How to Assess What AI Is Already Doing in Your Marketing System
"There is surely nothing quite so useless as doing with great efficiency what should not be done at all."
~ Peter Drucker, management consultant and author; from "Managing for Business Effectiveness", Harvard Business Review (May 1963)
Ask most founders whether AI is part of their marketing and you will get one of two answers. Some say "not really, we've played with ChatGPT a bit." Others say "yes, our marketing exec uses it for posts." Both answers usually turn out to be incomplete.
Picture an accountancy firm with 12 staff. The founder thinks AI is a side experiment. In reality, the marketing executive drafts every LinkedIn post with it, the freelance designer uses it for image concepts, the email platform has started suggesting subject lines, and someone in the team has been pasting client questions into a free AI tool to speed up replies. None of it was decided. All of it is shaping how the firm shows up.
Drucker wrote that decades before AI tools existed, but it fits. AI makes marketing work faster. Whether that work should be happening at all, and in the way it is, only becomes clear once you can see it.
So how do you find out? You assess what AI is already doing in your marketing system by listing every place it is used, tracing what each use touches in the customer journey, checking what the AI is working from, reviewing who checks the output, and then deciding what to keep, fix, stop or formally own.
Is AI Already in Your Marketing Without You Knowing?
For most SMEs, the honest answer is yes. AI did not arrive through a strategy meeting. It arrived through the side door, one tool and one shortcut at a time.
Adoption has moved quickly. Thryv's 2025 small business survey found that AI use among companies with 10 to 100 employees rose from 47% to 68% in a single year, with marketing support among the most common uses. That is exactly the size of business most founder-led firms sit in.
What founders see is a much smaller slice. The 2025 global study on trust in AI by the University of Melbourne and KPMG found that 57% of employees say they hide their use of AI at work, and only 40% say their workplace has any policy or guidance on it. In a small team, that gap does not need bad intent. People simply assume a quick shortcut is not worth mentioning.
The result is a marketing system where AI is doing real work, on real client-facing material, with nobody accountable for how.
Where Is AI Already Hiding in Your Marketing?
Before you can assess anything, you need a complete list. In our experience, AI tends to show up in four places, and most founders only know about the first.
Tools the team chose on purpose like ChatGPT, Claude, Canva's AI features, a transcription tool for meetings.
AI switched on inside tools you already pay for. Many platforms now include AI by default rather than as an extra. Google, for example, built Gemini into its Workspace Business and Enterprise plans from January 2025, so drafting help appeared inside Gmail and Docs without anyone buying it. Your CRM, email platform and website builder may have done the same.
AI used by people outside the business. Freelancers, agencies and contractors often use AI to produce what they deliver to you. That is not a problem in itself, but you should know where it sits in their process.
Personal accounts used for work. This is the one most likely to carry risk. If a team member is pasting client details, pricing or internal documents into a free personal tool, that information has left your control.
A simple way to start is to ask each person who touches marketing one direct question: "Where do you use AI in your work for us, including anything small?" Make it clear there is no wrong answer.
How Do You Assess What AI Is Actually Doing?
Once you have the list, the assessment itself is a short, structured review. You are working out whether each use is helping the system or quietly pulling it off course.
Trace what each use touches
For every use on your list, note which part of the customer journey it affects. Is it shaping how prospects first find you, how enquiries get answered, what existing clients receive, or what the founder sees in a report? A tool drafting internal notes carries far less weight than one writing your website copy or replying to new leads.
Check what the AI is working from
This step matters most. AI produces output based on the brief it is given. If there is a documented ideal client, a clear positioning statement and a voice guide, the tool has something solid to work from. If there is nothing, the tool fills the gaps with the most generic version of your category, which is how a specialist firm ends up sounding like everyone else.
Check who reviews the output
For each use, name the person who checks the work before it goes out, and what they check it against. "The exec reads it over" is not the same as "the exec checks it against our messaging guide." If nobody reviews it, mark that clearly.
Look for risk, not just quality
Ask three questions. Is any client or confidential information going into tools the business does not control? Could any output state something inaccurate about your services, pricing or results? Is anything going out under your name that nobody has read?
Decide: keep, fix, stop or own
Every use should end the review with one decision and one named owner. Keep what is working and well briefed. Fix what is useful but working from a weak brief. Stop what carries risk without enough benefit. Own anything important enough to become a documented part of how your marketing runs.
The value is that, for the first time, AI in your marketing becomes something you can see and manage rather than something you hope is fine.
What Your Marketing System Needs to Do Better
A pattern tends to appear. The AI problems are rarely about the tools. They are about what the tools were never given.
Content sounds generic because there is no agreed positioning to write from. Enquiry replies drift off-message because nobody has documented the core message. Reports get produced, but nobody can say which numbers actually matter, because the growth priorities were never set. In each case, AI is not creating the gap. It is exposing it faster and at greater volume.
That is the same conclusion we reached in What AI Can & Can't Do in Marketing, AI scales whatever foundation it sits on. This also connects to the wider diagnosis in How to Identify Marketing Gaps in Your SME, where the most costly gaps are often the ones hardest to see from inside the business.
So the most useful output of an AI assessment is not a list of tools. It is a clear view of which strategic decisions your marketing has been missing, and which of those AI has been quietly guessing at on your behalf.
Where the Strategy First Plan Fits
If the assessment shows AI working from thin briefs, the fix is not a better prompt. It is a better foundation.
The Strategy First Plan is how we build that foundation. Over three meetings and 45 days, a Fractional CMO leads the research and the decisions that every other part of your marketing depends on: who your ideal client is, how you are positioned, what your core message says, where prospects are dropping out of your customer journey, and which growth priorities deserve the next 90 days. The founder's time is around three hours.
Those outputs then become the brief. When AI is configured later in the MarketingOS, it works from your documented voice, message and priorities instead of a generic default. That is what lets AI speed up your marketing without diluting it, and it means every AI use in your table has something real to be checked against.
Quick Answers
How long does it take to assess AI in a small marketing team? For a team of up to around 30 people, the inventory usually takes a week of short conversations, and the review itself can be done in a couple of focused sessions. The time goes into getting honest answers about where AI is used, not into the analysis.
Should we ban staff from using AI in marketing? A blanket ban tends to push AI use out of sight rather than stop it. A better approach is to make AI use visible, agree which tools are approved, set clear rules on what information never goes into them, and give each important use a documented brief and a named reviewer. That keeps the benefits while closing the risks.
What should AI in our marketing be working from? At minimum: a defined ideal client, a clear positioning statement, a documented core message and a short brand voice guide. Without these, AI fills the blanks with generic category language. With them, it can draft faster while still sounding like your business, and your team has something specific to check every output against.
An AI assessment shows you where your marketing has been guessing. A Strategy First Plan replaces those guesses with decisions your team and your tools can actually work from.
Not ready for a call yet? Take the Free Marketing Strategy Assessment and see where your marketing stands today.
If you liked this post, check out The SME's Guide to Building a Human-Led, AI-Powered Marketing System.
