I built one workspace to replace the 6 AI tabs I had open — meet Inkfox

At this stage, I’d look at it in a very specific order: retention first, then paywall, then pricing.

With 118 users and $0 MRR, pricing is almost never the real issue yet. People usually don’t refuse to pay for something they rely on — they refuse to come back in the first place.

So the key question is: are users getting to a “second session moment”? If most users try it once and disappear, that usually means the product is still reading like a demo, not a workflow.

For a tool like this, the retention trigger is usually tied to one of two things:

• repeated creative use (people returning because it becomes part of their process)

• or asset continuity (projects, history, prompts, iterations that build over time)

Right now the strongest advantage you have is the unified workspace idea — everything tied to prompts and outputs. If that becomes a real “library of iterations,” not just generated files, it becomes harder to replace.

Paywall conversion only starts making sense once users feel loss if they stop using it. If there’s no “I need this again tomorrow” moment, the paywall will just sit there no matter how good it is.

So I’d focus less on optimizing monetization and more on answering one question: what makes someone come back on day 2 without being reminded.

I'd focus on understanding why people signed up in the first place. If 118 people tried it but nobody paid, that doesn't automatically mean pricing is the problem. It could mean they solved a curiosity problem, not a recurring problem.

I'd be talking to every active user I could find and asking what they were trying to accomplish, what they liked, and what stopped them from coming back. The first paying customer usually comes from solving a paiful problem, not from finding the perfect paywall.

Two things, one on the numbers and one on the offer itself. I think the second is where your real problem is.

On the numbers first: don't read anything into "0%" yet. With 118 users, a normal freemium rate of 1 to 2% predicts maybe one or two paying customers, give or take. Landing on zero six days in is ordinary luck, not a verdict. Even if your true rate were a healthy 1%, there's about a 1 in 3 chance you'd still be sitting on zero at this point. A rough rule of thumb: you need around 100 divided by your conversion rate, in users, before the first sale is even reliably due. At 1% that's about 100 users, so at 118 you are literally just reaching the line where the first one should show up. You don't have a 0% conversion rate. You have "too early to tell." So please don't pivot on this data, there isn't enough of it yet to pivot on.

The real issue, in my view, is the offer you lead with: "free, unlimited, no sign-up." That one line works against you in three ways at once.

First, the psychology of free. There's solid research on this (Dan Ariely's work on the "zero price effect") showing that free isn't just a cheap price, it changes behaviour. Once "unlimited and free" is the anchor in someone's head, any later request for money feels like a loss, not a deal. You've set the reference point at zero, and every paid prompt now has to fight that anchor.

Second, you gave away your own paid levers. The things you want to charge for are watermark removal and volume. But your headline already promises unlimited, which kills the volume lever, and most casual users honestly don't care about a watermark on a personal image, which makes the other lever weak. So the two doors you want people to pay through are either already wide open or not worth walking through.

Third, and this is the one I'd fix first: no sign-up means no relationship and nothing to lose by leaving. Your own origin story is the best part of the product, the desktop folder of half-named PNGs you kept losing, and every result tied to the prompt that made it. That saved workspace is the sticky, valuable bit, and it's exactly what an account gives you. Right now you give away the expensive part (the generation, which costs you real money per call) and fail to capture the cheap, sticky part (the saved history and library). I'd flip it. Let people try a few generations free, make the saved workspace the reason to sign up, and make the heavy stuff (HD, longer video, virtual try-on, commercial use, real volume) the reason to pay.

There's a unit-economics angle that makes this urgent too. You pay per generation, you promise unlimited, and you charge for a thing most free users don't need. Those three forces push conversion toward zero no matter how many users you get, and they're the same reason the bots can hurt you. "Unlimited free" of a per-call-cost product is a hole that gets deeper with every new signup.

So to your actual question, where to look first, pricing or the paywall moment or the product: I'd say none of those exactly. It's the shape of the free tier. The product and the site are genuinely good, your site even scores a perfect 100 on accessibility, which almost nobody manages. You don't need to remove the free tier. You just need to move the wall, from "everything, forever, free" to "a great taste, then sign up to keep your work, then pay to do it at full power." Then let your user count grow before you judge the rate at all.

Happy to go deeper on any of this, it's a fun problem and you're clearly building the right way.

I'd look at per-user cost and repeat use before touching pricing. For generative products, a simple cohort table of first-session cost, whether they come back within 48 hours, and which action triggers the second session tells you if the free tier is buying learning or just paying for drive-by curiosity.

I'm building TokenBar around a related lesson for AI coding tools: usage should be visible while decisions are happening, not only after the bill or limit surprise. For Inkfox, I'd make the cost/limit moment visible in-product before the upgrade ask, then compare conversion from users who hit that moment twice vs everyone else.

Appreciate the transparency, especially sharing the $0 MRR alongside the user numbers. Most launch posts only show the wins.

My first instinct wouldn't be to change pricing yet. With 118 users, I'd probably focus on understanding behavior before optimizing monetization. Are people generating once and leaving, or are they coming back multiple times? Retention feels like the strongest signal for whether the product is solving a real problem.

The consolidation angle is interesting because the value isn't necessarily that any one model is better—it's that users don't have to manage multiple tools and workflows. If people start relying on Inkfox as their default workspace, conversion may become a lot easier.

I'd also be curious what percentage of users are hitting the limits of the free tier. If very few are reaching that point, the issue might be engagement or retention rather than the paywall itself.

Either way, six days in feels far too early to draw strong conclusions. The fact that you've already gotten over 100 users and are openly sharing the numbers gives you a solid baseline to learn from.

Looking forward to seeing how the metrics evolve over the next few updates.

Love the radical honesty and the transparent numbers! Consolidating multiple AI workflows into one seamless interface is a genuine problem solver for creators.

Regarding your question about conversions: for a visual/creative platform, the first impression dictates user trust. Right now, your mascot logo is cute, but since Inkfox positions itself as a premium, all-in-one workspace for serious creators and business owners, transitioning into a highly professional, standard minimalist visual identity could instantly elevate the tool's perceived value.

When a platform looks like an industry-standard premium tool, users are naturally much more willing to cross that paywall. Exciting experiment, wishing you your first paid users soon!

I'd be careful treating this as a pricing, paywall, or product problem too early.

The harder question may be what conclusion actually deserves confidence from 118 users and 0 conversions.

Those sound similar, but they can lead to very different decisions about what gets changed next.

The danger isn't necessarily a lack of signal.

It's false confidence in the wrong signal.

I wouldn't make that call casually from the current data.