What's good management all about? Predictability and plans.
What's good innovation all about? Surprise and improvisation.
They operate at wildly different tempos, with different rules, trying to avoid totally different risks. You can't do both at the same time.
Unless you get good at blowing Uncertainty Bubbles.
An Uncertainty Bubble is a temporary space where the normal rules can be suspended. Juuust long enough for something new to form and become ready for reintegration.
These shimmering metaphorical spheroids have become utterly foundational for our work, and we have a hunch you'll find them handy too.
So today we're sharing an example of an Uncertainty Bubble we blew last year and how that played out. And over our next few emails, we'll go deeper on how to build them – and why most innovation attempts fail without them.
Were you forwarded this email? Subscribe for more.
Noticing we needed an Uncertainty Bubble
Last year we were asked to help a client with an app. This was a training app for executives that had clear, unique value. However – the team had struggled to get adoption for years, despite four rebuilds and lots of sales activity. The situation reached crisis point when one of their clients demanded some money back.
We dug in and quickly pattern-matched to an unlikely corollary: dance teaching.
When we started teaching Lindy Hop, we assumed that if you explained enough theory to students upfront, they'd be able to *think* their way into doing the dance right.
Guess what? This never worked. (What a surprise.)
What we do now—and what actually works—is to get people moving first. Before we say a single word in our dance classes, we put on a tune and get everyone to simply walk around the room in time with the music. Once everyone's in sync, we add rhythms, walk backwards, go double-speed or half-speed. We build, vary and stretch the physical experience first. The explanations come later, when students can connect the concepts to what their body already knows.
Well, client's training app was making the same theory-first mistake. The app had multiple books'-worth of theory about breathing and cardiology loaded into the onboarding journey. Then users were given a highly configurable but highly complicated tool and left to go and figure it out for themselves.
We did meet a handful of "true believers" in the app. The tiny fraction who had persisted, read the material, figured things out for themselves, designed their own practice, and achieved remarkable results. These people are inspiring because they are so rare (and they're brilliant to come across from a research perspective too, because they're the pathfinders).
Meanwhile, almost everyone else who'd ever tried the app had opened it exactly once, got frustrated or confused, and given up.
The app's core learning loop needed rethinking. We needed to get people moving first and hide the theory till later (or even save it exclusively for the devoted enthusiasts). We were looking for the equivalent of "walk in time with the music".
But—and here's the key point that will bring us on to Uncertainty Bubbles in a moment—we also realised that we wouldn't be able to rethink it fast enough if we went at the client's usual tempo.
We had dozens of directions to explore, and our first attempt to trial a new version of the app had taken over two weeks. We had to plan it out, get feedback from the CEO, battle with the app content system for a few days, test it with a few people, consider where and how it had failed, book a meeting for a week later, ...
Like in most established organisations, the default tempo was sustainable for maintenance and tweaking, but an order of magnitude too slow for figuring out anything totally new. The organisational conditions (or constraints, if that's your preferred jargon) were good for careful project management, but stifling for innovation.
That's where an Uncertainty Bubble comes in. You don't try to persuade the whole system to adopt the conditions needed for innovation. You create a temporary pocket within the existing system where you work at different tempo, operate in a different way, and where you can play with new frames.
Because when an organisation is really stuck, it always seems to come down to one limitation: which directions and data can even be considered within their current frame? When trying harder in the current frame stops working, getting unstuck requires finding a new frame.
When you find the right reframe, the stuckness melts away. A reframe can happen through a forcing function such as an existential crisis – often a self-inflicted one.
We prefer to create the conditions for the reframe on purpose. Before everything's on fire.
How we blew this Uncertainty Bubble
The hidden art of the innovator is both in "blowing" an Uncertainty Bubble when needed, and in reintegrating what gets created by the Uncertainty Bubble. We'll get to the second part later.
First though, how did we blow this particular Uncertainty Bubble?
During a client meeting, one of the team happened to mention a "challenge month" they'd run internally a couple of years back, when most of the staff had used the app every day and had "quite enjoyed it".
That was it: the opportunity, right there on a plate. This "challenge month" concept was already acceptable to the organisation. And the daily challenge structure would give us 30-odd cycles of learning and iteration.
It was 29th June, so we put together an email inviting everyone in the company to join the July Challenge. No overthinking it. No building anything yet. Just words on a screen.
To the 20-odd people who signed up, this looked like a planned, official programme. It appeared structured, purposeful, valuable. It had a clear start and end. Exactly the sort of things you'd want clients to see. The outside of the bubble looks safe and predictable to the wider system.
But to those of us who were running the programme behind the scenes? It would give us the scaffolding we needed to improvise and reframe daily.
Of course then we had to figure out how to deliver it.
The infrastructure was not glamorous. After considering a few of the options that were available at short notice, we settled on:
- the app as it was, for its biofeedback mechanism
- a daily Google Form to deliver instructions for that day and gather feedback as people worked through the exercises
- a WhatsApp group alongside email to share the links and handle questions and tech support
In practice, this meant we were on the hook to create a new Google Form each day, guiding participants through an exercise or two and then getting responses to both structured and open-ended questions.
The participants got to enjoy doing their daily challenge as if it was all planned out in advance. And we got near-real-time signals on what was landing and what wasn't, which meant that we were actually adapting and adjusting our plans continually as we went along.
Daily signals, daily changes, one big gorgeous mess
Early on in the month, we spontaneously designed an exercise where participants stated their intentions and planned their own practice, which worked surprisingly well. And we designed "fun" exercises for participants to do over the weekend ... which tanked. We also realised we had to invent ways to get people back into practicing when their habit had been derailed by a challenge that turned out to be too hard. As new ideas occurred to us, we incorporated exercises and feedback questions that would test those new ideas.
Honestly ... the process was what many would call a mess. But it was a productive mess. A mess that got us our desired iteration speed and much more besides.
You wouldn't want to run something like this over the long term or at scale. Nor could you run something in this way if the exercises or feedback questions had to go through any kind of formal review process or using official systems.
The improvisation with non-standard tools. The pressure. The need to respond to feedback in near-real-time. These properties are what you set up the conditions inside an Uncertainty Bubble to force, because these properties are what enable humans to generate that vital reframe that's needed to unlock progress.
By the end of week three in the July Challenge, the reframe we needed had started to stabilise. To cut a long story short: it became clear that the app by itself might never be enough. The structure, accountability and responsiveness that you get from a facilitated cohort was the value.
Now, writing that out, it kinda sounds facepalm-level obvious. But this is the power of Uncertainty Bubbles.
In the old frame that was held by the CEO and the rest of the organisation, the app was an app. A self-contained thing that had to work on its own. The concept of turning the app into a human-led cohort wasn't defensible – such an option wasn't even on the table. The experience of the July Challenge subtly laid new conditions that enabled a reframe.
The reframe isn't enough by itself, but it shifted the conditions, dissolving the stuckness and opening new directions for progress.
Closing the Uncertainty Bubble – you can't just pop 'n' go
Uncertainty Bubbles need to create rapid feedback loops with external reality ("how does the app actually work for customers?"), but they also need to create feedback loops that sense internal narratives and explore possible reframes ("the app is a self-contained thing ... or is it?").
One potential failure mode we faced? If we'd found a new working model for the app that couldn't integrate into the existing organisational narratives, and so would get ignored or orphaned.
Another failure mode: if we'd hooked the business on an exciting new narrative for the app but it ended up being another failed rebuild.
In short, you need to plan for closure and reintegration.
So we had to sense which reframings could work BOTH for customers and for the wider business in order to successfully close the bubble and reintegrate the new frame. Some of this work can be made explicit, but an awful lot is implicit and experiential. It really helped that the founders took part in the July Challenge.
In terms of explicit reintegration following the July Challenge, we also presented the journey at an all-hands, wrote a strategy document with Pivot Triggers baked in, and took the cohort model to market as a paid pilot with four real clients. This scaffolding enabled continued iteration and signal gathering, but now operating more comfortably at the tempo of the wider org and within official systems.
Following the paid client cohorts, the CEO concluded that the value of the offering was in the human touch. He came to his own realisation that the paid monthly subscription for the app was now incoherent, and he started planning for repeated training bursts and levels of progression. The app itself would become a free measurement tool to accompany the paid training.
The reframe made a radically different business model obvious. It also shifted the app from Annoying Problem to Energising Opportunity.
The full reintegration process took nine months. The new frame is now embedded.
Contrast this with an example of Uncertainty Bubbles that are, more often than not, left to wither and die: hackathons. If you've ever taken part in one, you'll probably know that the part of the job that gets missed is figuring out how to reintegrate those ideas after the party is over. It doesn't happen by magic.
Tom & Corissa x
