A marketing audit should answer two questions: where is growth leaking, and what should you fix first? Work through six areas, from positioning to team and process, run a short set of checks in each, and score every finding by impact and effort. The output is a ranked 90-day roadmap with owners, not a deck of observations.
What a marketing audit should answer
A leak is any point where you pay for attention, leads or customers and then lose them downstream: ads that bring the wrong visitors, demo requests nobody answers for two days, first-time buyers who never hear from you again. Getting to those two answers means settling three questions:
- Which numbers can we trust? If you can’t answer this, every other finding is a guess.
- Where is the biggest constraint? Traffic, conversion, retention or execution capacity. Usually one dominates.
- What do we fix in the next 90 days, in what order, and who owns each item?
Gather this before you start
- Admin access to ad accounts, GA4, Google Tag Manager, Search Console, CRM, email platform and store backend
- 12 months of revenue and marketing spend by channel, by month
- Closed-won deals (B2B) or orders with customer IDs (ecommerce) for the same period
- A list of every marketing tool, agency and contractor you pay, with monthly cost
- The number the business is actually judged on this year
One note on order: before judging any channel, compare last month’s conversions in each ad platform with the CRM or store backend. A large gap makes every channel number provisional.
Positioning and messaging
Positioning leaks are hard to see from inside. The symptoms show up elsewhere: decent click-through rates with weak conversion, sales cycles full of “we need to think about it,” and discounting to close.
- You can state who it’s for, the problem it solves and why you beat the main alternative in one sentence each, and the homepage says the same
- The reasons customers give for buying (sales call notes, reviews, win/loss interviews) match the reasons your site leads with
- Your ideal customer profile is written down and reflected in ad targeting and outbound lists
- Ads, landing pages, sales deck and onboarding make the same promise
- If you put a competitor’s logo on your homepage, a customer would notice
The last check is blunt, and when I audit companies it fails more often than founders expect. Generic category language (“all-in-one platform,” “premium quality”) weakens every stage at once.
Channel mix and spend efficiency
Look for money that isn’t working and dependence you haven’t priced in. List each channel with its 12-month spend, the revenue or pipeline you can tie to it, its cost-per-result trend and a verdict: scale, hold, cut or test. Then check:
- Blended efficiency (MER or blended CAC) moves with platform-reported ROAS. If platform ROAS rises while blended numbers stay flat, platforms are likely claiming credit for sales that would have happened anyway
- Brand search is reported separately, so it doesn’t flatter paid search
- No campaign, ad set or keyword has spent for 90 days with zero conversions
- The last budget increase held its efficiency. If cost per result jumped after scaling, you’ve found diminishing returns
- No single channel produces most new customers without a plan for what happens if costs rise there
- Organic traffic is growing on pages that convert (product, pricing, comparisons), not only on blog posts
For deeper passes, use the Google Ads audit checklist for paid search and the technical SEO audit checklist for organic.
Tracking and data integrity
Broken tracking hides every other leak and turns budget decisions into guesses, so it tops the roadmap whenever it fails.
- Each ad platform’s reported conversions are within an agreed tolerance of the CRM or store backend for the same period
- GA4 key events are business outcomes (purchase, qualified lead, demo booked), not scrolls or page views
- Where you use the Meta Conversions API, browser and server events share an event ID so they’re deduplicated
- With EEA or UK traffic, your consent banner and Google Consent Mode actually pass consent choices to your tags
- For B2B, lead quality or deal stages are sent back to the ad platforms, so bidding optimizes for qualified leads rather than form fills
- A written UTM convention exists and is followed, and “direct” or “(not set)” isn’t swallowing paid traffic
For the detailed setup, from consent to server-side tagging, see how I implement tracking.
Website and conversion
Look for the pages and steps where paid traffic disappears.
- Conversion rate is broken down by landing page and device. A mobile rate far below desktop on the same page points to a page problem, not a traffic problem
- The top five paid landing pages match the promise in the ads that send traffic to them
- Someone has submitted a test lead and a test order this month, end to end: did it arrive, reach a person and trigger the follow-up?
- For B2B, you know how long a demo request waits for a human reply
- Top landing pages load quickly on a mid-range phone over mobile data
- Pricing is findable and understandable without a sales call, or there’s a deliberate reason it isn’t
The end-to-end test takes ten minutes and often finds bigger leaks than a redesign would: forms emailing a former employee, a failing payment method, thank-you pages that fire no conversion.
Retention and lifecycle
If new customer numbers are growing and revenue isn’t, the leak is usually here. Acquisition gets the budget and the meetings; retention gets whatever flows someone set up two years ago.
- Repeat purchase rate (ecommerce) or net revenue retention (SaaS) is tracked by cohort, not only as one overall number
- Core automated flows are live and reviewed in the last six months. Ecommerce: welcome, abandoned cart, post-purchase, win-back. SaaS: onboarding, activation, expansion and renewal
- SPF, DKIM and DMARC are in place, as Gmail, Yahoo and other major mailbox providers require of bulk senders
- You know what share of revenue comes from existing customers, and which way it’s trending
- Churn and cancellation reasons are captured in a structured field, not only in support tickets
- CRM lifecycle stages are defined, and duplicates are cleaned on a schedule
Team, tools and process
Some leaks are organizational. Good plans stall because nobody owns them, and agencies optimize for the metric in their own dashboard.
- Every channel and every key flow has one named owner
- Agencies and freelancers work to written briefs with targets tied to the business metric
- Every tool has a named user and a reason to exist; overlapping tools and unused seats get cut
- A weekly meeting exists where numbers lead to decisions, not status updates
- Tests are logged with a hypothesis, a result and what changed because of it
- Budget moves are decided on blended numbers, not on each platform’s own reporting
If most of these fail, fix the operating rhythm before adding channels.
Scoring findings into a prioritized roadmap
A full audit usually produces a few dozen findings. Fixing them all at once is how audits die in a shared folder. Score each one instead:
- Impact (1-5): expected effect on the primary metric this quarter
- Effort (1-5): time, cost and dependencies. A 1 is an afternoon; a 5 is a multi-month project that needs engineering
- Priority score: Impact × (6 − Effort), which gives a range of 1 to 25
An example sheet, with made-up findings:
| Finding | Area | Impact | Effort | Score | Owner | Window |
|---|---|---|---|---|---|---|
| Demo requests land in a shared inbox with no alert | Conversion | 5 | 1 | 25 | Sales lead | Days 1-30 |
| Meta purchases double-counted, no deduplication | Tracking | 4 | 2 | 16 | Analytics | Days 1-30 |
| No post-purchase or win-back flow | Retention | 4 | 2 | 16 | Lifecycle | Days 1-30 |
| Brand and non-brand keywords mixed in one campaign | Channels | 3 | 1 | 15 | Paid media | Days 1-30 |
| Homepage doesn’t reflect why customers buy | Positioning | 5 | 4 | 10 | Founder and marketing | Days 31-60 |
| Three overlapping analytics tools | Team and tools | 1 | 1 | 5 | Operations | Backlog |
Then turn the sheet into 90 days:
- Days 1-30: anything that blocks trustworthy measurement, whatever its score, plus high-impact, low-effort fixes
- Days 31-60: two or three high-impact, higher-effort bets, each with a success metric and a kill threshold agreed upfront
- Days 61-90: review results, scale or cut, and re-score what’s left
- Backlog: low-impact items. Most never deserve to leave it
Score with two or three people who know the business, and write the reasoning next to each number, or it gets re-argued in every meeting. It’s the same sequence I use when building a growth strategy: the audit sets the order of work.
What you can self-audit, and where outside eyes pay off
You can run most of this yourself: the channel spend review, the test lead and order, the flow inventory, the tool list, the zero-conversion spend. These checks are mechanical and still find real money.
Outside eyes pay off in three places. Positioning, because you’re too close to hear how your own copy sounds. Tracking reconciliation, because the errors stay invisible until someone checks them against the backend. And anywhere the people auditing built the thing being audited, since flaws in your own setup are the hardest to see.
Get it built
If you’d rather have this audit run for you, that’s what my fixed-price Growth Audit is for: $1,500, ending in a scored roadmap rather than a list of observations, and credited if we continue working together. See pricing for what comes after, or get in touch to talk through where your growth is leaking.