Marketing Audit Checklist: 40 Things a Proper Audit Must Check (Free Template)
Most marketing audits are sales documents. They list 80 problems, give none a priority, and end with a quote for fixing all of them. A proper audit does the opposite: it scores what you have, ranks the fixes by cost and impact, and gives you a baseline to measure against in 90 days.
This is the checklist we would use on your business. It has 40 checks in eight areas, a 0 to 2 score for each, and a way to turn the result into a fix list. You can run it yourself in an afternoon. We give it away because the audit is not the hard part. Fixing what it finds is.
Quick answer
A marketing audit checks how a stranger finds you, understands you, contacts you and gets followed up, and how you measure all of it. Work through eight areas (offer, website, SEO, AI visibility, LinkedIn, capture, follow-up, measurement), score five checks in each from 0 to 2, and total out of 80. Fix the cheap, high-impact failures first. A good audit ends in a ranked list and a baseline, not a long report.
Eight areas, five checks each, 80 points. Your total is your baseline.
How to run this audit
- Block three hours. Open a spreadsheet with four columns: check, score (0, 1, 2), evidence, fix.
- Audit as a stranger. Use a private browser window, your phone and a different network if you can. Do not log in to anything.
- Score honestly. 0 = missing or broken. 1 = exists but weak. 2 = clearly good, with evidence you can point to.
- Write the evidence. A screenshot, a URL, a number. “Seems fine” is not evidence.
- Rank the fixes. Put each failure in a simple grid of effort against impact. Do the cheap, high-impact ones first.
The most common mistake is reading the checklist and nodding. Open the tabs and click.
Area 1: The offer (checks 1 to 5)
Answer first: audit the offer first, because every other channel sends people to it, and a weak offer wastes all of them.
If your marketing feels hard, check this before anything else. Often the real problem is the offer, and no amount of marketing will fix it.
| # | Check | 2 points when |
|---|---|---|
| 1 | Who it is for | A stranger can name the buyer from your homepage in five seconds |
| 2 | The outcome | You state a specific result, not “solutions” |
| 3 | Proof | A named result, number or quote appears next to the claim |
| 4 | The first step | A small, low-risk entry: audit, sample, pilot, short call |
| 5 | Reason to act now | There is a real cost of waiting or limited capacity |
Example. “We help UK manufacturers” scores 0 on check 1. “We help food packaging firms with 50 to 250 staff cut quote turnaround to one day” scores 2. The service has not changed. The sentence has.
Area 2: Website (checks 6 to 10)
Answer first: the site is a working machine, not a brochure: it has to load fast, read clearly on a phone and make the next step obvious.
| # | Check | How to test | 2 points when |
|---|---|---|---|
| 6 | Mobile clarity | Open the homepage on your phone | You know what they do and what to tap within five seconds |
| 7 | Speed | Search Console, Core Web Vitals report | Pages meet Google’s “good” thresholds: LCP 2.5 seconds or less, INP 200 milliseconds or less, CLS 0.1 or less |
| 8 | Dead ends | Tap every header, footer and call-to-action button | Every link goes somewhere real, no placeholder forms or empty links |
| 9 | One call to action per page | Look above the fold | One clear action on each key page |
| 10 | Page hygiene | View the source or use a crawler | No leftover template text, no wrong alt text, no duplicate page addresses |
What we usually find. Placeholder code left from a template (we have seen alt text about bed linen on a site that did not sell bed linen), a header “Contact” button pointing at nothing, and the same page live at two addresses with no canonical tag. To a person it looks fine. To a crawler it looks careless. A site that looks good is not necessarily a good site: machines read the code, not the design.
Common mistake: buying a redesign to fix a hygiene problem. Fix the hygiene first.
Area 3: Search (SEO) (checks 11 to 15)
Answer first: check that Google can find, index and understand your pages, and that you appear for the phrases buyers type.
| # | Check | How to test | 2 points when |
|---|---|---|---|
| 11 | Indexing | Search Console, Pages report | Every page you want found is indexed, and the “not indexed” list is explained |
| 12 | Own-name search | Google your business name | You are first, and the listing shows the right description, address and links |
| 13 | Keyword fit | Search Console, Performance report | Your pages appear for buyer phrases, not only your brand name |
| 14 | Titles and headings | Crawl or inspect source | Titles under 60 characters, one H1, headings that say what the section is |
| 15 | Authors and entity | Read your best articles | Named authors with bios, a clear About page, consistent business details |
Why check 15 matters. Expert content published under a generic company byline asks a search engine to trust a ghost. Put a real name, role and credentials on your best pages.
Area 4: AI visibility (checks 16 to 20)
Answer first: test what ChatGPT, Gemini, Claude and Perplexity say when a buyer asks your question, and whether they can read your site at all.
Ranking first on Google does not guarantee an AI tool recommends you. SEO decides which page you land on. It does not decide whether AI names you. Google says its AI features rely on normal search fundamentals: crawling allowed, internal links, good page experience. The rest of the work is trust and proof.
| # | Check | How to test | 2 points when |
|---|---|---|---|
| 16 | Citation test | Ask 10 buyer questions in each AI tool, with and without your location | You are named in several answers |
| 17 | Crawl access | Check robots.txt and any CDN or firewall rules | The crawlers you want are allowed, and key pages are not blocked |
| 18 | Structured proof | Read your site as a stranger | Credentials, research, data and case studies are in crawlable text, not only images or PDFs |
| 19 | One entity | Compare your name across site, LinkedIn, directories, Companies House | One consistent name, no old brand or second domain competing with you |
| 20 | Third-party proof | Search for reviews, directories, trade mentions | Recent reviews and mentions exist off your own site |
How to run check 16 properly. Write 10 questions a buyer would ask (“who are the best [service] providers for [buyer] in [place]”). Ask each in each tool. Log who is named, in what order, and what source is cited. Repeat every month. Dates matter: AI tools weigh when something was published, so old press is weak evidence.
What we find in audits. Sites that AI tools cannot read at all because of a page builder or plugin set-up, old rebranded sites that are still live and get read as a separate competitor, and calculators with no question-and-answer wrapper, so nothing citable exists.
For why this matters: Pew Research Center found people clicked a traditional result in 8% of Google visits when an AI summary appeared, against 15% when there was none. We cannot promise any tool will cite you. We can check that you give it every reason to.
Area 5: LinkedIn and personal brand (checks 21 to 25)
Answer first: for businesses that sell on trust, the founder’s profile does more than the company page, so audit the person.
People rarely affiliate with a company page. They buy from other people.
| # | Check | 2 points when |
|---|---|---|
| 21 | Headline | It says who you help and with what, in plain words |
| 22 | About section | Starts with the buyer’s problem, then proof, then how to reach you |
| 23 | Featured section | Holds your best case study, best guide and a booking link |
| 24 | Posting rhythm | Three to four posts a week, each standing alone, mostly teaching and stories |
| 25 | Conversation | You comment daily on posts by 10 to 20 people your buyers follow |
Example. A founder with a stock banner, 100 listed skills and no featured section scores low across the board. A one-hour profile rewrite lifts four of the five checks.
Common mistake: selling in the feed. Give value first. Let people get interested in you, then in the company.
Area 6: Capture (checks 26 to 30)
Answer first: audit what happens to a visitor who is not ready to buy today, because most are not.
Everything that is not connected to you capturing contact details is rented land. Ranking, ads and social reach all belong to someone else.
| # | Check | 2 points when |
|---|---|---|
| 26 | Lead magnet | A useful guide, calculator, checklist or sample trades for an email address |
| 27 | Form quality | Short form, clear promise, works on mobile, spam protection that does not block real people |
| 28 | Confirmation | The visitor sees a thank-you page and receives an email at once |
| 29 | Ways to enquire | There are at least two: form plus phone, WhatsApp or booking calendar, and all work |
| 30 | Compliance | A cookie banner and privacy notice that match current UK guidance from the ICO |
Test it. Submit your own form from a phone. Count the seconds until you get the confirmation. Check the alert reached your team.
Area 7: Follow-up (checks 31 to 35)
Answer first: audit the first reply and everything after it. Most leads are lost here, not before.
| # | Check | 2 points when |
|---|---|---|
| 31 | Speed of first reply | A named person replies within one working hour |
| 32 | Reply quality | The reply answers the question, proposes a next step and sounds like a person |
| 33 | Follow-up sequence | Stalled enquiries get at least three timed follow-ups |
| 34 | Review requests | Happy customers are asked automatically, with a direct link |
| 35 | Lead log | Every enquiry is logged with source, owner, time to reply and outcome |
How to test check 31. Send three enquiries over a week, at different times, from different email addresses. Write down who replied, when and how. The result tends to be uncomfortable, and useful.
Example. A review-request automation, a small sequence sent a few days after a job, takes an afternoon to set up. It works because most happy customers were simply never asked.
Area 8: Measurement (checks 36 to 40)
Answer first: if you cannot say how many people enquired, from where and what they were worth, you cannot improve anything.
| # | Check | 2 points when |
|---|---|---|
| 36 | Key events | Enquiries, calls, bookings and purchases are tracked as key events in Google Analytics 4 |
| 37 | Source attribution | You can see where each enquiry came from, and “how did you hear about us?” is asked on the call |
| 38 | Search Console | Verified, reviewed monthly, sitemaps submitted |
| 39 | Third-party view | At least one tool the person doing the work does not control |
| 40 | Revenue link | You know revenue per session or per enquiry, even roughly |
Google describes a key event as an action that is particularly important to the success of your business. Decide what yours are, mark them, then test that they fire. We tell clients we use third-party tools so we do not engineer our own tools to measure our work. You should expect the same from anyone you hire.
Turn the score into a fix list
Add up your points out of 80.
| Total | What it means | First move |
|---|---|---|
| 0 to 30 | The funnel leaks badly | Do Areas 1, 2, 7 and 8 before anything else |
| 31 to 55 | Solid base with gaps | Fix every 0 score, then lift 1s to 2s |
| 56 to 80 | Strong foundation | Spend on traffic, content and testing |
Every failed check goes somewhere on this grid. Start top left.
Then write the top-10 list: take every check scored 0, tag each Low, Medium or High effort, tag each High, Medium or Low impact, and put the Low-effort, High-impact ones first. A good audit hands you a list of ten, not a list of eighty.
What a good audit output looks like
A proper audit, from us or anyone else, should leave you with:
- A score out of a clear total, by area, so you can compare next quarter.
- Evidence for each finding: a screenshot, URL or number.
- A ranked fix list with effort and impact, and who should do each one.
- A baseline sheet of the numbers to beat in 90 days.
- A plain-English summary of one page.
- A view on what to ignore, because some things are not worth fixing yet.
Use this to judge any audit you are handed, including ours.
Red flags in agency audits
- No scores. Without a score you cannot track progress.
- No priorities. Eighty findings with equal weight is a to-do list for the agency’s invoice, not for you.
- Generic findings. “Improve your meta descriptions” appears in every audit ever written. Look for findings specific to your site, with the URL.
- Automated report with a logo. A tool output with a cover page is not an audit.
- The quote on page one. If the price arrives before the diagnosis, it is a pitch.
- Free audit, no explanation. Ask the person who did it to walk you through it. If they cannot, they did not do it.
- Fixing requires them. A good audit lets you do the cheap fixes yourself.
Questions people ask
What should a marketing audit include?
It should cover your offer, website, search visibility, AI visibility, social presence, lead capture, follow-up and measurement. Each area needs a score, evidence and a recommended fix, ending in a ranked list and a baseline you can measure against.
How long does a marketing audit take?
Running the 40 checks yourself takes about three hours. A professional audit of a small or mid-sized business typically takes a few days, mostly spent testing as a buyer and analysing data. Be wary of a one-hour automated report.
How often should I run a marketing audit?
Run the full checklist once a quarter and the quick tests (reply speed, AI answers, form test) monthly. Re-scoring the same 40 checks each quarter shows whether the work is moving the numbers.
Can I audit my own marketing?
Yes, and you should. You know your business best. The risk is going easy on yourself, so require evidence for every score, test as a stranger would, and ask a colleague or outside partner to check any 2s that matter.
What is the difference between an SEO audit and a marketing audit?
An SEO audit covers how well search engines can find and rank your pages. A marketing audit is wider: it includes the offer, AI visibility, social, lead capture, follow-up and measurement, so it shows where enquiries are lost after the visit as well as before it.
Sources
- Google, “Core Web Vitals”, web.dev, 2025
- Google, “Define key events”, Google Analytics help, 2025
- Google Search Central, “AI features and your website”, 2025
- Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results”, 2025 (browsing data from 900 US adults, March 2025)
Want us to do this for you?
If you would rather someone else ran the 40 checks and handed you a ranked fix list, we are happy to. Get in touch and we will tell you what we find. You might also like our guides to AI search optimisation, website fixes and how to choose a marketing agency.
Work with us