Every CMO I talk to thinks their team is ‘using AI.’ Most of those teams are using AI badly, on the wrong tasks, with no shared playbook. Here’s the 20-minute diagnostic that surfaces the real gap and the spend you’re losing because of it.
Most marketing teams have surface-level AI use (writing captions, brainstorming subject lines) and zero structural use (campaign analysis, audience research, lifecycle automation). The gap is invisible in dashboards but costs a typical 20-person team £50k+ a year in lost productivity. Diagnose it in 20 minutes, then decide whether to train, hire, or buy your way out.
When I ask CMOs whether their team uses AI, the answer is almost always “oh yes, we’re all in on it.” When I ask what specifically, the answer narrows: “social captions, mostly. Some brainstorming. The content team uses it for first drafts.”
That’s the gap. Your team is using AI for the easiest, lowest-leverage 10% of marketing work. They’re not using it for the high-leverage tasks where it would actually save real time: campaign analysis, audience segmentation, lifecycle automation, ad creative testing, performance reporting, briefing creative agencies.
This pattern shows up across every marketing org I’ve audited in the last 18 months. Forrester’s 2026 marketing AI maturity research puts roughly 70% of marketing teams in what they call “tactical adoption” (surface-level use on low-stakes tasks) and only 15% in “strategic adoption” (AI integrated into measurement, planning, and execution). [1]
This is a 5-question survey you send your marketing team. Anonymous responses, takes each person 4 minutes. The results give you a clear picture of where the gap is and which roles are pulling their weight on AI usage.
Multiple choice: ChatGPT, Claude, Gemini, Copilot, Perplexity, an AI feature inside another tool (HubSpot, Salesforce, Canva, Adobe), other (specify), none.
What it reveals: the breadth of AI tool exposure. If most answers are “none” or “ChatGPT only,” you have a foundational adoption problem before you have a skills problem.
Open text. Force specifics: “social posts” is not specific enough, “writing 5 LinkedIn captions for the product launch” is.
What it reveals: the actual depth of use. If 80% of responses are content drafting and brainstorming, your team is in tactical adoption. If you see audience analysis, campaign post-mortems, brief writing, segmentation, or lifecycle optimisation, you have someone operating strategically and that person is a multiplier you didn’t know you had.
Multiple choice: 0 hours, 1-2 hours, 3-5 hours, 6-10 hours, more than 10 hours.
What it reveals: the realised ROI per role. Top performers using AI well report 5-10 hours saved weekly. People reporting 0-2 hours either aren’t using it or are using it on tasks that didn’t actually save time.
Open text. This is the most valuable question. It surfaces the high-leverage tasks people sense AI should help with but can’t access.
What it reveals: the highest-ROI training topics for your team. If 5 people independently say “I wish AI could help me analyse campaign performance,” you have a clear training priority.
1 = avoid it, 5 = use it for almost everything possible.
What it reveals: the confidence distribution. The gap between the top quartile and bottom quartile predicts how lopsided your team’s output is becoming.
Once you have 10+ responses, look for these patterns:
If 80%+ of Question 2 responses are content drafting or brainstorming, your team is in tactical adoption. They’re using AI as a typing assistant. The real wins are not happening. The fix is structured training on 3-4 high-leverage marketing workflows (campaign analysis, brief writing, audience research, performance reporting), not more “intro to ChatGPT” sessions.
If Question 3 responses cluster bimodally (a few people reporting 8+ hours saved and the rest reporting 0-2), you have a leverage problem. Two power users are getting big wins; the rest of the team is invisible to them. The fix: have the power users teach a 60-minute internal session, then run a 4-week cohort to bring the bottom quartile up. The gap between top and bottom shrinks fast when you structure the knowledge transfer.
Question 4 is your training roadmap, delivered by your team for free. Cluster the answers. The top 3 most-wished-for tasks are your priority list for the next quarter of upskilling. If the top wish is “campaign post-mortems,” that’s the workshop you need to run, not another session on prompt basics.
If a meaningful chunk (more than 25%) reports confidence of 1-2 AND 0 hours saved, those people aren’t going to teach themselves out of it. They’ll quietly fall further behind every quarter unless someone designs the path for them. This is where formal training matters most.
This is the conversation that gets training budget approved. Walk your CEO or CFO through this math.
Assume a 20-person marketing team. Fully-loaded average cost: £75,000 per head per year. Top-quartile AI users save 8 hours a week. Bottom-quartile users save 1 hour a week. Median user saves 3.
If you could move the median from 3 hours to 6 hours saved weekly across 20 people, that’s 60 extra productive hours per week. At an average rate of £37/hour fully loaded, that’s £2,220 in weekly productivity recovered, or roughly £111,000 a year. That’s net new capacity without hiring.
The cost of a 4-week structured AI training program for a 20-person team typically runs £15-£40k depending on depth. So the payback period on real training is usually under 2 months, and the run-rate annual benefit is 3-7x the investment. [2]
This math is conservative. It doesn’t count the second-order benefits: faster campaign turnaround, better personalisation, fewer external agency hours, lower turnover from people who feel their skills are growing. Those compound on top.
If your diagnostic shows the team is willing but unskilled (low confidence, high “wish I could” responses), structured training is the highest-ROI move. The team already knows your brand, your customers, your campaigns. Giving them AI capability layered on top is faster than hiring net-new people.
Look for cohort-based programs that work on your actual marketing tasks rather than generic AI courses. Our corporate workshops for marketing teams are designed around this exact problem: real campaigns, real prompts, real productivity gains by week 4. We’ve run these for marketing teams from 8 to 80 people.
If your team is small and the leadership doesn’t have time to drive a training program, one strategic hire can shift everyone. Look for a marketing operations or growth hire with deep AI fluency. They become the in-house champion who teaches the rest of the team in 1:1s and team meetings. Lower upfront cost than training the whole team, but takes longer to land.
For specific tactical tasks (ad creative, email subject lines, social copy), AI-native tools like Jasper, Anyword, or Mutiny abstract away the skill required. Buy these for the workflows where you don’t need your team to learn the underlying capability. Don’t expect them to close strategic gaps like campaign analysis or planning. Tools augment skill; they don’t replace it.
Don’t run another generic “intro to ChatGPT” lunch and learn. You’ve probably done one already. They produce a moment of enthusiasm and zero follow-through. Specifically for the reasons covered in why most corporate AI training fails, single events don’t change behaviour. Your diagnostic results will give you the data to demand something better.
Pair this with our pieces on measuring AI ROI for the broader budget-defence framework, and AI skills in 2026 job descriptions for the talent-market context.
4-5 minutes is the right length. If it takes longer, the response rate drops sharply and the data gets worse. The five-question format hits the sweet spot of useful insight without survey fatigue. Send it on a Tuesday morning, close it Friday afternoon.
Yes, especially for Questions 3 and 5 (hours saved and confidence). People underreport usage when they think their manager will see it tied to their name. Anonymous responses give you the real picture and the team feels safer being honest.
It’s still useful but you’ll need qualitative follow-up. With 5 people, the survey gives you a directional read; have a 15-minute 1:1 with each person to dig into their Question 2 and 4 answers. Treat it as a structured conversation tool rather than a statistical exercise.
Tools handle specific tactical tasks well (writing ad copy, generating subject lines) but don’t build the underlying skill. If you only need tactical lift, tools are faster and cheaper. If you need your team to think strategically with AI (analysis, planning, decision support), training is the only path. Most marketing teams need both.
Three filters: do they design around your specific marketing workflows or run a generic curriculum, do they include 4+ weeks of practice with real coaching or just lectures, and do they measure adoption at 30/60/90 days. If the answer to all three is yes, they’re serious. If they sell one-day workshops with no follow-up, they’re selling an event.
About this guide
This article was written by Hina Mian, co-founder of Future Factors AI, drawing on 10+ years in marketing strategy and direct work auditing AI adoption across marketing teams from 5 to 200 people, plus Forrester and PwC research on marketing AI maturity in 2026.
Sources
Weekly tips. Real examples. Practical help for busy professionals.
We care about your data, check out our privacy policy.
