How to unearth strategy magic
(... and resist becoming the turkey that votes for Christmas)
Over the past few months we had the privilege to chat with a group of APG award-winning planners on a project partnering with the excellent Account Planning Group.
We chatted to nine authors of award winning papers from the 2025 APG Awards, with a loose conversation guide, but with a centre of gravity around: What mix of research actually powered the work?
Across the conversations, spanning everything from Müller Rice to Dulux to McDonald’s, one thing came through consistently:
AI and Big Data are now table stakes.
But they’re not where the magic comes from.
AI is powerful. But it’s a mirror.
No one we spoke to dismissed AI. Quite the opposite. It’s fast. Useful. Increasingly embedded in the day-to-day reality of planning.
But it has (current) limitations and reflects what already exists. It spots patterns. It summarises. It optimises, and that can be a limitation for a hungry irrational planner’s mind:
As Yolanda from BBDO put it, talking about her Libresse paper:
“I think AI is just a mirror of society… it isn’t able to invent and create something new… It would have just told us, ‘make it blue’.”
Sometimes the best work doesn’t come from pattern recognition, it comes from breaking patterns and logic.
Data tells you what. People tell you why.
Big Data brings reassurance. It gives scale and confidence to an increasingly risk-averse industry and client base. A sense that “60 million people can’t be wrong” (to paraphrase Neil Godber).
But again, it has its flaws.
Across the conversations, there was a shared view that data can point you to the problem but that it rarely unlocks the answer.
There’s a sense that the ‘unlock’ only happens when you get closer to people.
Stuart, talking about his V&A paper said:
“The data showed what the problem was [young people aren’t interested], but the human-led research explained why [it doesn’t feel like a place for me]”
When you understand not just what people do (which the data might tell us), but why they do it, and how it feels to them, you can elevate your work.
You can’t feel data.
Good enough is easy. Magic isn’t.
There was a clear tension running through all the work - the fact that AI and data can get you to a “good enough” answer quickly. And in an industry under pressure to deliver faster, that’s tempting.
With so many planners under ever-increasing pressure and in what I call ‘output mode’ - relying on quicker tools that are ‘good enough’ is becoming ever more important.
But “good enough” doesn’t win APG awards, and probably won’t win you the pitch.
The standout strategies - the ones that land emotionally, culturally and commercially often come from places that feel illogical, unexpected or “dumb” (to quote Matt from VCCP) - and we found that those strategies weren’t powered by AI and Big Data alone - but by a combination of AI, Big Data AND human-led research.
Matt from VCCP said of his Müller strategy:
“I don’t think AI would have ever spotted that daft culture connection of Declan Rice and Müller Rice.”
The difference is texture
Again and again, planners talked about moments they picked up human texture. Not dashboards. Not charts. Moments.
Standing in someone’s living room, noticing what’s on the shelves.
Hearing how something is said, not just what is said.
Spending time where customers hang out.
The small, human details that don’t show up in data. The texture of reality.
And it’s in that texture that the work improves and takes leaps it wouldn’t have taken otherwise.
Craig talked about not what was said, but how it was said for his Poppy’s strategy:
“The caller paused, then asked, “do you do normal funerals?”… It was the way the funeral director was reporting that conversation and the look on her face. [A transcript] wouldn’t have captured the body language of the people or the feeling in the room. It’s as much to do with the way that the funeral director was reporting that conversation and the look on her face as she talked about it.”
The power of n=1
One of the more uncomfortable truths that we found was that a lot of breakthrough thinking in the papers didn’t come from large datasets.
They came from a single moment - often from a single person. A sentence. A reaction. A pause.
Something that made the authors stop and think: There’s something in that.
But acting on that requires confidence, because in an industry that embraces ‘the data told me so’ as a security blanket, “n=1” doesn’t look convincing in a deck.
Tass - author of the McDonald’s award-winning paper - expressed this when he said:
“‘Have you validated it quantitatively?’ You don’t need to. You don’t need to know that 70% of people steal fries… You just need to know that it’s true. And it resonates.”
So the role of the planner, and the wider agency team, becomes critical to not just find the insight, but to back it and sell it.
How to work with human-led research: Reality first, Brand second.
A consistent theme from the winning papers was how they worked with human-led research: The best work didn’t start with trying to validate an answer toward the end of a strategic process, it started with people early in the process.
Not with a tightly defined brief designed to prove a hypothesis, but with open exploration about customers of a brand, prospects of a brand, or just… people.
A lot of the planners did this at the beginning of the process, and chatted with real humans without knowing what they were really looking for.
They trusted that something more interesting would emerge. And it did.
Stuart said of his V&A strategy:
“We were never using research to commission a finding. It wasn’t that we had this hunch that we were desperate to prove. We went into it open-mindedly.”
If you do it the other way and “commission a finding”, you’ll probably get one, but you might miss a stronger truth.
How to work with human-led research: ‘Shed the snobbery’ and get out!
This sounds obvious, but it’s not happening enough. I’m not convinced it ever has happened enough, but it feels like it’s happening a lot less than the past.
And in the current climate there’s lots of discussion about working from home versus working in the office… yet there’s very little thought about working where your customers or clients are! Spending time with them, observing them, chatting with them. Soaking up their behaviours, language, culture, nuance.
Every author we chatted to - across different levels, and in different sized agencies - talked about the need to get out of the office, to spend time in people’s homes, to walk around shops and to sit with clients inside their businesses.
To see and feel the world as it actually is, not as it appears through a screen.
Frances said of her clients, Very:
“Spending time at Very in Liverpool... Liverpudlians are fun. They are fabulous. They are maximum… big hair, big lashes, big nights out. That is what Liverpool is. And Very as a company, you really feel that... it’s flamboyant, fun people. They love a laugh. They love a party. And you really get a sense of that when you spend a bit of time up there.”
And in this, there’s a laziness - and an arrogance - that planners, agencies and brands are building strategies and creative platforms without ever involving real people.
We’re investing millions of pounds to try and change human behaviour, to get humans to buy a product, yet we’re becoming arrogant enough to think we can do it without chatting to real humans.
We expect them to ‘love’ our brands, and yet we can’t be bothered to even speak to them.
Shame on us.
The risk: becoming the turkey
I want to be clear that the findings aren’t an anti-AI argument.
AI is here. It’s useful - an essential tool in modern planning and it’s only going to get better.
But there’s a warning emerging from all of this.
If planning leans too heavily on AI and Big Data for “good enough” answers we are effectively training the system to replace us.
Every shortcut teaches the machine what we do.
And if what we produce is incremental, pattern-based, predictable work…
It becomes replicable by an “AI slot machine” run by Meta, Google or others, and negates the need for creative agencies or planners…
Max from VCCP said:
“Great work relies on planning to safeguard the strategic skills technology can’t replicate - taste, nerve, intuition and timing. And the only way to nurture them is through proximity to real people”
So what should planning do?
Jo Arden closed the APG session with an important set of reflections.
If planning wants to thrive, not just survive it needs to:
Protect time for getting close to real people
Make sure investment is seen as valuable
Be entrepreneurial in how it can access its audiences
Appreciate that strategy magic can come from small numbers of people
Provide a blanket of security and confidence for those leading human-led research
Because one of the enduring roles of planning hasn’t changed: To represent the voice of real people.
(Maybe I’ll ask Jo to pen a guest post to expand on those points above if we can find the time).
The Reality Advantage™
What this project reinforced is something we’ve been exploring for a while - The Reality Advantage™
The best strategies don’t come from choosing between AI, data, or human understanding - they come from combining them.
But with one important shift: Reality first. Brand second.
Because AI can mirror the world.
Data can measure the world.
But only real people can show you how it actually feels and that’s still where the magic is.