What Did I Take Away From Miro Canvas 26 (London)?
Solutions to problems that were never problems to begin with... and stickers.
This week I spent the day at Miro’s Canvas 26 in London. Though I went for the tool, I also had three questions in mind as I went in:
When everyone can build 10x faster, are we actually going anywhere, or just going?
Where does a tool genuinely help a team decide, rather than create more for them to decide about?
And when the AI hoovers the whole organisation’s data into one place, does that make us clearer, or just busier?
Side Note: getting to a conference about humans and machines collaborating meant crossing a city half shut down by tube strikes. A very human, very analogue act of resistance to a change people did not choose. I am not sure anyone on stage clocked the irony, but it turned out to be the truest thing I saw all day. Hold that thought…
AI scaled the creation. It has not scaled the decision-making.
Founder and CEO Andrey Khusid opened on the gap he thinks is the single biggest opportunity in business right now: individuals can now be 10x with AI, while their companies are still stuck at 1x. The organisation simply cannot absorb the speed of its own people. He has a name for the friction, too, the “collaboration fracture”: we now juggle three kinds of collaboration (human to human, human to agent, and agent to agent) with no shared place where all three happen and stay visible to everyone.
CPTO Jeff Chow took the idea up a level, pitching the canvas as the “decisioning layer” for the agentic era. One slide framed the ambition as a flip, from 80% doing and 20% thinking toward 80% thinking and 20% doing. Lovely in theory. But buried in the same section was the sharpest line of the day: AI has scaled creation, and it has scaled output, but it has not scaled decisions/decision-making.
Sit with that as a Product Ops person, because it is our entire remit in a single sentence. We are not in the creation business. We are in the alignment, prioritisation and decision quality business. If the rest of the company is now producing ten times the artefacts and the machinery for choosing between them has not budged, the bottleneck does not vanish. It lands on us, and it gets heavier. More decisions made faster is not remotely the same thing as better decisions. Faster building mostly just buys you the capacity for faster bad calls.
This absolutely mirrors the same sentiment from MXP (Mixpanel) two weeks prior.
A shared surface is a genuinely good idea. A new source of truth is not.
The thing everyone came to see was the agentic canvas itself. Mark Boyes-Smith demoed voice Sidekicks (conversational agents you brief and then literally talk to) and Flows: visual, multi-step AI workflows that live on the board, that the whole team can see and run, and that reach out into your other tools. One Flow pulled data from external systems (Jira was the example), shaped it into something like a roadmap, and posted the update straight back into Slack. There was plenty more in this vein too: Code to Prototype turning a description into a working mockup on the board, Mermaid and Markdown support, connectors into Claude, ChatGPT and Copilot so context is not trapped in one tool.
Here is the part I genuinely liked. Most automation today is invisible, owned by one person, and dies quietly the moment that person leaves. A workflow that is multiplayer, visible, and built to be handed off is exactly the thing Product Ops spends its life trying to create and almost never gets the time to. The direction of travel is sensible: bring the intelligence to where people already work, rather than dragging them into yet another interface. If Flows lowers that bar, I am interested.
But…
Here is the part that gave me pause. Underneath Flows sits a quieter ambition: bring all your data onto the canvas. Pull from Jira, from your analytics, from your spreadsheets, and build your backlog right there on the board. My margin note was blunt: why Miro for this, over an actual source of truth?
It is the same instinct behind everything Antonia wrote in Stop Buying Tools You Can’t Defend:
A whiteboard is, by design, a place for thinking. It is loose, expressive, and deliberately low-fidelity. The moment you treat it as a system of record, you have lifted your data out of the thing that governs it, and signed up for drift, duplication and the slow rot of “which version is right?” Product Ops spends years trying to shrink the number of competing truths in a business. I am wary of anything whose growth depends on adding one more.
And, the Miro canvas is unstructured data. Taking tasks that ladder up into epics and reports and initiatives and OKRs and sliced and diced and analysed, out of the structure and into text boxes… I suddenly got a cold sweat thinking about unpicking this.
I will also gently confess that the voice demo lost me. Talking to a Sidekick looked slick, but I could not work out what it solves that typing does not, and the claim that vaguer, more general instructions somehow produce sharper output went by unexplained. Happy to be shown. Genuinely.
When building gets cheap, thinking gets expensive.
The theme that kept surfacing, on stage and in the networking spaces, was that the constraint has moved. When building is no longer the slow part, the quality of thinking becomes the slow part. Daniel Hulme, WPP’s Chief AI Officer, put a number on the backdrop that stuck with me: 78% of employees are already using AI at work, most of them with no sanctioned tools and no training. The technology is already in the building. What is missing is the judgment to aim it.
That is a people problem, not a tooling one, and it is exactly where I kept wishing the day had gone deeper. Change management earned a passing mention, and then the conversation hurried back to features, which is precisely backwards. And, look, this was the Miro expo, to show off their features for us to drool over and spend more money. I get it. And I cannot deny, the location, setup, a staggering number of Miro staff, the pomp and high-fiving on/off stage - was all very impressive! But I’m in the business of practical answers to the real problems I have, or am going to have.
AI hands us focus and time for deeper thinking, but only if we actually use it that way, and there is a real risk of decision fatigue when suddenly every decision feels possible. Without clear principles for which decisions even matter, more capacity to decide just manufactures more noise.
Martin Eriksson’s Decision Stack, which I reviewed recently, is a great tool here.
So did I find what I went for?
I went looking for the new tools, and to get inspiration for how I could be better using the tool overall with my teams - we are fairly basic whiteboard users to date. So, for our money, what can we get out of it.
But I came out with some new truths about the AI-obsessed SaaS industry… chief amongst them is that the traditional lanes platforms have remained in are now blurring, which is fantastic news for their CEOs, sales teams and shareholders, but adding much more confusion than ever before to those tasked with choosing the best products for the right purposes, clouded by the vibe-coding mantra ‘It can be whatever you want it to be’ that is yet to be truely proven, and even less sustainable.
I also got AI-generated, machine-printed, human-handed-out stickers. Truly human-to-tech collaboration.
Back with my Product Ops hat back on, what does this mean for us?
Tools have never been top of the priority list for Product Ops. In fact, we should always be making it one of the last things we change or promote. But we’re clearly in a world where tools are getting a lot of attention again, doing much more than ever before, handing over much of our busy work to them. There is excitement to be trying new tools again, whether home-made or off the shelf, with encouragement to be ‘trying them out’ all across the business landscape.
And yet, the responsibility to implement, maintain and scale those platforms still remains with the chosen few to ensure they are maintainable, scalable, secure, robust - all the things that drain the excitement from water-cooler conversations, and yet make them viable. And when we have platforms land-grabbing to replace a greater number of competitors or claim to be the right single source of truth, it makes the time we spend evaluating those potential solutions that much more vital. I’m already contending with so many vibe-coded solutions that sit as a support structure to core business needs. I built a good number of them!
And I am sure there will be those reading who are thinking ‘Lighten up, Graham, let teams explore, use whatever tools get the job done and ship faster!’. Please get in touch or leave a comment to dive into this more… because every fibre in my being says that thinking leads to some very expensive dead ends to unpick. I’ve seen it, I’ve been there, and it is really not pretty.
It is not just the technical unpicking either, but the behaviours and habits your teams have built up too that need unpicking, and if you have read much of my writing, you know change management is a whole new scale of problem to tackle.
So, your research evaluation and scrutiny of tools is as important as ever, as is your focus on the landscape of your business in 12 months+ living with your new relationship, not just the next few weeks honeymooning with the shiny new toys.
BUT, I want to balance this before I finish. I like Miro, its a great tool to make your ideas visual, collaborative, and at its heart, the new tools shown off at the show were in this vein still. The technology is impressive, and I love a shiny new toy as much as anyone. So I will be exploring Sidekicks and Flows. I have a really great Claude-powered sync with Airtable going great guns. And the Widget Creator, sitting right next to those Claude connectors, might just be the no-code home I have been hunting for to host the little internal tools my team keeps building.
I just struggle to get past this thought: Miro Canvas 26 sold the canvas as the cure. The longer I sat with it, the more I think the problem was never the canvas.
Graham
Did you attend the show? What did you come away with? Tell me your thoughts to all this.