Exact on day one
Your rules are compiled into the model as they are, so recommendations are right from the first visit. No weeks of collecting data first.
For fashion stores
Complete-the-look recommendations for fashion stores. Live in one line, proven with your own A/B test.
Your products are stars; we show your shoppers the constellations.
No product feed: Outfiy reads your catalog from your own pages. Your merchandising rules work exactly from day one, then learn within limits you set. No cookies, no shopper profiles.
Works with Shopify · WooCommerce · ikas · Ticimax · Next.js · React and any custom store
Product page · up-sell. The item is sold out: close alternatives, a step up in quality and price, keep the sale.
Cart page · cross-sell. Products that go with what is already in the bag, so the order grows before checkout.
Home page · personalisation. A returning shopper sees picks based on what they viewed.
Why it’s different
Not another “you may also like” strip.
Learns within guardrails
Weights learn from clicks and carts, but only inside the range you allow. If performance falls below your plain rules, it rolls back on its own.
Proves its own uplift
10% of visitors don’t see the widgets. Your report shows the difference against that control group, with a confidence interval.
Live demo
Walk through our demo store.
No sign-up. Open a product, add it to the bag and watch the recommendations appear. Turn on the merchant view to see why each item was picked.
Open the demo store- Open a productThe look and similar items appear under it.
- Keep browsingPicks follow what you look at.
- Open the merchant viewSee the rules, move a weight, watch the ranking change.
How it works
One line to install. Everything else is automatic.
I
Add one line
Paste the script into your site or Google Tag Manager.
II
It learns your catalog
Outfiy reads products from your pages as people browse.
III
Recommendations go live
Widgets appear on product, listing, home and cart pages.
Under the hood
Your merchandising rules, readable and editable, compiled into a model that can learn.
Each rule becomes a small piece of a neural network. On day one the network scores products exactly like the rules; after that, every weight can only move inside its allowed range. Point at a rule to follow its path.
Same category
The strongest boost: alternatives from the same category come first. Learning can move it, but never outside its range.
A hard filter built only from product data. It is frozen: learning can never let an out-of-stock product through.
Women’s pages never show men’s products, and the other way round. Also frozen.
The strongest boost: alternatives from the same category come first. Learning can move it, but never outside its range.
Products in a similar price range score higher; a wider ±40% band adds +0.5. The band edges can learn too.
A 12 × 12 table of colour pairs (navy with beige, black with white…); each entry may move ±0.5.
Shared descriptive words in product names (linen, oversized, chelsea…), rare words weighing more.
Popular products get a small lift, in five levels.
Fixed · never learned
What you’ll see
A report that shows what the widgets earned.
Every store gets the same measurement: 10% of visitors form a control group without widgets, and the difference is reported with a confidence interval.
yourstore.com · last 14 daysSample report · figures are illustrative
Recommendation click rate6.4%
Add to cart from recommendations2.8%
Revenue from recommendations₺48,300
Uplift vs control group+9.6%
Add-to-cart rate per session, with 95% confidence interval
+9.6% uplift (95% CI +3.1% … +16.2%).
Compared
How Outfiy differs from the usual options.
| Outfiy | Free built-in recommendations | Recommendation SaaS | Personalisation suites | |
|---|---|---|---|---|
| Setup | One script tag | Built in | App or script, plus feed setup | Integration project, often weeks |
| Product feed | Not needed | Not needed (platform data) | Usually required | Required |
| Day-one quality | Your rules, exactly | Same category or bestsellers | Generic until data builds up | Configurable; models need data |
| Fashion complete-the-look | Built in: top → trousers, shoes, bag | Rarely | Sometimes, often manual | Usually, with setup |
| Learning transparency | Readable rules, bounded weights, automatic rollback | No learning | Mostly a black box | Varies, often opaque |
| Privacy | No cookies, no shopper profiles | Platform defaults | Often cookies and profiles | Cookies and profiles |
| Uplift measurement | Built-in 10% control group with confidence interval | None | Sometimes | Separate A/B tooling |
Typical approaches; check each vendor.
Pricing
Free until it proves uplift.
Start with a 30-day free pilot. Keep Outfiy only if your own report shows it earns more than it costs.
30-day pilot
Free
- All widgets, including complete the look
- Your own A/B report against a control group
- Help with installation
After the pilot: [price to be announced]
Request a free pilotStarter
€49/month
Up to [visit limit] monthly visits
Growth
€149/month
Up to [visit limit] monthly visits
Scale
€399+/month
[visit limit] monthly visits and more
Free pilot
Try it on your store.
Tell us about your store. We set up a 30-day pilot with you and share the uplift report at the end.
- One script tag, no product feed.
- 10% control group from the first day.
- Remove it any time in a minute.
Thanks, your request is in.
We’ll reply by email to set up your pilot.
FAQ
Questions, answered.
Will it slow my store down?
The script is about 32 KB (gzipped), loads asynchronously and does its work in small idle slices after your page has loaded. Widgets reserve their space before they render, so nothing jumps; our automated tests check every build for zero layout shift and no long tasks.
What about GDPR, KVKK and the cookie banner?
Outfiy sets no cookies and builds no shopper profiles. In the visitor’s browser it keeps a random test-group ID, the products viewed in this session and a product cache; events reach us without raw IP addresses, and Global Privacy Control and Do Not Track are honoured. Whether your banner should mention it is for your legal team to decide; our privacy notice lists exactly what is stored.
What happens if my store’s theme changes?
Outfiy reads products and finds its place from the page each time, so most theme changes need nothing from you. If it can’t find the product block, the widget goes above the footer; you can always pin it with a <div data-wirerec="similar"> container.
Does it work on SPA and headless storefronts?
Yes. It follows client-side navigation (React, Next.js, Vue, Nuxt) and reads each new page; a single-page version of our demo store is tested on every build.
How long does removing it take?
A minute: delete the script tag or the Tag Manager tag. We can also switch the widgets off for your store from our side, without touching your site; it takes effect within minutes.
How is it different from free built-in recommendations?
Built-in widgets usually show the same category or bestsellers. Outfiy completes the look across categories (a shirt gets trousers, shoes and a bag), follows your own rules, and measures its uplift against a control group, so you know whether it pays.
Beyond fashion
Also works for home, beauty, electronics…
Fashion is where we start. The same rules engine has complement tables for other catalogs too.
