The Hiring Question That's Quietly Shifted
A year ago, agency job specs looked like this: "Must have hands-on experience with Midjourney and ChatGPT. Strong prompt engineering skills preferred. Familiarity with Adobe Firefly a plus."
You still see specs like that. They're becoming a tell.
Spark AI's 2026 report found that 65% of agency leaders now prioritise aptitude over specific technical AI skills when hiring. The shift is happening fast, and it tracks a real insight: in a landscape moving this quickly, the ability to learn holds more long-term value than any specific tool proficiency.
Tools change every six months. The person who could use last year's model at expert level is now competing with someone who just picked up the new one. Tool knowledge decays. Aptitude doesn't.
What Aptitude Actually Means
Aptitude is a fuzzy word, so let me be specific about what it looks like in hiring.
It's the candidate who walks into an interview and asks unprompted how your agency is thinking about AI. Not because they want to demonstrate knowledge. Because they're genuinely curious about how you've approached it.
It's the candidate who, when you describe a problem your team is facing, immediately starts thinking out loud about how they might build something to solve it. Not with specific tools — with structure. "You'd need a place to put the brand guidelines. Then the thing would have to check against them. Then someone would have to evaluate the output before it ships."
It's the candidate who tells you about a time they learned something complex on their own, without being asked, because it would make their work sharper.
None of that depends on which tools they've used before. All of it predicts how they'll adapt when the tools inevitably change.
Chris Murphy, MD of Tuncarp, Nailed It
One of the sharpest quotes in the Spark report comes from Chris Murphy, MD and Founder of Tuncarp — the production agency that restructured around AI and got hired as Publicis's content studio.
"You need people who are willing to go down rabbit holes. Curiosity is the real skill now."
That's the hire now. People who follow the thread. People who open the unfamiliar tool. People who ask "what happens if I do this differently" and have the patience to find out.
If you're interviewing creatives for an AI-forward team, the best question might not be "what tools have you used." It might be "tell me about a workflow you figured out on your own that your team didn't know you were using." The answer tells you everything about how they'll respond when the landscape shifts again in six months.
The Trap: Aptitude Without Structure
Here's where I have to give leaders the bad news.
Hiring for aptitude alone doesn't solve your AI capability problem. It creates the potential for solving it. Whether that potential actually converts depends on what you do after they start.
The Spark report is blunt about this: "Aptitude brought into the organisation will not sustain itself without the right environment. The curiosity and drive that made someone a strong hire can stall just as quickly if the structures to support them are not in place."
I've seen this inside agencies dozens of times. They hire a brilliant, curious junior. That person wants to push AI forward. They get met with: no approved tools, no protected time to experiment, no interest from leadership in what they're building, no recognition when it works.
Six months later they're less curious. Twelve months later they've taken that curiosity to a competitor that actually lets them use it.
Aptitude is a raw material. Leadership has to build the factory around it. Otherwise you're paying for potential that quietly evaporates.
What to Put in Place
Four things, in roughly this order.
Protected build time. If you hire curiosity, you have to give it somewhere to go. Two hours a week minimum, named in the calendar, defended against client overrun. I wrote an entire piece on this — it's not a luxury, it's the hiring promise you're honouring.
Approved tools on enterprise tiers. The tools your team actually uses, paid for by the agency, with the good versions. Not the free tier. Not personal accounts. The annual cost is less than one mid-level contractor. The signal it sends is massive.
Visible leadership attention. When someone builds something real — a custom GPT that saves the team five hours a week, an automated briefing workflow, a brand-voice checker — the leadership team should know what it does, by name. A monthly show-and-tell is the cheapest possible way to make this happen.
Promotion criteria that reflect AI contribution. If AI capability is important, it should show up in performance conversations. If building workflows matters, it should factor into promotions. The Spark data is clear: when AI adoption isn't rewarded, it stays a side project. When it's tied to career progression, it becomes a priority.
The Interview Shift
Three changes worth making to how you hire, based on this.
Change the job spec. Stop listing specific tools as requirements. List outcomes and ways of working — "builds workflows that scale across the team," "shows initiative in learning new tools," "demonstrates ability to evaluate AI output critically."
Change the interview. Less "what tools have you used." More "how do you approach something you've never done before." Watch how they handle ambiguity. That's the signal.
Change the trial. If you're doing trial projects, make the trial about solving a real agency problem — not about demonstrating a specific tool skill. The candidate who builds a scrappy working solution with whatever tools they pick is a better hire than the one who executes flawlessly on a tool they already mastered.
The Through-Line
The reason this matters so much right now is that the kinds of skills that separate an average AI user from an exceptional one are not static. They're moving. And the people who will define what excellent looks like in two years are the ones who can keep learning when the ground shifts.
Hire those people. Then build the structure around them that lets their curiosity compound into organisational capability.
Either half of that alone won't get you there. Together they're the advantage.