Metrics from the pilot.
A/B test · vs existing third-party recommender · 70K+ sessions · 7-figure GMV fashion store
Read more →Where current engines fall short.
Same products for every guest.
80%+ of ecom traffic is anonymous — most visitors never log in. History-based engines need history they don't have, so first-time visitors all get the same default bestsellers.
Stale preferences shape new visits.
76% of consumers get frustrated when recommendations don't fit. Profile-based engines recommend from past preferences — so a shopper who bought a red coat in November returns in May and still sees winter coats.
Most of the catalog is never shown.
Most of the catalog never reaches a shopper. Engines trained on engagement reinforce what's already winning — 20% of products drive ~80% of sales, with impressions even more skewed. The long tail gets no air and never breaks through.
Recommendations stay in one category.
Engines that recommend within a single category miss the cross-sell — a 10–30% share of ecommerce revenue. A shopper browsing dresses gets more dresses recommended, not the matching shoes or bag.
New products start cold.
Historical recommenders have no behavioral signal for new SKUs — they stay invisible until enough engagement is accumulated, which won't happen if they're never shown. New launches and refreshed inventory all hit the same wall.
How ob.session works.
ob.session attaches to your storefront and personalizes every page load from how each visitor engages with your catalog in the moment — what they pause on, scroll past, return to.
Delivered as widgets, Collection page personalization, or both — across your storefront.
Like a good salesperson reads behavior — not names.
What catches attention.
How each visitor moves on the page — where they pause, hover, scroll past, etc. The micro-behavior shows which products are actually looked at, not just shown.
What shoppers look for.
The kind of product the visitor is moving toward — a category, a price band, a style. Inferred from the path through your store, not from a stored profile.
What kind of visit this is.
Where each visitor came from, their device, how the session has unfolded so far. A quick mobile visit looks different from a long desktop browse.
< 200ms · no warm-up · no identity
Watch it work.
One minute on a live Shopify store: a first-time visitor, and how the page changes.
Transcript
This is a live demo of ob.session, a recommendation app for Shopify. This is an ordinary Shopify store. I've never been here before: no account, no history, nothing for a recommender to look up.
Let's go to any collection. These products were chosen at page load for this visit. This is the order it starts in, because the session hasn't shown anything yet. The grid, the cards, the theme are all untouched. The order is the default collection order, and the store just works as is.
Now I behave like a shopper with a preference. I came looking for shoes, classic ones. I looked at a few pairs, opened one, added one to the cart, with some clicking and hovering around. I like dark colors.
Now we go back to the trousers page. Same collection, same theme, same grid, same products — a different order, and only for this session.
The widget on the product page follows the same signal, so the whole visit stays consistent. Not two systems guessing separately. That's ob.session: reading what a shopper wants while they're still deciding, and shaping the whole visit around it.
Where ob.session is different.
| Popularity-basedPopularity | Collaborative FilteringCollab Filtering | Session-basedSession-based | ob.sessionob.session | |
|---|---|---|---|---|
| Serves anonymous visitors | ✓ | — | ✓ | ✓ |
| Personalizes per visitor | — | ✓ | ✓ | ✓ |
| Real-time, in-session | — | — | ✓ | ✓ |
| Uses behavioral signals (hover, scroll, etc.) | — | — | — | ✓ |
| Collection page personalization | — | — | — | ✓ |
| Recommends from the long tail | — | — | — | ✓ |
Fully private by design.
ob.session sees what they do, anonymously — never linked to a name, email, or account.
No names, no emails
ob.session never sees names, emails, accounts, or payment information.
Never visible to others
Your catalog, orders and shopper data are never shared with, or made visible to, another merchant. A model trained for your store is never applied to another.
Consent-gated, GDPR-aligned
Nothing tracked before consent. Pseudonymous data handled per GDPR. The DPA is public.
Questions we hear most.
Will this slow down my store?
No. The script downloads alongside your page and runs only after the page is built, so it never blocks your store from rendering.
Recommendations have a strict time limit. If a response is late, shoppers see your store exactly as it would look without ob.session.
Why does it lift sales when the signals are so weak?
One hover says little. A hundred signals in sequence — what was opened, skipped, returned to, and for how long — say a lot about what a shopper is after.
Most visitors are not logged in and have no history, so the session is the only data there is, and most engines do not read it. In an A/B test against an existing third-party recommender, over 73,039 sessions, ob.session brought +43.3% revenue, +30.3% orders and +10.0% average order value.
How much does it cost?
There is no base fee and no monthly minimum.
You pay a share of the sales that come through ob.session recommendations — the "attributed" revenue — graduated per 30-day cycle: 4% on the first $10,000, 3% from $10,000 to $100,000, and 2% above $100,000.
How long does setup take?
Four steps: accept the agreement, approve billing, enable the app embed, and wait while we prepare recommendations for your store. The last step usually takes up to 15 minutes.
Does it work for first-time visitors?
Yes. Recommendations come from the current session, so a visitor with no account and no history gets personal recommendations within the same visit.