‘ve spent 13+ years building marketing and growth functions from scratch — most recently as Head of Growth at a CRM agency, where I built a department’s entire operating structure before AI tools became standard kit. So I’ve watched the shift from a slightly unusual vantage point: not as someone who arrived in marketing ops because of AI, but as someone who built processes the old way first, and is now watching AI quietly rewire them.
Here’s what I’m actually seeing — not the hype-cycle version, the working-marketer version.
AI Has Already Taken Over the Boring Parts of Paid Media
I’ve managed seven-figure paid budgets across markets like Lenovo EMEA and Motorola EMEA, and the single biggest shift I’ve witnessed in performance marketing is this: bid management isn’t really a human task anymore. Platforms like Google’s Performance Max and Meta’s Advantage+ now handle the kind of granular, real-time optimisation that used to eat entire days of an account manager’s week. AI-driven PPC bid management can reduce wasted ad spend by around 37% and increase ad ROI by roughly 50%, and that tracks with what I’ve seen anecdotally across accounts I’ve overseen — the waste that used to come from slow, manual bid adjustments is largely gone.
What’s interesting is that this hasn’t reduced the need for strategic oversight — it’s changed what that oversight looks like. The quality of inputs — creative assets, product feeds, landing page structure — now drives performance more than manual bid strategy does. In other words, the job shifted from “manage the bids” to “manage the inputs the algorithm is learning from.” That’s a meaningfully different skill, and it’s one a lot of experienced marketers haven’t fully adjusted to yet.
Personalisation at Scale Is Real Now — Not Just a Slide in a Pitch Deck
Early in my career, “personalisation” usually meant segmenting an email list into three or four buckets and hoping for the best. That’s no longer where the bar sits. AI-powered personalisation can improve conversion rates by as much as 202%, and the majority of consumers now actively prefer personalised experiences over generic ones.
Having led growth strategy across 20+ client accounts spanning SaaS, eCommerce, and B2B, I’ve seen first-hand how much variance there is in how “ready” different organisations actually are for this. The brands seeing real gains aren’t the ones with the flashiest AI tool — they’re the ones with clean, well-organised customer data and a genuine understanding of their customer journey. AI personalisation is only as good as the data foundation underneath it, and that foundation is still very much a human responsibility.
The Productivity Gain Is the Part Nobody Argues With
Whatever scepticism still exists about AI in marketing, the time-saving case is no longer up for debate. Marketers using AI save an average of 6 to 13 hours a week, with senior practitioners saving the most — and that broadly matches my own experience leading teams through this transition. The biggest unlock isn’t that AI does the strategic thinking; it’s that it removes the friction around reporting, first-draft content, campaign setup, and the dozens of small repetitive tasks that quietly consume a marketing leader’s week.
That time has to go somewhere, though. The teams I’ve seen benefit most are the ones who’ve been deliberate about redirecting those reclaimed hours into strategy, creative judgement, and client relationships — not just absorbing them as quiet downtime.
Where I Think Human Judgement Still Wins — And Will For a While
This is the part I feel most strongly about, having built departments and managed brand reputations directly. Roughly 30% of marketers see generative AI as a real risk to brand safety, and a similar share remain cautious about the accuracy of AI-generated content — and from where I sit, that caution is earned, not paranoid.
Having led campaigns for brands like Coca-Cola, Pampers, and the NHS — where tone, sensitivity, and trust matter enormously — I don’t think AI is close to replacing the judgement calls that actually protect a brand’s reputation. It’s excellent at the first draft, the data crunch, the pattern it would’ve taken a human days to spot. It’s not yet trustworthy with nuance, cultural context, or knowing when a campaign needs to be pulled because it simply feels wrong in a way no dataset would flag.
My Honest Take
I don’t think AI is replacing marketing leadership — I think it’s raising the bar for what marketing leadership actually needs to look like. The marketers who’ll struggle aren’t the ones using AI “too much” or “too little” — it’s the ones who never developed strong strategic judgement in the first place, because that’s precisely the skill AI can’t replicate, and increasingly, the only thing left for humans to genuinely own.
The teams I’d want to build now look less like a stack of specialists each running one channel, and more like a smaller group of sharp strategic operators, directing AI tools rather than competing with them.