AI & the Future
Using AI, and using it well
Graeme Stiles is a Chief AI Officer, and his answer to how hard you should go in on AI is to do more of the thinking yourself. The two divides, and the discipline that crosses the second one.
Blog / AI & the Future
Darren Morris runs Lucky Hustle, a data-driven Joburg agency, and his first marketing trend for 2026 is deliberately worded. Not the use of AI, the acceptance of it.
"It was almost like we were in a space where agencies had to hide that they're using it — like something shameful. Marketers are now seeing that these tools are here to help us. Once we accept AI, this industry changes for the good forever."
The tell that acceptance has arrived is commercial rather than technical. Clients now ask whether the agency is using AI, because they want their budget used as efficiently as possible. We've watched the same shift from the client side, with teams choosing agencies partly on whether the spend works as hard as the tools allow. In 2024 an agency whispered about AI. In 2026 it leads the credentials deck with it.
Acceptance has a sharp edge most agencies haven't faced yet. If AI collapses the hours a deliverable takes, then charging by the hour is, in Darren's words, "going to end up shooting them in the foot."
"Marketing should be output-based. We deliver X, Y and Z based on your objectives. How we get there is completely up to us — and that's where we use all the tools 2026 has."
For anyone buying agency services, this is a procurement question worth asking at your next review: are we paying for hours that AI has already eliminated? For anyone selling services, agency or freelance, it's more urgent. Reprice around outputs and outcomes before your clients do the maths for you.
Three questions locate you honestly:
Declared. Could you list, to a client or your board, exactly where AI sits in your workflow, without embarrassment in either direction? The shame cycle has inverted, and hiding it now reads as being behind.
Priced. Does your pricing reflect outputs and outcomes, or hours that no longer exist? Model what your three most common deliverables cost you now versus a year ago, and sit with the difference.
Proven. Can you show one deliverable where the tools measurably improved cost, speed or performance? Not a demo, a result.
The warning that runs under all of this comes from Matt Brownell, at the Marketing Cafe: "AI is a race to the middle. It's removing barriers to entry, so anyone can become a marketer. What it's done is put pressure on really good marketers to become really great ones."
Yudhvir Ranchod of Peach Payments, an enthusiastic adopter whose team uses AI avatars to cut production costs exponentially, puts the corollary from experience: "As amazing as AI is, we've seen businesses use it in the most generic way. You had this great tool, but you misused it, so now it looks like everybody else." The technology that makes everyone competent makes distinctiveness the entire game.
Try this week: write your one-paragraph AI disclosure (what you use, for what, and what stays human) before a client or boss asks for it. Then, if you sell services, model your three most common deliverables on value rather than hours.
Sourced from recorded, on-the-record interviews on the South African Digital Marketing Podcast. Every practitioner quoted above is named in the text.
First published on the SA Marketers' Circle.
I'm Terence Mentor. These articles come from my conversations on the SA Digital Marketing Podcast. Through TM Digital, I look at what businesses spend on marketing and tell them plainly what's working.