I Built an AI Ebook Creator to 700+ Users as a Solo Dev - Here's What Actually Worked for Me

TL;DR: Solo founder, no funding, no paid ads. Built Inkfluence AI - an AI ebook creator, from scratch since October 2025. 700+ users, 200+ indexed pages, and recently started getting cited by Microsoft Copilot. Sharing the full playbook below.

The Problem

Coaches, course creators, and solopreneurs know they should have a digital product - an ebook, a lead magnet, a workbook - but actually making one takes weeks. You have to write the content, format it, design a cover, export to PDF/EPUB, maybe create an audiobook version. Most people give up halfway.

I built Inkfluence AI to collapse that process into minutes. Pick a topic, generate an outline, write chapters with AI, design a cover, export. Done. It supports 23 content types - everything from novels to cookbooks to study guides.

Pricing: Free tier that produces a real 5-chapter ebook (not a crippled demo), Creator at $6.99/mo, Premium at $12.99/mo. The free plan is deliberately generous because the conversion path is organic: people create something real, see the quality, upgrade when they need EPUB export or longer books.

Tech Stack (for the devs)

• App: React + Vite + Tailwind CSS v4 on Vercel
• Marketing/SEO site: Astro 5 (this was the pivotal decision - more on that below)
• Backend: Firebase Auth + Firestore, Vercel serverless functions
• Background jobs: Inngest - book generation takes 60-90s per chapter, way beyond Vercel's 30s timeout. Inngest handles retries, 300s timeouts, and step functions out of the box. Best infrastructure decision I made.
• Payments: Stripe Checkout + Customer Portal
• AI: OpenAI GPT-4.1 (mini for Creator tier, full for Premium)
• iOS: Native app, same Firebase backend

What Actually Grew the User Base

No paid ads. Here's what moved the needle, ranked by impact:

1. GSC-driven content (not guessing, not vibes)

I have 84 blog posts and 200+ indexed pages. But the important part isn't the volume, it's the process.

Every piece of content starts in Google Search Console. I look for queries where I'm getting impressions but zero clicks, then either write new content or rewrite the title/meta of existing pages. Example: I noticed "free ai book writer" pulling 136 impressions at position 38. Wrote a comparison of 7 free AI book writing tools targeting that exact cluster. Same with "amazon kdp ai disclosure" - saw the query, wrote the guide.

The weekly loop:

1. Pull GSC data, sort by impressions with low/no clicks
2. Create or optimise content for those queries
3. Interlink aggressively (every post links to 3-5 related posts + tool pages)
4. Submit sitemaps + IndexNow on every deploy

This compounds. Pages that rank start generating data for new keyword opportunities, which spawn new pages, which rank and generate more data. Flywheel.

2. AI citation optimisation (the next SEO)

This one surprised me. Microsoft Copilot has cited my pages 591 times and it's accelerating, 53 citations in a single day recently. I started placing "direct-answer blocks" at the top of key pages: a single, definitive paragraph that comprehensively answers the page's core question. AI chatbots love to quote these.

If you're doing SEO in 2026, you need to be thinking about AI citation alongside traditional rankings. The traffic source is different (users inside Copilot/ChatGPT clicking your citation link) but it's real and it's growing fast.

3. Comparison pages and free tools

I have 15+ "vs" pages (Inkfluence vs ChatGPT, vs Jasper, vs Sudowrite, etc.). These convert well because people searching "[tool A] vs [tool B]" are actively making a buying decision. The trick is being honest about where competitors win, biased comparisons get bounced.

I also have standalone free tools (/book-outline-generator, /lead-magnet-generator, /kdp-tools) that rank for their own keywords and funnel people into the main product.

Mistakes I Made

Starting with a React SPA. Zero SEO for months. Migrating to Astro for the marketing site was the single biggest growth unlock. If your product needs organic traffic, start with a framework that does static/SSR. Don't bolt it on later like I did.

Ignoring internal linking. I spent a full session today just adding contextual links between 9 related blog posts. Tedious work, but my well-interlinked pages rank noticeably better than orphan pages. Google rewards topical clusters.

Not going mobile on my admin dashboard. I've been working 10+ hour days since October and only today made my admin dashboard responsive. Four months of being chained to my laptop just to check metrics. Embarrassing in hindsight.

Underestimating non-obvious use cases. My cookbook/recipe-book blueprint turned out to be surprisingly popular. Users kept finding it organically while I focused marketing on "ebook creation." If you're building a content tool, pay attention to which templates people actually use - it might not be what you expect.

By the Numbers

| Metric | Value |
|--------|-------|
| Registered users | 700+ |
| Blog posts | 84 |
| Indexed pages | 200+ |
| AI citations (Copilot) | 591 (trending sharply up) |
| Ad spend | $0 |
| Infra cost | ~$20/mo (Vercel + Firebase + domains) |
| Time to build | 4 months solo, 10+ hr days |

What's Next

Conversion from free to paid is the biggest lever right now. I have the top of funnel working, organic traffic is healthy - but free-to-paid activation needs work. If anyone has cracked this for freemium dev tools, I'm all ears.

Beyond that: more AI citation optimisation (growing faster than traditional search), expanding content blueprints based on user requests, and continuing the GSC-driven content flywheel.

AMA

Happy to go deep on any of this - the Astro + React architecture, the GSC-to-content workflow, Inngest for background jobs, pricing strategy, AI citation tricks. Fire away.