AI in Marketing
AI in marketing: where it creates value, and where it's just activity.
Most AI advice in marketing is either hype or fear. Neither is useful. The questions worth asking are commercial ones: where does AI change a number that matters, and where is it just producing more activity that looks like progress. I get asked about this often enough that it's worth answering properly, organised the way the questions actually come up.
AI Strategy & Business Impact
How do we identify the best AI opportunities in our business?
Start with the business problem, not the technology. Look at where your team loses the most time, where customers hit friction, or where decisions are slowed by manual work. The right question isn't “where could we use AI?” It's “where would AI create a commercial result we could measure?”
How do we measure the ROI of AI investments?
Most businesses end up measuring how much AI they're using rather than what it's changed commercially. The measures that matter are cost reduced, decisions made faster, conversion improved, or new revenue created. If an initiative can't be tied to one of those, it isn't producing ROI, it's producing activity.
Where should we start with AI adoption?
With a business priority you already care about, not a tool you've heard about. Pick a cost that's too high, a process that's too slow, or a decision that takes too long, and test AI against that specific problem. Prove it works before you scale it.
Which AI initiatives will have the biggest commercial impact?
Rarely the ones being pitched to you hardest. Vendors sell what demos well, not necessarily what's most valuable for your business. Ranking initiatives by likely commercial impact before any budget is committed means investment follows evidence rather than whoever presented most recently.
How do we align AI investments with business strategy?
AI investment should follow your existing growth priorities, not set new ones. If you can't draw a straight line from an AI initiative to something you've already committed to as a business, it's a distraction wearing the language of progress.
How can AI improve our profitability?
Only when it's aimed at cost, conversion or the speed of decisions, not layered on top of what you already do. Any margin improvement needs to be measured against what would have happened anyway, otherwise you're spending more money to feel more modern.
How do we create an AI roadmap for our business?
A useful roadmap sequences initiatives by commercial impact and readiness, not novelty. Every item should tie to a specific business outcome and a credible way to prove it worked.
How do we avoid expensive AI projects that deliver little value?
Prove it small before you scale it. Decide how you'll measure success before the project starts, not once it's finished. Most expensive disappointments come from skipping a modest, well-measured pilot in favour of a rollout too large for anyone to properly assess.
How do you build an AI business case?
Anchor it to one specific commercial outcome and a credible way to measure whether it happened. A case built on efficiency claims and vendor benchmarks rarely survives contact with a CFO. One built on a clear before-and-after usually does.
How do you prioritise AI projects?
By expected commercial impact and how quickly you can prove it, not by which is easiest to build. The right sequence gives you a track record to point to when you're asking for the next round of investment.
What AI KPIs should executives track?
Commercial ones: cost saved, revenue lifted, time reinvested into higher-value work. Not usage metrics like how many tools have been rolled out. A KPI that doesn't tie back to the P&L is a vanity metric with better branding.
What does successful AI transformation look like?
Quieter than most people expect. A business that's become measurably more efficient and effective, with leadership able to explain exactly where and why. The loud version, heavy on announcements and light on evidence a year later, is usually the one that didn't work.
What's the difference between AI strategy and AI marketing advisory?
AI strategy is where AI fits your business priorities overall. AI marketing advisory is the applied version inside marketing: which activities benefit, how you measure it, and how it fits your existing customer and brand strategy. I work at that applied level, grounded in the broader commercial picture.
How do you future-proof a business against AI disruption?
Not by chasing every new release. By building the discipline of measuring what's actually working and staying close to where your customers get value, so you can adapt quickly whichever way the technology moves. That habit matters more than any specific tool you adopt today.
AI & Marketing Performance
Will AI replace our marketing team?
Not the marketers who are good at their jobs. But marketers who use AI well will increasingly outperform those who don't. The real opportunity is removing low-value work and giving your team more time for strategy, creativity and understanding customers.
How can AI improve marketing effectiveness?
By taking the guesswork out of decisions that used to rely on instinct: what content performs, which segments respond, where budget is being wasted. It adds the most value applied to measurement and decision-making, not just to producing more content faster. Faster output isn't effectiveness on its own.
How do we use AI without losing our brand voice?
Treat it as a drafting tool, not a decision-maker on tone. Businesses that keep their voice intact use AI to generate options and speed up production, while a person still owns the final call on what actually sounds like them.
What AI use cases create the fastest ROI?
Tasks with a clear before-and-after, like content production speed, campaign optimisation or response times, since the baseline is easy to establish. Judgment-heavy work like strategy or positioning takes longer to prove, and rushing it tends to cost more than it saves.
How do you integrate AI into your marketing strategy?
It should sharpen execution of a strategy you've already set, not become the strategy itself. If your AI initiatives aren't clearly in service of priorities you'd already committed to, you're building capability without a direction for it to go.
What is AI marketing advisory?
Independent guidance on where AI genuinely strengthens marketing performance, and where it doesn't, without an incentive to sell you a platform. It sits above the tools; the job is judgement about commercial impact, not implementation of any particular product.
AI Governance & Adoption
What governance do we need before adopting AI?
Less than most businesses assume, but more than none. Clarity on what data your AI tools can access, who reviews AI-generated output before it reaches a customer, and how you'll catch errors or bias. Governance should match the risk of the use case, not slow everything down to the pace of the riskiest one.
How do we introduce AI without disrupting the business?
Sequence it around your team's capacity to absorb change, not around what's technically possible. Businesses that adopt AI smoothly treat it as a change management exercise first and a technology rollout second.
What mistakes should businesses avoid when adopting AI?
Adopting tools before defining the problem. Skipping measurement. Rolling out faster than the team can absorb. Almost every expensive AI mistake I see is organisational, not technical; the technology itself usually works fine.
How do we know if we're ready for AI?
Readiness isn't about infrastructure. It's whether you have clear priorities, data clean enough to trust, and a way to measure results. Businesses that skip straight to tools before answering these tend to generate a lot of activity and not much value.
What should our first 90 days of AI adoption look like?
Pick one well-defined problem, decide how you'll measure success before you start, run a small pilot, and be honest about the result. Ninety days is enough time to know whether an approach is worth scaling. It isn't enough time to rescue a project that started without a clear problem to solve.