The aha use case: why AI adoption sticks for some teams and dies in others

"The largest barrier to usage, and we see this globally, is finding your aha use case. It's about discovering the one thing that adds a lot of value — and from there, people start using these tools more frequently."

Her example was aimed at me: if I only used AI to write guest introductions I already enjoy writing, I'd conclude it adds nothing. Point it at the part I dread — "look at this person's LinkedIn and suggest questions" — and it becomes indispensable overnight. The adoption lesson for anyone rolling AI into a team: stop demonstrating what the tool can do, and start finding what each person hates doing. The aha is personal, or it doesn't happen.

High-empathy and low-empathy work

The distinction Imke borrows from the creative industry is the cleanest sorting tool in the whole AI conversation:

"Low-empathy tasks — generating the 96 formats you need to put one ad on social media — that's productivity; automate it. High-empathy tasks — looking at a customer persona and really feeling the need that underpins it — AI can be a thought partner, but not a replacement, precisely because of the empathy."

It rhymes with Michelle Geere's automation split (machines for optimisation, humans for headspace) and Matt Brownell's pyramid flip — three practitioners, three vocabularies, one model.

The finder

Run this for yourself, then each team member:

1. List the dreads — the five recurring tasks this person most avoids. 2. Sort by empathy — mark each low (mechanical, format, volume) or high (judgement, feeling, relationships). 3. Match the lowest first — pair the most-hated low-empathy task with a tool and prove the win within a week. That's the aha. 4. Then add the thought partner — for one high-empathy task, use AI as sparring partner only, with the human keeping the pen. 5. Tell the story — spread adoption by showing the benefit, not the technology. Imke's own craft: selling AI is storytelling about benefits, "not the terabytes of data underpinning the model." Nobody adopts a spec sheet.

One reassuring note she offered: contrary to the fear narrative, the research shows African users are broadly excited about what generative AI can do — clear-eyed that skills will change, but largely positive. The fear is louder in the discourse than in the data.

Try this week: find your own aha — one dreaded low-empathy task, one tool, one honest verdict. Then run the finder with your team in a single 45-minute session, so everyone leaves with one match to test.

Sourced from recorded, on-the-record interviews on the South African Digital Marketing Podcast. Every practitioner quoted above is named in the text.