Marketing audit 2026: 40 questions and a checklist in one working day

Every time I start a new project, the first day of work is an audit. Not getting to know the team, not a strategic session - just an audit. Below is my exact 40 question checklist that I have been using on every project since 2019.

During this time, I conducted 30+ marketing audits - from small e-com projects with a budget of 100K ₽/month to B2B with a turnover of 200M ₽/year. Patterns repeat. About 80% of the problems are the same, just on a different scale. At the end of the article I discuss the most common findings.

80% of the projects I work on spend money on advertising with broken analytics. They don't know what works.

1. Why an audit is needed (and who should do it)

An audit is not about a report. An audit is about a list of 5 problems that need to be fixed tomorrow.

Most of the “audits” that I have seen from the outside end in a thick 30-slide document with conclusions like “the quality of content needs to be improved” and “it is recommended to optimize advertising campaigns.” This is not an audit - it is an imitation of work. A good audit should lead to concrete changes after a week. If nothing had changed a month after the audit, it was useless.

Who should do it: an external specialist sees problems that the internal team has become accustomed to and has stopped noticing. I myself noticed things with clients that were “always there”: UTM tags on half of the campaigns, a form without confirmation of submission, managers responding to requests after 8 hours. The team did not consider this a problem. This is a problem. An internal checklist is good for quarterly monitoring - like a thermometer, not like a diagnosis.

The structure of my audit is 4 blocks of 10 questions. I allocate a fixed time for each block. Only 6–8 hours with ready access.

2. Block 1: analytics and data (10 questions)

I always start with analytics - because if the data is dirty, all the following conclusions about channels and funnel are useless. This is the foundation. In 24 out of 30 audits, a critical problem was already discovered here.

I write in detail about setting up a stack in separate article about marketing analytics. Here are only diagnostic questions.

1. Are GA4 and Yandex Metrica counters worth it? I check it using GTM Preview and the Tag Assistant extension. It occurs even on large projects: the counter is “available”, but only on the main page - after the redesign they forgot to distribute it to other pages.

2. Do you have goals set for key conversions? Not abstract goals of “page visit”, but events: filled out a form, clicked the order button, added to cart, made a purchase. Without this, you only see traffic, not results.

3. To what extent do analytics and CRM data diverge? I take the last 90 days: how many leads are in CRM, how many conversions are in GA4/Metrica. A discrepancy of up to 15% is normal (different attribution models, data delays). 20–30% is a reason to look into it. More than 30% is a critical problem, the data is unsuitable for decision making.

4. Are there UTM tags on all paid channels? I am uploading a report on sources/channels. If I see a large share of traffic in (direct)/(none) or (not set) during active advertising campaigns, it means that the tags are partially placed. A typical loss is 15–40% of attribution.

5. Is GTM used, or are the tags written into the code directly? Without GTM, every tag change requires a developer. This is not critical in itself, but it greatly slows down the work with analytics.

6. Is the dataLayer configured for key events? This is already an advanced level. Without dataLayer, you cannot transfer the purchase value, user ID, or order status to analytics. On e-com, this blocks the calculation of ROAS based on real data.

7. Are offline conversions connected to advertising accounts? For B2B and complex products with long cycles - critical. If a sale occurs through a call or meeting, the advertising office must know about this, otherwise autobidding is optimized for the form, and not for the actual sale.

8. Is there an attribution setup with a model other than Last Click? Last Click lies in most cases with multi-channel funnels. Read more about this in the article about channel attribution.

9. Is the dashboard set up with key metrics? It doesn’t have to be beautiful - Looker Studio or DataLens with data from GA4 and CRM is enough. If the data is only looked at “when something happened”, this is not analytics, but a post-facto investigation.

10. Who on the team is responsible for data quality? If “no one”, the data will degrade. UTM tags disappear after the next “quick launch”, goals stop working after the redesign, no one notices the discrepancy with CRM. We need at least one person who checks the analytics once a week.

3. Block 2: channels and budget (10 questions)

Here I’m not looking at campaign settings, but at the big picture: what channels are there, how much does a lead from each cost, is there any logic in budget allocation. A detailed audit of specific campaigns is a separate job for 2–3 days.

11. What is the CPL for each channel for the last 90 days? I compare it with unit economics: at what CPL the project makes money, at what CPL it loses. On most projects this table simply does not exist. The budget is divided “by feeling” or “by inertia”.

12. Are there channels where the CPL is higher than acceptable, but the budget continues to flow? In 18 out of 30 audits - yes. Most often this is a display ad or a content channel that “once worked.” Nobody stopped it because “what if it starts working now.”

13. What is the budget structure between channels? I record it as a percentage. Typical picture in the Russian Federation 2026: 70–80% — Yandex Direct, 10–15% — VK Ads, 5–10% — Telegram Ads. If the distribution is very different, I want to understand why. Sometimes there is a justification, sometimes it just happened historically.

14. Is there seasonality in expenses and is it taken into account when planning? Most projects are aware of their seasonality, but the budget is still the same month by month. During the busy season, there is a budget deficit and lost sales. When it’s cold, money burns to nothing.

15. How often is A/B testing of ads carried out? I heard the answer “never” in 12 out of 30 audits. Without testing, campaigns degrade: ads burn out, CTR drops, CPL grows - but this happens slowly and imperceptibly.

16. Is retargeting used and for which audiences? I check: is there retargeting at all, at what stage of the funnel, what is the lag between the first touch and retargeting. A typical mistake: retargeting only for “added to cart”, and catalog and longread visitors leave without a follow-up.

17. Are there look-alike audiences based on real customers? For 60% of projects - no, although there is a customer base and it is enough for uploading to Y. Auditorium or VK.

18. What percentage of the budget goes to brand requests? Brand traffic is the cheapest and highest converting. But if he is not in the office, competitors take over. I look at the share of brand expenses and compare it with the share of brand conversions.

19. Is there organic traffic, and how does it compare to paid traffic? The ratio of paid and organic traffic is an indirect indicator of the project’s dependence on the advertising budget. SEO and content marketing - in competitor analysis There is a section about organic search.

20. Is ROAS or DRR considered for each channel? For e-com this is mandatory. For lead generation, CPL and lead-to-sale conversion are considered. If the answer is “no,” the budget is spent blindly.

4. Block 3: funnel and conversions (10 questions)

This is the most painful block - because here you can see not only marketing problems, but also product, sales, and UX problems. Marketing brings traffic, but what happens to it next depends on dozens of factors outside the advertising account.

21. What is the conversion rate from click to lead? I count for each landing and each channel separately. The average conversion for a landing page in the Russian Federation is 2–5%, depending on the niche and traffic temperature. If it’s less than 1%, there’s most likely a problem with the landing page or the relevance of the traffic.

22. What is the conversion rate from lead to qualified lead? How many applications are actually suitable clients, and how many are non-targeted. If qualifications are low with a normal CPL, there is a problem in targeting or in the landing offer.

23. What is the conversion rate from qualified lead to sale? This is already a sales zone, but the marketer must know this figure. Without it, it is impossible to calculate unit economics. More details about calculations can be found in the article about unit economics.

24. How long does it take from the application to the manager’s first contact? This is the number one conversion killer in lead generation. On one project with a CPL of 1,400 ₽, the first call occurred in 6–8 hours. Sales conversion - 4%. After implementing the “first contact in 15 minutes” rule, the conversion increased to 9% - without changes in advertising.

25. Are there automatic follow-ups after an application? Email or messenger with application confirmation and next step. Absent in half of the projects - although it can be set up in an hour.

26. Do all landing pages work correctly on mobile devices? I check the main devices manually and through Webvisor. Typical problem: the form does not work properly on iOS, the button goes under the keyboard, the banner overlaps the content.

27. What is the loading speed of landing pages? Via Google PageSpeed Insights. Loading time of more than 3 seconds on mobile means losing 40–50% of users before they even saw the offer. On three projects, speed optimization reduced CPL by 15–25% without changes in advertising.

28. To what extent does the offer on the landing page coincide with the offer in the ad? A simple test: I take an ad, click - does what is promised in the ad match what I see on the landing page? Divergence = loss of trust = drop in conversion. Occurs in 30% of projects.

29. Are there repeat touches with those who didn’t buy? Email marketing, retargeting, “did not buy” segment in CRM. Most projects live by the principle “they brought traffic - either they bought it or they didn’t.” Repeated touches with warming content are +15–30% to the final conversion on long cycles.

30. Are returns and reasons for refusals tracked? For e-com - the share of returns by category and channel. For services - reasons for refusals in CRM. If it is not tracked, half of the information about the quality of traffic and product is simply lost.

5. Block 4: team and processes (10 questions)

This block takes the least time - 1 hour - but often explains why the previous three blocks look the way they do.

31. Who makes decisions on the marketing budget and based on what data? For 8 out of 30 audits the answer was: “the founder, according to the feelings.” This is not necessarily bad at the start, but with a budget of 500K ₽/month it’s already a problem.

32. Does the team have OKRs or KPIs with specific numbers for the quarter? “Increase sales” is not the goal. “CPL no higher than RUB 1,800, 150 qualified leads per month, sales conversion 8%” is the goal. The difference is that the latter can be monitored weekly and decisions can be made based on fact.

33. How often does the team look at analytics data? If the answer is “when things go wrong,” that’s reactive management. The weekly rhythm of working with data is the minimum for projects with a budget of 200K ₽/month.

34. Is there a content plan and editorial policy? For projects with content marketing or SMM - a basic planning tool. Lack of a content plan = chaotic publications “according to the mood” = no cumulative effect.

35. Are there regular retrospectives on advertising campaigns? We look not only at the current results, but also at what worked and didn’t work over the past period. Without this, the same mistakes are repeated from quarter to quarter.

36. Is there a CRM and is it fully used? I don’t mean “does there have an amoCRM account,” but do managers use all the fields, tags, and stages of the funnel. Typical picture: there is a CRM, but the data in it is only a contact and the “new” status. It's almost useless for marketing.

37. Are there demarcated areas of responsibility in marketing? Who is responsible for traffic, who is responsible for conversion on the site, who is responsible for CRM and the speed of processing leads. When “everyone is for everything” - no one in particular is for anything.

38. How do contractors report results? I check the format of reports from agencies and freelancers. If the report contains only “impressions, clicks, CTR” without CPL and sales conversion, the contractor is optimizing what is easy to show, and not what is important to the business.

39. Is competitor analysis carried out on a regular basis? At least once a quarter - their offers, positioning, advertising activities. Without this, you work in a vacuum. About the methods - in the article about competitor analysis.

40. Is there documentation on marketing processes? Instructions for launching campaigns, checklist for publishing content, rules for processing leads. Without this, every new employee or contractor starts from scratch.

6. What to do with the audit results

After going through all 40 questions, you have a list of problems. Usually there are 15–25 points. This is fine. A bad audit result is when 2-3 problems were found because they weren’t looking carefully.

The next step is prioritization. I use a simple matrix: impact on revenue or CPL (high/low) and complexity of fix (high/low). Anything with high impact and low complexity - do it immediately. Usually there are 3–5 tasks.

Of the 30+ audits I've done, the "immediate" list almost always included one of two things: either fixing analytics (UTM, goals, dataLayer) or speeding up lead processing. Both changes are made in 1-3 days and show results within the first two weeks.

The format of the final document: not a 40-slide presentation, but a table. String = problem. Columns: description, suspected cause, impact on metric, next step, responsible, deadline. All. One page.

If the audit is done for a client, I separately record 3 “quick wins” that can be realized in the first 2 weeks. The client sees the result, confidence in the process grows, and it’s easier to continue working.

7. Frequent finds

Statistics have been collected over 30+ audits. Here's what's most common, with frequency and approximate impact on CPL:

I would like to note separately: most often problems are clustered. If there are no UTM tags, there is most likely no normal attribution. If there are no CPLs for channels, most likely there are no OKRs either. Correcting one leads to several related ones.

One of the strangest-looking cases: a project with a CPL of 3,200 ₽ reduced it to 1,800 ₽ without changes in advertising campaigns - they simply corrected the UTM tags, and it turned out that half of the “direct” traffic was actually from an email newsletter. Email worked, but all conversions were attributed to paid advertising, which was then “optimized” to the detriment of email.

If you want to conduct an audit yourself using this checklist, but at some point you get stuck, write to Telegram or through form. Starting consultation - 0 ₽.

Related articles: KPI dashboard template, ROAS/ROMI calculator, setting up marketing analytics from scratch, unit economics for marketers, analysis of the contractor's report.