What we're worried might go wrong on launch day (T-6 pre-mortem)

We launch Avery NXR on Product Hunt in 6 days — July 9, 2026.

Founders often write post-launch retrospectives about what went wrong. Useful but too late.

I'm writing the pre-mortem. Naming what we're worried might go wrong BEFORE July 9, so we plan for it now. This is the version of the retrospective I'd want to have written if I could rewind time from a hypothetical July 10.

Worry 1: Comment volume overwhelms us

Product Hunt launch threads can get 300+ comments in a single day. Response quality drops when volume overwhelms capacity.

What could go wrong: We fall behind on replies. Late replies hurt engagement rate. Engagement rate affects ranking. We drop from #2 to #5 because we couldn't respond fast enough.

What we're doing about it: Pre-committing 5 team members to comment moderation on launch day. Rotating shifts across timezones so the thread has active attention for the full 24 hours. Pre-drafting replies to the most predictable questions (pricing, architecture, competitor comparisons).

Residual risk: We can only do so much. If volume is 500+, even 5 people is thin.

Worry 2: Launch video has a rendering issue

Videos rendered on our machines look great. Videos on other people's devices sometimes don't.

What could go wrong: Launch video shows up with a watermark artifact on lower-resolution devices. Some users see a broken viewing experience in the critical first hour.

What we're doing about it: Testing the video this week on at least 10 different device configurations (mobile, tablet, older laptops, various screen resolutions). Re-rendering if we find issues. Backup version prepared in case we need to swap mid-launch.

Residual risk: We can't test every device. Something we didn't test might break.

Worry 3: A key customer becomes unavailable

We have customer testimonials + demo participation planned. Life happens.

What could go wrong: A customer scheduled for a launch-day quote or short video gets sick, has an emergency, or just becomes unresponsive. We scramble for backups. Backup quality is 20% less compelling.

What we're doing about it: Pre-recording final customer content by T-4 (July 5), not T-1 (July 8). Confirming attendance twice — once at T-3 and once at T-1. Backup customers identified for every planned appearance.

Residual risk: Customers have lives. Some percentage of planned participation won't happen. Plan for 80% attendance, not 100%.

Worry 4: Onboarding funnel bugs surface under load

Our onboarding is stress-tested. Production load is a different animal.

What could go wrong: A subtle bug in the signup → first-agent-configured flow surfaces at 50x normal traffic. Users see a broken intermediate state. Confusion + abandonment.

What we're doing about it: Load testing this week with 100x normal traffic simulation. Manual QA on the exact onboarding flow (not just infrastructure). Engineering on-call throughout launch day with priority to fix onboarding issues within 30 minutes.

Residual risk: Some bugs only surface with real users doing weird things. Some abandonment is unavoidable.

Worry 5: Rank ambiguity

We could be #2 by 10 upvotes or #4 by 30 upvotes. Small differences, big narrative implications.

What could go wrong: We finish #4 instead of top 3. The launch feels like a disappointment even though the underlying numbers are strong.

What we're doing about it: Setting expectations honestly with the team + investors + supporters BEFORE launch. "Top 5 is a success. #1 is a stretch goal. #10 is disappointing." Publishing our success framework in advance (see [post 303]) so the narrative isn't determined by rank alone.

Residual risk: Emotional weight of rank is real regardless of what we say in advance.

Worry 6: Competitive launch collision

We don't know who else is launching on July 9. It's a real risk.

What could go wrong: A well-funded competitor launches on the same day. They have more upvote-mobilizing power. We rank behind them despite building the better product.

What we're doing about it: Checked the visible PH pipeline for the week of July 9. Nothing obvious as of T-6. Diversifying our launch impact across channels beyond just PH (Twitter, LinkedIn, direct email, press) so PH rank isn't the only signal that matters.

Residual risk: PH's calendar isn't fully transparent. Something could surprise us.

Worry 7: The "hunter no-show"

Our hunter is committed but life happens.

What could go wrong: The hunter can't be as active on launch day as planned. Comments referencing them go unanswered. Some PH users interpret hunter absence as low-quality launch.

What we're doing about it: Confirming hunter's July 9 availability at T-4, T-2, and T-0. Backup engagement plan if hunter is limited (team members respond, hunter posts confirmation of "team is running this today" in the thread).

Residual risk: Hunters have other launches to attend to. Full-day presence isn't always feasible.

Worry 8: Energy fizzling too early

Launch day has a natural rhythm. Early hours (Asia daytime), mid-day (Europe), late day (US Pacific). Energy that fizzles by mid-day loses the US audience.

What could go wrong: We peak in Asia+Europe hours + lose momentum by the time US Pacific wakes up. We finish behind competitors who saved energy for the US window.

What we're doing about it: Team shift rotation so someone is always fresh. Pre-scheduled content pushes across timezones (blog posts, tweets, LinkedIn drops timed to each region's morning). Pre-briefed US-based supporters for late-day engagement push.

Residual risk: Sustaining 24 hours of momentum is genuinely hard. Some drop-off is inevitable.

Worry 9: The "critical comment" thread

Every launch gets at least one comment that could go either way — from constructive skeptic to hostile takedown.

What could go wrong: We respond poorly to a critical comment. The response goes viral in the wrong direction. Public thread becomes a discussion of our defensive response instead of our product.

What we're doing about it: Ruban + I both pre-approved to draft responses to critical comments. Personal, human, defensive-but-not-defensive tone. Pre-written framework: acknowledge, engage the technical substance, decline to escalate emotionally.

Residual risk: In the moment, tone is hard. We might miss.

Worry 10: Personal exhaustion

I run point on the launch. Yesterday's version of this list assumed I could go 20+ hours.

What could go wrong: I hit hour 12 and my judgment noticeably degrades. Later comments are shorter, less thoughtful. Real damage to relationships that would have been fine with a better version of me.

What we're doing about it: Mandatory handoff at hour 8. Ruban takes over the primary voice for 3 hours while I rest. I return for the US Pacific window with full energy.

Residual risk: Adrenaline convinces you you're fine when you're not. Discipline required to actually take the break.

Worry 11: A minor bug that becomes a major story

Small bugs surfaced at launch scale become disproportionate stories. Twitter amplifies.

What could go wrong: A minor rendering issue on the marketing website goes viral. A tiny UI inconsistency in the product becomes the launch narrative.

What we're doing about it: Full audit of website + product surfaces this week. Ship fixes for anything we can find. On launch day: monitor Twitter for our brand name aggressively. Respond to any brewing narrative within 30 minutes.

Residual risk: Twitter is unpredictable. Small things can become big fast.

Worry 12: Our own team dynamics under stress

Launches stress-test team relationships. Fatigue + high stakes bring out the worst version of everyone.

What could go wrong: A minor disagreement about launch tactics escalates. Two team members have a snappy exchange. Someone feels unheard. Relationships fray at exactly the moment we need each other most.

What we're doing about it: Explicit team norm before launch day: "Assume best intent, especially when tired." Ruban + I as designated de-escalators. Post-launch dinner scheduled where we celebrate + reconnect no matter what happened.

Residual risk: Emotions are emotions. People are people.

What could go REALLY wrong

Beyond the individual worries, three scenarios that would be genuinely bad:

Scenario A: Infrastructure failure. Website goes down for 3+ hours during peak launch traffic. Customers can't sign up. Damage is direct + measurable.

Mitigation: Redundant infrastructure across regions. Auto-scaling. On-call engineering with 30-minute response SLA.

Scenario B: Security incident. A vulnerability gets disclosed publicly on launch day. Our launch narrative becomes a security narrative.

Mitigation: Security audit completed 4 weeks ago. Bug bounty program active. Incident response playbook ready.

Scenario C: Ruban or my personal emergency. Life happens. One of us becomes unavailable mid-launch.

Mitigation: Second-in-command briefed. Cross-trained team members. Launch continues if one of us falls out.

What we can't plan for

Some risks aren't pre-mortem-able:

→ A macro event (major news, market crash, natural disaster) that dominates attention on July 9 → A competitor pulling a coordinated attack we don't anticipate → A key influencer randomly posting something negative about us the night before → Something we haven't imagined at all

For these, the plan is: adapt in real time. Trust the team. Keep the customer + community at the center of decisions.

The reason to name worries publicly

Some founders would hide these worries. Present a confident face. "Everything is on track."

I'm choosing to name them because:

1. Publicly naming risks forces us to plan for them. If we said "everything is fine" we might miss things.

2. Founders reading this benefit from honest pre-mortems. Launches are hard. Everyone worries. Sharing the worries normalizes the experience.

3. If any of these go wrong on July 9, we've already told you it might. No pretending in the retrospective. Honesty from the start.

Six days to launch

Six days to fix what we can fix. Prepare for what we can prepare for. And accept that some risks will remain no matter how much we plan.

The launch will happen. Some things will go wrong. Some things will go right. We'll iterate.

→ avery.software — Product Hunt launch: July 9, 2026. Get the alert: producthunt.com/products/avery-nxr?launch=avery-3