AI in marketing: where it helps, where it hurts the brand, and how to build it into the team
AI speeds up the drafting station. The editor's desk decides whether that turns into better marketing.
BY STEFAN MARITZ··6 MIN READ
TL;DR
- AI in marketing works best on the production line around the creative work: research, briefs, first drafts, repurposing, and reporting. The judgement calls on positioning, voice, and what to publish stay with people.
- Budgets have moved faster than readiness. CMOs now allocate an average of 15.3% of marketing budgets to AI, while only 30% report mature AI readiness.
- Volume is the trap. 49% of US consumers believe generative AI has made content quality worse, and search engines reward original, people-first work regardless of how it was produced.
- AI has not reduced spending on people in marketing. Labour's share of marketing budgets rose from 21.9% in 2025 to 24.5% in 2026, and lack of internal AI expertise is the barrier marketing leaders cite most.
- Build a shared system before adding more tools: brand knowledge the AI can draw on, agreed workflows, and clear points where an editor signs off.
Marketing teams were among the first to pick up AI, and many are now living with the side effects. Output went up quickly, and many teams found that quality and brand voice slipped with it.
I have spent a decade building marketing teams and, more recently, the automations inside them. This guide covers where AI helps marketing, where it does damage, why more output rarely means better results, and how to build it into a team so the gains last.
How marketing teams are using AI in 2026
Investment is significant and still rising. Marketing leaders expect AI-driven automation of marketing work to more than double, from 16% in 2026 to 36% by 2028.
Readiness lags behind the spend. 70% of CMOs say their marketing processes are not mature enough to implement and scale AI, and lack of internal expertise is the most cited barrier to AI-driven efficiency, at 38%. That matches what we see. Most marketing teams we meet began with individuals using AI tools on their own, and still have no shared way of working the rest of the team can follow.
Where AI helps marketing most
AI earns its place in the parts of marketing that are repetitive, structured, and checkable. The table below covers the uses I would start with.
| Use | What AI does well | Where a person stays in the loop |
|---|---|---|
| Research and insight | Summarises customer feedback, reviews, and competitor pages quickly | Deciding which findings matter and what they mean for the plan |
| Briefs | Turns a messy request into a structured brief with audience, goal, and constraints | Approving the objective before any work starts |
| First drafts | Produces a usable starting point grounded in brand guidelines and examples | Editing for voice, accuracy, and a point of view |
| Repurposing | Turns one strong piece into email, social, and web versions | Checking each version still says something worth saying |
| Reporting | Assembles channel and campaign reports from analytics tools | Interpreting the numbers and deciding what changes |
| Search and AI visibility | Structures content so search engines and AI assistants can find and cite it | Making sure the substance is original enough to deserve citing |
Notice what sits in the right-hand column. Every row keeps a person on the decision, and that is where the value of a marketing team has always been.
Where AI does damage
The fastest way to lose trust with AI is to publish content that sounds like everyone else's. Generic prompts produce generic copy, and audiences are noticing: among Gen Z and millennial consumers in the US, the share who think generative AI has made content quality worse rises to 57%.
Search is no shortcut either. Google's guidance is to create helpful, reliable, people-first content, and it lists extensive automation to produce content on many topics as a warning sign. Using AI to support well-researched, well-edited work is fine. Using it to flood a site with thin pages is a strategy with a short shelf life.
Brand voice is the quieter casualty. When five people each prompt their own assistant with no shared guidelines, the brand starts to sound like five different companies. That is a systems problem, and it has a systems fix.
Why more output rarely means better results
Think of marketing as a production line: brief, draft, edit, approve, publish. AI makes the drafting station very fast. If editing and approval capacity stay the same, finished work piles up at the editor's desk, and either quality drops or the queue grows. Speeding up one station of a line without looking at the next one is a familiar mistake in operations, and many marketing teams are making it now.
The numbers point the same way. Labour's share of marketing budgets rose from 21.9% to 24.5% between 2025 and 2026, which suggests teams are investing in people and capability alongside the tools. AI can help a team run more experiments, faster. Deciding which experiments matter is still the job, and it is a strategy question.
So measure marketing AI by outcomes, the way you would measure any AI ROI: cost per asset, time from brief to publish, conversion, and the share of drafts that need heavy rework. Volume produced is an input.
How to build AI into a marketing team
- Write down the brand's voice properly. Tone, banned phrases, good and bad examples, and the positions the brand holds. This becomes the material every AI tool draws on, and it is the core of a company brain for marketing.
- Pick one workflow that takes the most time today, usually content production or reporting, and record how long it takes now.
- Redesign it end to end. Decide which steps AI drafts, which a person edits, and who signs off. Put the brand knowledge into the tool, so nobody has to paste guidelines into every prompt.
- Protect editing capacity. If drafting gets faster, move time into editing and approval, or the bottleneck simply moves.
- Agree the approved tools, so people stop paying for their own assistants on personal cards with company information inside them.
- Make it the default way of working. Update templates, briefs, and checklists so the new workflow is the easiest one to follow. Adoption sticks when the work itself changes, and fades when it depends on a few enthusiasts.
- Review at 90 days against the baseline, then move to the next workflow.
Frequently asked questions
How is AI used in marketing?
Mostly for research, briefs, first drafts, repurposing content across channels, and reporting. The strongest teams use AI to speed up the structured parts of the work and keep people responsible for positioning, voice, and what gets published.
Will AI-written content hurt our SEO?
Search engines judge whether content is helpful and original, however it was produced. The risk comes from publishing large volumes of thin, generic pages, and well-researched work edited by a person is a different matter.
Can AI keep to our brand voice?
It can get close when it has clear guidelines, good examples, and a person reviewing the output. Without shared guidelines, each person's assistant drifts in its own direction.
Does AI reduce the size of a marketing team?
The budget data so far points the other way, with labour's share of marketing budgets rising as AI spend grows. The work shifts towards editing, judgement, and running the system. We do not work on headcount-reduction programmes, and we help teams become more capable instead.
Where should a marketing team start with AI?
With one workflow that takes a lot of time today, such as content production or reporting. Write down the brand voice first, redesign that one workflow, measure it, and use what you learn for the next.
If you would like help building this into your own team, our AI consulting for marketing teams page explains how we work with marketing leaders.
WORK WITH PRAXES
Ready to move from reading about AI to getting your organisation ready for it? Start with the diagnostic.
Take the AI Readiness Diagnostic