Week 4 of Building PageBolt: What We Shipped, What's Broken, and What's Next

63 users. 0 paying customers. 0 trials. But 60+ signups this week.

That gap is the story.

What We Shipped

No-Card Trial — Started the week at 0 trials. Launched a "start free, no credit card required" flow Thursday. By Sunday, 60+ signups. This was the missing piece. We had a product people wanted to try, but "add payment method to demo" was killing interest before they even started.

Dashboard v4 Redesign — Shipped a complete dashboard overhaul. Charts are faster. The API key management UI doesn't suck anymore. Usage history is actually readable. Small things, but they compound.

MCP Registry Listing — Got accepted into the official Model Context Protocol registry. One line in a JSON file, but it means our MCP server is now discoverable for Claude Desktop, Cursor, and Windsurf users. This is the channel to AI developers.

Chrome Extension Audit Review — Submitted our PageBolt Recorder extension to Chrome Web Store. Still in review, but if approved this week, it's another distribution channel.

210+ Blog Articles — We shipped a metric ton of SEO content. Tutorials on screenshots, PDFs, Python, Node.js, CI/CD, website monitoring. Real code, no fluff. Dev.to and Hashnode syndication automated. This is paying off: organic search traffic doubled mid-week.

What's Broken (and What We Found)

The Trial Conversion Killer — We launched the no-card trial, then... nothing happened. Signups came in, but nobody converted to a paid plan. Something was broken in the trial→upgrade flow.

Dug into it. Found it: we were showing trial users "Free plan — 100 requests/month" in the dashboard, but not telling them their actual trial usage. They thought they were on the free plan already. Why pay?

Fixed it Saturday: made the trial status explicit, showed remaining trial requests, and made the upgrade CTA show actual plan pricing instead of generic labels.

Haven't seen conversion yet (it's only been ~24 hours), but the fix feels right.

Metrics

• Users: 63 (up from 50 last Monday)
• Trials: 0→60+ this week (no-card removed the barrier)
• MRR: $0 (still, but that's expected — trials convert over days, not hours)
• Blog traffic: 2x organic search clicks vs. last week
• Signup flow: 5+ people watched the video demo, clicked "get started", got stuck. (This is the next bug to fix.)

What's Next

Trial Conversion Monitoring — Watch the trial→paid funnel closely this week. If the conversion rate is still under 5%, there's another blocker we haven't found yet.

Signup Flow Fix — We know people are dropping off after clicking "get started." The Loom video demo works (I've watched people watch it), but something after that video is breaking the flow. Probably an edge case in the no-card trial creation.

Content Machine — Keep shipping tutorials. The SEO compounds over time. We're now ranking on "screenshot python", "screenshot javascript", "pdf from html nodejs". These aren't million-search-volume terms, but they're high-intent. People searching "how to screenshot with python" are ready to buy.

Stripe Integration — Need to get paid plans actually working end-to-end. Right now we can collect payment, but the usage enforcement and plan limits aren't connected. This needs to be rock-solid before we start pushing people to upgrade.

The Honest Part

Zero paying customers after 4 weeks stings. Especially when I know the product works and people are signing up.

But the pattern is clear: no-card trial → 60 signups in 3 days. That's a 20x lift. The product has product-market fit for someone. Now it's about finding the path from trial to paying customer.

The trial conversion bug we found (showing "free plan" instead of "trial active") is probably killing 80% of our conversions. If we fix that and hit even a 5% conversion rate, we're at 3 paying customers this week.

Not a business yet. But a signal.

What I'd Ship If I Had More Time

• Webhook events (users asking for this constantly)
• Bulk screenshot API (for the "1,000 screenshots in 5 minutes" use case)
• Screenshot comparison API (highlight what changed between two URLs)
• Video recording with custom branding/watermarks

All of these came from user requests. None of them are hard. Just time.

The bottleneck isn't product. It's distribution and conversion.

---

If you're building something: What's your biggest bottleneck right now? For us, it's clear — we have demand (60+ signups), we have product (people use it), we just need to convert trials to customers. DM if you want to chat about it.

If you've used PageBolt: I'd love to hear where you got stuck. The trial sign-up flow, the API call, integration with your stack, pricing decision — whatever it was. That feedback drives the next week.

Next week: trial conversion numbers, Stripe enforcement working, and probably 10 more tutorials.

—

PageBolt is a web capture API (screenshots, PDFs, video recording with narration). Try free — 100 requests/month, no credit card.