How this started
This piece started with a simple internal question: "How much are you experimenting at the moment?"
We asked the team to pick one of three options: "Nope", "A bit", or "Yeah, loads", plus a word or two on what they were trying. The replies were honest, varied and very human, from people constantly testing new tools to others who commit deeply once something proves its value.
We paired those real-world anecdotes with wider research on experimentation and innovation, to understand not just what people are trying, but why it matters for how we work.
Why experiment
Experimentation is what separates teams that keep the lights on from teams that get better over time. Organisations facing constant change need continuous experimentation, not just fixed strategy, to stay ahead. Innovation and experimentation go together; without the willingness to test something new, genuine innovation does not happen.
This is not reserved for a dedicated "innovation team". Everyone, in product, marketing, design, operations and client service, needs the tools and permission to test and learn independently. Defaulting to "the way we have always done it" feels safer in the short term, but it is also how teams slowly fall behind.
Making room for it
The biggest barrier to experimentation is rarely willingness, it is time. A useful way to think about this is the 70/20/10 model, popularised by Google and inspired by 3M's earlier "15 percent time" policy that produced the Post-it Note:
Allocation - Focus
70% - Core, day to day work that keeps things running
20% - Adjacent experimentation, improving or extending existing tools and processes
10% - New exploration, with no immediate deliverable attached
Google's own experience is a useful reminder that the model only works if the time is genuinely protected. Well known products emerged from it, but usage of the full allowance was inconsistent, because core workload naturally expands to fill whatever time exists.
The same logic applies to production and delivery. A culture of experimentation here is rarely about big new ideas, it is about better systems: fewer manual handoffs, more repeatable templates, less time lost to human error on everyday tasks. The tools worth testing here are often unglamorous, plug-ins, scripts, small automations, but they quietly remove time and risk from the work that keeps a business running.
"We are always exploring new ways of working with clients, including project management tools and dashboard reporting, so that everyone involved can see and manage a project in one place. We have also restructured how briefing works across the agency and our clients, so that the right information is captured upfront, in the right order, before a project starts moving."
— Claudia Edelstein, Digital Account Director
This shows experimentation is not only about adopting new tools, it is also about actively redesigning the process behind the work. Restructuring the briefing process reduces drag and misunderstandings before a project even begins, and exploring more collaborative project tools gives clients better visibility and control over their own projects as they progress.
But the replies we got back made one thing clear straight away: experimenting well is not about doing more of it, or doing it constantly. It is about knowing what is worth trying, what is worth keeping, and when to stop looking. That idea runs through everything that follows, and it is the thread we want to pull on properly across this series.
Experimentation isn't one thing
The first thing our replies showed us: there is no single "right" way to experiment. Some people try everything, some want a tool shown to them once before they commit, others experiment mostly by paying close attention to the world rather than chasing new platforms.
"I am a visual, practical learner, so I get the most out of a tool once someone shows me it in action. Show me something new and I will use it. I have always been that way."
— Angela Brown, Director
"I do like to experiment with tools, be it AI or otherwise, but that creative mind exists in the real world, and seeing what is out there, good and bad, and living with it."
— Steve Donohue, Creative Lead
Neither approach is more valid than the other, they are just different routes to the same outcome: finding what genuinely improves the work. We go deeper into these different experimenting styles, including a fascinating account of what happens when constant testing becomes exhausting rather than exciting, in the next piece in this series.
What actually earns a place
Trying something is easy. The more interesting question our replies raised is what actually survives contact with real work, and what quietly gets dropped.
"Overlord, the Figma/After Effects integration, is still a bit sloppy with text formatting issues, and sometimes kicks in a tonne of layers, but is way better than exporting to Illustrator, layering manually and then importing to AE."
— James Adams, Film and Motion Designer
"I tried Canva AI through ChatGPT Business for a pitch deck. It promised a lot, and while the output wasn't quite ready to use as is, it was still a useful exercise in seeing what's possible, and I went back to my usual PowerPoint approach for that one."
— Angela Brown, Director
One tool stuck around despite its flaws, because it beat the alternative. One did not, and that is a perfectly good outcome too. The next piece in this series looks at what separates the tools people actually keep from the ones they quietly drop.
When do you stop experimenting
Perhaps the most surprising insight is that constant experimentation is not actually the goal. At some point, the aim shifts from testing everything to protecting focus, and knowing when you have found something worth committing to.
"My sister showed me a favourite product 25 years ago and I have stuck with it ever since, because once something proves itself, I commit to it."
— Angela Brown, Director
"So much is happening, not only in our sector but in business as a whole, and there is so much going on. I started to find that monitoring it all, testing it all, became more and more frustrating."
— Costas Michalia, Strategy and Innovation Director
That tension, between staying open to what's new and having the discipline to stop looking once something works, is arguably the most interesting question this whole exercise raised.
The thread through all of it
The same idea kept surfacing across every reply, from research to creative work to the tools running the business day to day: experimenting well is not about trying everything, it is about knowing what is worth trying, what is worth keeping, and when to stop. The next three pieces take each of those questions further.