- Platform-reported numbers are labelled as platform-reported. An ad platform's conversion value is that platform's claim about its own contribution. Adding two platforms' figures together and calling the total revenue double-counts the same deals.
- The attribution window is stated on every view. A 7-day and a 90-day window produce different answers from the same data. Neither is wrong; an unstated one is.
- Demand creation and demand capture are reported separately. Capture converts this month and creation compounds over quarters. Blended into one number, the slower channel always looks worse.
- Email engagement is not measured on opens. Apple's Mail Privacy Protection prevents senders from seeing whether a message was opened and hides the reader's IP address, though it does not extend to links. Use clicks, replies and meetings instead.
- Analytics retention is set on purpose. In Google Analytics 4, user-level data retention can be set to two months or 14 months, and the setting governs explorations and funnel reports rather than the standard aggregated reports. A year-on-year funnel analysis is impossible on a property still set to two months.
Those first three are the house rules on our own reporting, published on our demand generation page, alongside one more: nothing published that we cannot show you the source for. They are a fair standard to audit anyone's reporting against, including ours. For what to compare your numbers to once they are clean, see our guide to B2B marketing benchmarks.
Then the Channels: Spend Against Pipeline
Only now is a channel review worth the time, because you know who the programme is for, how a lead becomes an opportunity and which numbers can be believed. The question for every channel is the same: what did it contribute to qualified pipeline, at what cost per opportunity, and is it reaching the accounts in the profile?
| Channel | The question to ask | The evidence to pull |
|---|---|---|
| Paid search | Is it capturing demand from in-profile buyers, or buying clicks from anyone? | Cost per opportunity, search terms against the ICP, win rate on search-sourced deals |
| Paid social and LinkedIn | Is it reaching the buying committee at target accounts? | Account coverage and engagement on the target list, not reach or clicks |
| Content and SEO | Is it creating demand for the problem you solve? | Branded search volume over time, organic-sourced pipeline, self-reported attribution |
| Email and nurture | Does it move accounts forward, or just send? | Clicks, replies and meetings per thousand sends, stage movement after a sequence |
| Events and partners | Do the conversations become opportunities? | Meetings booked, opportunities opened within 90 days, deal value |
We report cost per opportunity, not cost per lead, on our demand generation work, and the audit should do the same. A cheap lead that sales rejects costs more than an expensive one that becomes a qualified opportunity, and only the CRM can show the difference.
Watch for one pattern in particular: a team told its cost per lead is too high will lower it, usually by widening the targeting until cheaper, worse leads arrive. The channel report improves and the pipeline does not. If the ICP check found a lot of out-of-profile pipeline, this is usually where it came from.
The Stack: Less Software, Configured Properly
The technology review comes last because it is the easiest to overrate. Intent platforms and account-based software are useful. None of them can decide which accounts are worth winning or who follows up on Tuesday.
On our account-based marketing engagements, the stack review is an honest audit of CRM, marketing automation, intent data, ads and enrichment against what the programme actually needs, and the recommendation is often to use less software, configured properly, rather than buy more. We are not resellers and do not take vendor commissions, which is the only reason that recommendation is believable.
Three questions do most of the work:
- Is the CRM the source of truth? If opportunity source, stage dates and deal value are not reliable there, nothing above it can be reported honestly. Our CRM systems work is built around getting every lead into one place, so that it can be.
- Are first-party signals used before bought ones? Your own site, CRM, email and sales engagement data are under-used almost everywhere. Exhaust them before paying for an intent data platform.
- Is anything paid for and unused? Seats nobody logs into and features nobody configured are the cheapest savings an audit finds.
What a B2B Marketing Audit Should Hand You
A good audit ends in decisions, not a score. If the output is a colour-coded spreadsheet of 140 issues, nobody will act on it. Ask for something closer to this:
- Findings ranked by what they cost in pipeline, not by how many there are or how easy they are to fix.
- An owner and a date for each one, agreed with sales, because most of the important findings are not marketing's alone to fix.
- A costed plan for the next 90 days, with what each change should move and how you will know.
- A written answer to the question you did not ask, including "this motion is wrong for you," if that is what the evidence says.
That is the shape of the diagnostic that starts most of our account-based work: a short, paid, fixed-scope assessment that ends in a written deliverable you own either way, covering the ICP, a first tiered target account list, the buying-committee coverage you already have, whether your data and reporting can support account-level measurement, and a costed plan for the first ninety days. If the honest answer is that you do not need account-based marketing, that goes in the document too.


