ShopOwl
An AI shops your store. Finds what kills your sales.
I'm WoXille, a solo dev from France, and a Computer Science Engineering Student. Today I'm launching ShopOwl, the project I wish existed when I was helping a friend fix his Shopify store last year.
The problem 🦉
Most Shopify merchants are passionate craftspeople, not UX experts. They stare at their store every day from the admin panel, but they never actually experience it as a first-time visitor would. So when sales drop, they don't know if it's the price, the product page, the checkout, the trust signals, or the mobile layout.
The "professional audit" alternative? €500-2000 from an agency, 1-2 weeks of waiting, a 40-page PDF full of jargon (CTA, conversion rate, bounce rate, funnel...) that most non-technical merchants can't act on.
What ShopOwl does
You paste your store URL. An autonomous agent (Playwright + Chromium) does what a real customer would:
🔍 Lands on your homepage and times the page load
🛍️ Finds a product, opens the page, looks for the buy button
🛒 Tries to add to cart and reach checkout
📱 Repeats the journey on mobile (375×812)
🧐 Checks 60+ signals: reviews, trust badges, payment methods, contact info, image count, mobile sticky cart, page speed (TTFB / FCP / weight), legal mentions, SEO basics, and more
Then AI transforms the raw data into a report a baker could read. No "above the fold" - we say "visible without scrolling". No "CTA" — we say "buy button". Findings are split into 🔴Critical / 🟡 Improve / ✅ Working well, each with a concrete action.
What's different
- 2 minutes end-to-end. Paste URL → grab a coffee → done.
- No install, no plugin, no Shopify app permissions, no scope.
- Plain French/English — we relentlessly remove jargon. If a baker can't understand a finding, we rewrite it.
- Honest gaps shown alongside wins — the AI is explicitly prompted to highlight what's working, not just complain.
Stack (for the curious)
- 🐘 PHP 8.2 + Apache (the boring monolith)
- 🌐 Node.js + Playwright (the autonomous agent)
- 🐬 MySQL 8
- 🤖 Claude Haiku 4.5 (Anthropic) for the report writing
- 🐳 Docker Compose + Caddy (auto SSL) on a single DigitalOcean droplet
- 💸 Stripe for payments
Lean POC architecture, mono-droplet, ~24 €/month infra.
Pricing
(keep in mind that there is -50% for the "Early Birds")
- 🎁 First report free, no credit card (3 free reports for the Early Birds)
- 🎟️ One-shot from €4.50
- 🚀 Subscription from €9/month (Early Birds 50% off — locked in for
life)
What I'd love from you
This is a solo POC, day 1 in production. I'd love:
- Your honest reaction to the report on YOUR store (or any URL)
- Feedback on tone, jargon I missed, frictions you'd want analyzed
- DMs if you spot bugs or weird findings — I'll fix in hours, not days
- Give some feedback at the bottom of your report to help me upgrade this tool.
The bests beta testers will be rewarded with free subscription or free reports !
Try it for free : https://shopowl.app
ShopOwl is available in english and in french.
Thanks for being here !
4 Comments
The "we say 'buy button' not CTA" language rule is the actual moat. Every Shopify audit tool aimed at non-technical merchants eventually drifts into jargon because the founders are technical and their language becomes the default. Holding the line on plain merchant language is harder than it looks.
The pattern we see at Hivemind across founder-led repositioning: the wedge is almost never the feature stack, it's the language layer. ShopOwl's actual product is "the only audit a baker can read."
Pricing is the open question though. €4.50/one-shot and €9/mo means you're competing more with your own free first report than with €500 agency audits. Worth thinking about whether the model is one-time diagnostic or continuous monitoring with alerts when something regresses after theme updates. Different product, different LTV.
Thanks — I'll be coming back to this comment more than once.
On the language point: what you're describing is something I live without having framed the difficulty this clearly. I tend to assume the work was done once at positioning time, when in practice I'm pulling out a "CTA" or an "above the fold" on every page review. Your line about technical founders whose vocabulary becomes the default norm makes me little uncomfortable — that's exactly the drift I'm at risk of.
On pricing, I hadn't been looking at it from that angle. You're right: at €4.50 a one-shot, I'm competing with my own free tier, not with the €500 agency audits. And my €9/mo Starter, as it stands, is just the one-shot with a quota on top — it isn't offering anything meaningfully different.
The "continuous monitoring + alert when something regresses after a theme update" framing is genuinely shifting the way I think about the product as I write this reply. It's probably a different product, with a different perceived value, and I hadn't really done the work of separating it from the one-shot diagnostic.
I don't know yet what I'll do with this. If you have models in mind — monitoring products that have managed to land with a non-technical audience, or common mistakes founders make on this kind of pivot — I'd genuinely value any reference. This is clearly an angle where I'm short on pattern recognition.
"Love the positioning — most merchants really can't tell where they're losing sales. I do human conversion audits for the same problem. Different approach, same mission. The automated report + human interpretation combo could be powerful. Worth connecting?"
Yeah, send me an email at contact@shopowl. app