How Miro's CEO builds an $18B company in the AI era

Executive overview

Building a product people love still comes down to solving a real problem and obsessing over product-market fit. AI has made software cheaper to build but hasn't changed the fundamentals — it has, however, made the pace of change so fast that planning beyond 12 months is no longer realistic.

Andrey Kusid, CEO of Miro, rebuilt his product's entire interface around AI collaboration, shifting from a manual whiteboard tool to a platform where teams move from idea to prototype in hours rather than months.

The core insight: brand, mission, and founder-market fit matter more now, not less — because AI has commoditised the build, and only products people genuinely trust will break through.

Starting out: problem first, scale second

• Miro started as a simple idea — bring a whiteboard into a browser — born from a personal pain point running a creative agency with remote clients.
• Early goal was not scale but break-even; ambition grew only after seeing clear signal that the product was working.
• First growth inflection came in 2015 (flash to HTML migration), then enterprise sales traction in 2018–19, then pandemic drove 5M to 50M users in 18 months.
• Growth levers in order: product virality and invite loops, SEO, then intentional marketing and sales layered on top.

Finding product-market fit

• Start with: is this a real problem, is it a big problem, is the market large enough?
• Run open-ended customer interviews to prove or disprove hypotheses — don't ask what people need, observe how they react to a prototype.
• For deep qualitative signal, 7–20 interviews is sufficient.
• Track whether failure rate sits around 30%: if every experiment succeeds, you are not pushing far enough; if most fail, your bets are too speculative.
• Distinguish between a wrong problem and a wrong solution — if you believe the problem is real, iterate hard on the solution before giving up.

Brand and trust in a commoditised market

• Renamed from Real Time Board to Miro deliberately: the goal was to move from descriptive name to love mark — something that activates a feeling, not just a function.
• Inspiration: artist Joan Miró, canvas as metaphor — aspirational and creative, not another enterprise tool.
• Three tiers of company naming: descriptive name → brand → love mark. Aim for love mark.
• With AI commoditising software quality, brand trust has become a primary differentiator.

Planning in the AI era

• Miro used three-year "painted picture" planning cycles — the 2022 picture was well-executed, but AI was not predicted.
• Now planning in 6-month committed windows; 12-month directional vision at most.
• You can't meaningfully predict beyond 12 months given the pace of LLM change, shifting customer behaviour, and uncertainty around build-vs-buy decisions.
• Key strategic question: where do you have permission to play, and more importantly, where do you have permission to win?
• Mission ("empower teams to create the next big thing") stays constant; tactics change constantly.

Reinventing Miro for AI collaboration

• Original Miro: people collaborate manually, each adding pieces to a shared canvas.
• New Miro AI Canvas: teams collaborate with AI as a participant — AI helps move from step to step in a workflow, not just assist individual tasks.
• Key distinction: individual AI productivity gains don't shorten a project if team coordination remains slow. Miro's bet is on multiplayer AI — team-level throughput.
• A workshop that previously produced sticky notes and took months to reach a prototype can now collapse into hours, ending with a deliverable ready for customer testing.
• The AI Canvas is a separate mode running alongside the core product; whether they merge is an open experiment.

Where Kusid sees the biggest AI opportunities

• Legal, marketing, and coding are leading verticals for reinvention.
• Coding specifically is shifting engineers toward context and alignment work rather than line-by-line writing.
• Marketing end-to-end — from content creation through channel optimisation to ROI — is being redesigned from scratch.
• Miro is trialling 30–40 AI startups internally through a fast-track experimentation programme.
• Consolidation across the AI landscape will happen fast — likely within 18–24 months. Move now or get absorbed.

Qualities for building in this environment

• Continuous curiosity: the builders who thrive are energised by ambiguity, not overwhelmed by it.
• Critical thinking: understanding the competitive landscape and where you can actually win is as important as building speed.
• Resilience: the pace of change will only increase — physical AI (agents in the real world) will add further disruption.
• Founder-market fit: know what kind of builder you are. Don't chase markets you aren't wired for.

Recommended resources

• High Growth Handbook — collected insights from CEOs who scaled fast; useful if you're not in a hub where you learn by osmosis.
• High Output Management — fundamentals for scaling teams; worth rereading as the business grows.
• Favourite AI tools: Granola (note-taking), Perplexity, ChatGPT, Claude — used interchangeably for different jobs.