Failure is not a detour on the road to building great products, it is the road itself. That is the honest truth Panay recently articulated, and it is worth sitting with. When you are building something new, especially in a space as fast-moving as AI-native tools, you will encounter failures that range from minor bugs to catastrophic missteps. The question is not whether you will fail, but whether you will recover and learn. Panay's point is that recovery is always possible, and the learning is where the real product emerges.
This matters for anyone building or adopting technology today. Consider the trajectory of companies like Ramp crosses $2B in revenue as 85,000 customers adopt its AI platform. Ramp did not cross $2 billion in revenue by getting everything right on the first try. It iterated, stumbled, and refined its AI platform based on real-world usage, and 85,000 customers are the proof. Or look at Figma opens its platform to agents and expands code access in beta. Figma's decision to open its platform to agents and expand code access is a direct result of learning what designers and developers actually needed, not what the company assumed they wanted. Both stories reinforce Panay's insight: failure is only a dead end if you refuse to treat it as data.
Our take is straightforward. The best product teams do not fear failure. They build systems to capture it, analyze it, and turn it into fuel. That means creating environments where mistakes are surfaced quickly, without blame, and where the path from failure to fix is short. For spreadsheet users who feel stuck in legacy tools, this is especially relevant. The tools that will transform your workflow are the ones built by teams that have already failed, learned, and adapted. You do not want the product that has never broken, you want the one that broke, got fixed, and emerged stronger.
What this means in practice is that you should evaluate tools not just on their feature lists, but on their willingness to evolve. Ask yourself: Does this company treat user frustration as a signal to improve, or as an edge case to ignore? The products that will empower you tomorrow are the ones that have already survived their worst failures today. Panay's point is not a feel-good platitude. It is a practical filter for deciding where to invest your time and trust. The specific consequence to watch is how quickly a product team responds when something goes wrong. That speed, not the absence of error, is the true measure of confidence.