Latest / The CTO Podcast with Fexingo: Technical Leadership, Architecture, and Engineering Org / How Stripe Built Engineering Trust Through Incident Transparency
Transcript
- Lucas: When I think about the most misunderstood lever in technical leadership, it’s not code quality or hiring velocity. It’s how you talk when things break. Most companies treat incidents like security breaches, something to be contained and buried. But one specific payments platform decided to do the exact opposite, treating every outage as a public relations asset rather than a liability. Luna: That sounds counterintuitive until you realize that silence creates suspicion. If you don't tell people what went wrong, they assume you're hiding something worse. Lucas: Exactly. And the data supports this shift in mindset. We are looking at November third twenty twenty six, and the market is still rewarding operational maturity over raw feature speed. The company in question built what they call an architecture contract, but the real innovation was in their postmortem process. They stopped writing internal reports for managers and started writing detailed, public-facing narratives for customers. Luna: So they were essentially gamifying honesty? Making engineers compete to find the root cause rather than compete to avoid blame? Lucas: Precisely. They eliminated the performance review component of incident handling. You cannot have a blameless culture if admitting a mistake hurts your bonus or your promotion track record. They decoupled accountability from punishment. Accountability became about fixing the system, not finding the person who clicked the wrong button. Luna: That requires a massive cultural shift. Did developers push back on being so transparent? I imagine some feared legal exposure or just felt exposed. Lucas: They did push back initially, but the leadership team protected the process with ironclad policies. No names in the report. No references to individual performance. Just the timeline, the trigger, the detection gap, and the fix. They also published a monthly summary of all incidents, showing not just the big ones, but the small failures that would have gone unnoticed otherwise. Luna: Publishing the small failures seems risky. Wouldn't that make the system look fragile to enterprise clients who prefer perfection? Lucas: It makes the system look honest. Enterprise clients are sophisticated. They know software breaks. What they fear is a cover-up. By publishing the small stuff, they proved that their detection systems were working. If you only report outages, you might be missing minor degradations. Showing the full spectrum builds credibility. Luna: I saw a report last week where a different fintech tried this but got bogged down in legal review. How did they keep the turnaround time fast enough to matter? Lucas: They created a template that was legally pre-approved. The structure was fixed: what happened, why it happened, how we fixed it, how we prevent recurrence. Engineers filled in the blanks. Legal reviewed the template, not every single report. This reduced the review cycle from days to hours. Luna: That is a brilliant operational hack. It turns a bottleneck into a pipeline. So the metric that changed wasn't uptime, it was time to trust. Lucas: time to trust is the new latency. If a customer experiences an issue and receives a clear, empathetic, technical explanation within twenty four hours, their retention probability stays high. If they get a vague email three days later, they start shopping for alternatives. The speed of the apology matters as much as the speed of the fix. Luna: It reminds me of how airlines handle delays. The best carriers give you a reason, even if the reason is bad weather. The worst give you a generic 'mechanical issue' and leave you guessing. Lucas: Right. Ambiguity breeds anxiety. Specificity breeds calm. When you tell a customer exactly which server failed and why the backup didn't catch it, you respect their intelligence. You invite them into the engineering reality rather than shielding them from it. Luna: But doesn't that expose too much vulnerability? Competitors could use these reports to map out your weaknesses. Lucas: Competitors are already mapping your weaknesses through your API documentation and support forums. Hiding the truth doesn't hide the weakness; it just hides your competence in solving it. A public report shows you have solved the problem. That is the only defense that works. Luna: So the strategy is: show the scar, but show the healing. Make the recovery part of the product story. Lucas: Exactly. And this ties directly into the broader market environment in November twenty twenty six. Companies that are cutting costs are often cutting their incident response teams first. That is a short-term gain with long-term reputational damage. The winners are keeping those teams fully staffed because they know trust is a hard asset. Luna: It’s interesting that you frame trust as an asset rather than a soft skill. Most CFOs still see it as overhead. Lucas: They should see it as insurance. Every time you publish a good postmortem, you are depositing goodwill into the customer relationship bank account. When the next crisis hits, you have the balance to draw from. If you haven't been making deposits, you bounce immediately. Luna: I wonder if this approach scales to smaller startups. Do they have the volume of incidents to justify the overhead? Lucas: Startups benefit even more because they are building the habit from day one. If you wait until you have ten thousand users to start being transparent, it looks suspicious. If you do it with ten users, it looks like curiosity. They write shorter reports, but the principle is identical. Luna: Shorter reports mean less friction for the engineer. It’s about removing the ritualistic aspect of documentation. Lucas: Yes. Documentation should serve the reader, not the auditor. If the customer is the primary audience, the language changes. Less jargon, more context. More empathy, less defensiveness. It forces the engineering lead to think like a communicator. Luna: That’s a huge skill gap. Most senior engineers are great at debugging but terrible at explaining the debug to a non-technical stakeholder. Lucas: And that is why this has to be a leadership priority. You can’t expect engineers to naturally pivot to customer-facing communication without training. The best teams I’ve seen pair a senior engineer with a product manager for the drafting phase. Two brains, one voice. Luna: Pairing engineering with product is the secret sauce. It bridges the technical truth with the user impact. Lucas: It ensures the report answers the right questions. Engineers ask 'what broke.' Customers ask 'why does this matter to me.' The merged document answers both. Luna: So the output is a hybrid artifact. Part technical log, part customer service letter. Lucas: Part of both. And ultimately, it becomes a knowledge base entry that prevents the same mistake twice. The transparency is the marketing; the prevention is the value. Luna: I love that distinction. Marketing isn't adding noise; it's reducing uncertainty. Lucas: Uncertainty is the enemy of adoption. People adopt tools they understand and trust. When you pull back the curtain, you demystify the black box. Luna: Before we dive deeper into the metrics, I want to check in with our listeners. If these conversations about building resilient systems are useful for what you're running... Lucas: we deliberately don't run ads on these episodes. If you want to support that choice and keep the signal clean, the link is buy me a coffee dot com slash fexingo — that's F. E. X. I. N. G. O. It really helps us keep producing deep dives like this without corporate interference. Luna: Yeah, and it keeps the focus on what matters to you, the listener, rather than what advertisers want to sell. Lucas: Back to the metrics then. Once you publish these reports, what data points do you track to measure success? It’s not just about fewer outages. Luna: It’s about the sentiment around the outages. Do customers complain less after a transparent report than after a silent one? Lucas: The data shows a forty percent reduction in support tickets following a well-written postmortem compared to a standard outage notice. The tickets drop because the answer is already there. The customer reads the report and says, 'Oh, I see what happened,' instead of calling to ask. Luna: That’s immediate operational efficiency. You’re defusing the ticket before it arrives. Lucas: And you’re freeing up your support team to handle complex issues rather than repetitive explanations. It compounds over time. The better your library of postmortems, the less time you spend on reactive communication. Luna: So the investment is upfront labor, but the payoff is ongoing leverage. It’s a classic compounding asset. Lucas: Exactly. And here is the kicker. These reports become SEO gold. Prospective customers search for ' downtime' or ' reliability'. If your top results are helpful, honest analyses, you win the sale before the demo. Luna: Inbound marketing through honesty. That feels almost illegal how effective it is. Lucas: It feels illegal because everyone else is lying by omission. When you tell the truth, you stand out. It’s a competitive moat built on integrity. Luna: I’m curious about the internal morale side. Does publishing these reports make the team feel defensive or proud? Lucas: Proud, once the fear subsides. Engineers want to solve hard problems. Being able to say 'here is a hard problem we had, and here is how we mastered it' is deeply satisfying. It turns a negative event into a badge of competence. Luna: It reframes failure as learning. That is the core of a growth mindset. Lucas: And it attracts better talent. Top engineers want to work for companies that learn. They don't want to join a place where mistakes are punished. They want to join a place where mistakes are documented and solved. Luna: So the recruitment funnel improves as a side effect of the incident process. It’s a holistic cultural loop. Lucas: It is. You fix the system, you build trust, you attract talent, you improve the system further. It’s a virtuous cycle. Luna: What’s the biggest barrier for a company trying to start this tomorrow? Is it the writing, or the courage? Lucas: The courage. Writing is easy. Publishing is hard. Leadership has to publicly endorse the transparency and shield the team from backlash. Without that protection, the initiative dies in week two. Luna: So it starts at the top. The CEO has to be willing to take the hit for the sake of the long game. Lucas: Absolutely. If the CEO is worried about stock price dips on news days, they will kill the transparency program. You need leaders who are confident enough to let the chips fall. Luna: Confidence in the brand’s resilience. Belief that the customers are smarter than the headlines. Lucas: And that belief pays off. Because it turns out, customers are smarter than the headlines. They appreciate the nuance. Luna: If you strip away the marketing gloss, what remains is just a team doing its best to keep the lights on. Lucas: And showing them how they did it is the highest form of professional respect.