If you search for "marketing audit framework," you'll find a standard template: review your website, review your SEO, review your paid ads, review your social media, review your email. Produce a document. Present it.
This format produces long lists of findings and almost no identification of the actual constraint. It's useful for consultants who need to demonstrate effort. It's not useful for founders who need to know what to fix first.
Here's what a marketing audit needs to cover to be actionable.
Layer One: The Measurement Foundation
An audit of marketing outputs is only useful if the measurement system can tell you what's true. The first question is not "what are our results?" It's "how would we know if our results were accurate?"
What to check:
GA4 configuration. Is the tracking code on every page? Are conversions defined and firing correctly? Are the goal definitions aligned with actual business outcomes (not just "contact form submitted" but "qualified lead form submitted")? Is session attribution set up correctly for your mix of direct, organic, paid, and email traffic?
UTM discipline. Every paid campaign, every email campaign, every partner link should have consistent UTM parameters. If UTMs are missing or inconsistent, the attribution in GA4 is wrong. You cannot make channel allocation decisions on corrupted attribution data.
CRM pipeline data. Are leads being logged consistently from all sources? Is there a source field tracking where each lead came from? Is the pipeline stage definition consistent — does "qualified opportunity" mean the same thing to every sales rep?
If the measurement foundation is weak, everything downstream is guess work. Most audits skip this because it's boring. It's also the part that makes the rest of the audit meaningful.
Layer Two: The Positioning and ICP Audit
The most common root cause of underperforming marketing is positioning that's too broad — a value proposition that's true for too many people to be specifically compelling to any of them.
What to check:
ICP definition specificity. Can you describe your best-fit customer in enough detail to know whether a specific company is in or out? Company size, industry, and tech stack are necessary but not sufficient. What's the trigger event — the thing that made them start looking for a solution? What's the pain point in their language, not your marketing language?
Positioning validation against closed-won. Take the last 10 closed-won customers. Did they match the ICP definition? If not, the ICP is wrong. Also: how did they describe the problem before they found you? If that language doesn't appear anywhere on your website, the positioning is misaligned with how buyers actually search.
Competitive displacement. When you win, who do you displace — or what do you replace? "We're better than the alternative" is not positioning. "We're the solution for founders who tried the agency model and found it produced deliverables rather than results" is positioning. Do you know your specific displacement story?
Layer Three: Channel Performance Analysis
This is where most audits start. It's the third layer, not the first, because channel performance data is only interpretable against the backdrop of accurate measurement and clear ICP.
What to check:
Conversion rate by channel and by content type. Not overall conversion rate — segmented. Organic traffic from branded search converts differently than organic from non-branded. Paid traffic from retargeting converts differently than cold. Newsletter subscribers convert differently than cold inbound. Know the numbers by segment before drawing any conclusions.
CAC by channel. For any channel receiving meaningful budget, calculate the fully-loaded CAC including time cost (someone is managing it) and direct spend. Compare to average contract value. If the payback period exceeds 18 months, the channel math is broken regardless of traffic volume.
Content performance by topic cluster. For companies with content programs: which topics drive traffic that converts, and which drive traffic that bounces? A high-traffic article that produces zero leads is not an asset. An article that gets 200 visitors and produces 3 qualified leads is.
Email list health. Deliverability rate, open rate trend (not the absolute rate, the trend), reply rate, unsubscribe rate. A declining open rate is the early signal that your list quality is degrading — either through churn of engaged subscribers or accumulation of disengaged ones.
Layer Four: The Handoff Audit
Marketing-to-sales handoffs fail more often than any other part of the system and get audited less often than any other part.
What to check:
Time from lead submission to first sales contact. This should be under 4 hours for inbound leads. Every hour over that threshold reduces conversion probability. Know the actual number, not what the process says it should be.
Lead information completeness at handoff. When a sales rep receives a new lead, do they know what content the lead engaged with, what problem they indicated, what company they're from, and what the lead source was? Or do they receive an email with a name and a form fill? The quality of the handoff determines the quality of the first sales conversation.
MQL-to-SQL rate. What percentage of marketing qualified leads become sales qualified leads? If this number is below 20-25%, either marketing is generating too many unqualified leads or the SQL definition is too restrictive. Know which problem you have.
Layer Five: The Strategic Constraint
After the four layers above, there's usually one finding that stands out: the specific constraint holding the system back.
Almost always, it's one of:
- Attribution is broken, so channel investment decisions are made on bad data
- ICP is too broad, so content and paid traffic attract unqualified intent
- Handoff is slow or information-poor, so qualified leads don't convert
- Proof doesn't speak to specific problems, so the bottom of the funnel underconverts
The value of a real audit is identifying which of these is the primary constraint — the one that, if fixed, moves the most critical metric. Not all four have equal weight in every system. Finding the one that matters most, in the right order, is what makes the engagement valuable.
The list of everything that could be improved is infinite. The list of things worth fixing first, ranked by impact and dependency, is usually 2-3 items. That's the deliverable a good audit produces.