I Looked at My SaaS Conversion Data and Realized I Was Selling the Wrong Thing

I run a free AI image and video tool. Last week I pulled my Stripe subscriptions in cohort order. The data was... not what I expected.

For context: I run ZSky AI. Photographer with aphantasia, healed from a TBI through creative work, built an AI tool because I needed to externalize what I couldn't visualize in my mind. 36,000+ creators have signed up since launch. We have a generous free tier and four paid tiers above it.

My four paid tiers, from cheapest to most expensive:

1. Skip the Line — $9/month. No extra credits. Just removes the queue wait. You use your free-tier credits but they process immediately on dedicated GPUs.
2. Pro — $19/month. 3,000 extra credits per month. Same instant generation. Standard "more of everything" tier.
3. Ultra — $49/month. 6,000 credits, 4K upscaling, more advanced features.
4. Max — $99/month. 20,000 credits, API access, white-label options.

When I built the pricing page, I assumed Pro would be the workhorse. Pro is the classic SaaS sweet spot — enough features to feel substantial, low enough to feel approachable. Skip the Line was almost a throwaway tier for people who wanted to support the project.

Last week I pulled the actual data:

2026-04-10 | Pro Monthly $19
2026-04-10 | Skip the Line Monthly $9
2026-04-09 | Skip the Line Monthly $9
2026-04-09 | Skip the Line Monthly $9
2026-04-08 | Pro Monthly $19
2026-04-08 | Skip the Line Monthly $9
2026-04-08 | Skip the Line Monthly $9
2026-04-07 | Skip the Line Monthly $9
2026-04-06 | Skip the Line Monthly $9
2026-04-05 | Pro Monthly $19
2026-04-04 | Pro Monthly $19

9 of the last 11 paid signups were Skip the Line.

Not Pro. Not the workhorse tier I designed. The almost-throwaway tier I added because I felt bad putting an instant-generation paywall on the free tier.

What I learned

For weeks I was telling people my product was about "AI image and video generation." That's the surface description. It's true, but it's not what people are paying for.

What people are actually paying for is escaping the queue.

The free tier already does what they need creatively — quality, speed, no watermark on video, full commercial use. The free tier is good enough that the only thing left worth paying for is the wait time. And Skip the Line at $9 is the cheapest, simplest way to remove that wait.

This sounds obvious in retrospect. It's a classic "the constraint is the product" insight — when your free tier removes every other constraint, the constraint that remains is the one people pay to fix.

But I had to look at the data to see it. My intuition (and frankly, every SaaS pricing tutorial I've ever read) said the upgrade path goes Free → Cheap → Mid → Expensive. The reality at ZSky is the upgrade path goes Free → "please just stop making me wait."

What I'm changing

1. Skip the Line is now the FIRST upgrade option mentioned everywhere.

My homepage used to say "Free tier and paid plans starting at $19/month" because Pro was the entry-level paid tier in my mental model. I just changed that to "Skip the Line ($9/month) for instant generation." Same goes for the schema descriptions, the meta descriptions, the related-card pricing summaries on every landing page. Skip the Line gets the first slot now.

2. Every CTA explains the constraint.

Instead of "Try free or upgrade to Pro," I'm rewriting CTAs to "Try free, or skip the queue with Skip the Line — $9/month, instant generation on all 7 GPUs." The mention of GPUs is intentional: it makes the value tangible. People pay $9 to put their requests on dedicated hardware instead of standing in line.

3. The marketing email is being rewritten.

The old weekly recap emphasized features. The new weekly recap leads with: "36,854 creators are joining ZSky every day. The queue is getting longer. If you want to skip it for $9/month, here's how." Honest framing of the actual value exchange.

Why I'm telling you this

If you run a SaaS product with multiple tiers, pull your Stripe data in cohort order before assuming you know which tier converts best. Don't trust the framework. Don't trust your own pricing intuition. The data will tell you what people actually want to buy, which is often very different from what you thought you were selling.

The other lesson I'm taking from this: don't be afraid to add a tier whose only job is to remove a single constraint. Skip the Line is unusual because it doesn't add features. It just deletes friction. Most pricing pages assume tiers should stack value. But sometimes the most valuable thing you can sell is the absence of an obstacle.

If you want to see how I'm restructuring the pricing flow, the changes are live at zsky.ai/pricing. Or if you just want to see the product, zsky.ai/create — free tier, no credit card, 200 credits at signup.

— Cemhan

---

I run ZSky AI, a creativity engine made by artists for artists. I write here on dev.to about indie SaaS, AI tools, and what I learn building one. If you want monthly data reports, they're at zsky.ai/data-reports. CC-BY-4.0, freely citable.