Natural language
A need in ordinary words ("something for dry skin under ₪100, nothing with fragrance") becomes a category, a set of filters and a sort.
Conversational commerce, grounded in your catalog
The advisor your website never had
A shopper describes what they want in their own words. The assistant renders your familiar product grid (same cards, same filter rail, same sort) already filtered to the request, and adds to the real cart in one click. There is no API to build, no customer data to hand over, and no page to rewrite.
Something for dry skin under ₪100, nothing with fragrance. Face moisturisers, all 34 products.
- Answers, not links
- It reads the request, picks the products and says why, in the shopper’s own language.
- Zero engineering
- One tag and one CSP line. No API, no re-index, no deploy.
- Only your catalog
- Every product, price and stock level is a live row in your catalog.
- Amazon In the search bar and a dedicated chat window. Answers questions, compares products, adds to cart.
- Walmart An "Ask Sparky" button across all shopping categories in the app.
- eBay Appears inline on the page the shopper is on. The closest example to proactive placement.
- Alibaba A sourcing agent with more than 10 million monthly users by the company’s account.
This is no longer an experiment. Four of the largest marketplaces in the world put a conversational shopping assistant on their own sites in the past two years.
None of them settled for a better search box. All of them chose a conversation.
The platform
Every kind of question a shopper arrives with
A search box serves the one shopper who already knows the product name. The rest describe a need, a person, an occasion or a doubt, and every one of those is a sale your storefront is currently losing in silence. These are the four things the assistant does about it.
01Understands the ask
Gifts and occasions
Requests with no category behind them: a gift, a trip, a routine, "what do I need for…". Each product type is verified to exist before it is offered.
Any language, and voice
It answers in the language the shopper opened in, and takes spoken input where typing is the friction.
02Answers from your catalog
Your own product grid
Same cards, same filter rail, same sort: pre-filtered to the request, with every filter it chose left as a live chip the shopper can edit.
Product questions
Questions about one row answered from stored product knowledge (the label, the spec, the size) rather than from the model’s memory.
Side-by-side compare
A shortlist compared on the attributes that actually differ between them, and a recommendation with the reason attached.
03Closes on your storefront
Your real add-to-cart
It calls your own add-to-cart function, so the badge, the mini-cart and your analytics all update exactly as they do today.
Never navigates away
A product named in a sentence opens over the conversation. The page underneath is handed back unchanged when the panel closes.
Honest about gaps
When you do not carry something it says so, instead of quietly showing the nearest unrelated thing and losing the shopper’s trust.
04Tells you what happened
Every question, logged
What shoppers asked, what was shown, what was tapped and what was added. The demand your search logs never recorded.
Catalog gaps
The asks that found nothing worth showing. A reading of what your range is missing, written by the people trying to buy it.
Product knowledge, pre-built
We answer the questions your catalog can support ahead of time, per product, so the assistant is grounded before the first shopper arrives.
And three things it is not
- Not a support bot: no order status, returns, refunds or account questions.
- Not a search platform: nothing is re-indexed and your PLP is not replaced.
- Not a data grab: no account, no profile, nothing kept between visits.
The model writes the query. Your catalog writes the answer.
Most shopping chatbots generate a sentence and hope it is true. This one does not get to. The model’s job is to turn a request into a structured query: a category, a set of filters, a sort. A deterministic search against your own catalog produces the products.
01
The shopper writes
"Something for dry skin under 100 shekels, nothing with fragrance." Typed or spoken, in Hebrew or English, in whatever words they have.
02
The model writes the query
Not the answer. It emits a category, filters and a sort, and every filter it chose becomes a live chip the shopper can edit or remove.
03
Your catalog writes the answer
The listing the shopper already knows, pre-filtered. Every product named is a row in your catalog, at your price, with your availability. Add to cart is your own add-to-cart.
What the published results say
The figures below measure an assistant running on the retailer’s own site, not traffic arriving from an external AI product. They are real in the sense that they were measured and published, and none of them was measured against a control group. We show them as they are, and mark what each one does not prove.
+31%
conversion at retailers running an on-site agent, against those that do not
Salesforce, 2026 · cohort comparison
4×
conversion for shoppers who engaged an assistant vs. everyone else
Dashly benchmark · self-selected
3×
site conversion after deploying an assistant, +38% basket
Tatcha, vendor case study · no holdout
2.69%
of LLM-referred visits engage a shopping assistant
329-brand study · engagement, not lift
So the pilot is a measurement, not a launch
We propose a holdout from day one: a randomly assigned share of traffic never sees the entry point, and the comparison is engaged-eligible against holdout, not engaged against everyone. It is the only design that tells you what the assistant caused rather than what it attracted. We would rather find out with you in month one than defend a self-selected number in month nine.
And four things that are true before any of it is measured.
1 tag
to install, plus one line in your CSP. Nothing else changes.
0
customer records, cookies or cart data ever reach us.
100%
of products, prices and stock read live from your catalog.
Days
from catalog access to an assistant answering on your site.
It sits on top of the stack you already run
No API to build, no authentication, no CORS, no re-index, no code deploy. Nothing about your storefront, your search or your PLP has to change. The assistant sends a product code and receives a boolean. The work is a tag manager entry and a line on a list your team already maintains.
However your storefront is built
- Shopify
- Adobe Commerce
- Salesforce Commerce Cloud
- BigCommerce
- WooCommerce
- Wix
- A storefront you wrote yourself
<!-- Google Tag Manager → Custom HTML → All pages -->
<script src="https://<widget-origin>/tag.js" async></script>Content-Security-Policy:
frame-src https://<widget-origin>;What the assistant will not do to your page
Never navigates away
A product named in a sentence is a button that opens it over the conversation, not a link.
Never touches your history
It refuses to write into the top page’s session history on your behalf; your back button stays yours.
Never reads your page
Cross-origin framing means it cannot see your DOM, cookies or cart even if it tried.
Never leaves the layout changed
When the panel closes, the page is handed back exactly as it was: same widths, same scroll.
Your customer data never reaches us. Not because we promised. Because it cannot.
The assistant loads as a separate frame inside your page, and the browser keeps the two apart. It receives only what the page chooses to send it. Your cookies, the shopper’s account and the cart contents it neither receives nor can read, not because we undertook not to look, but because the browser itself does not permit it, and will not permit it even if there is a bug in our code.
What never reaches us
The browser blocks it, not our code.
- Cart contents, order history and your customer records.
- Cookies, sign-in identifiers and anything your site stores in the shopper’s browser.
- Payment data of any kind. The assistant is never on a checkout page.
- A shopper profile. No account, no cross-session identifier, and no memory between visits.
What does reach us
Isolated to you, deletable on request, never used for training.
- What the shopper types or says, in order to answer it.
- Which products were shown, which chip was tapped, which product was added, so a complaint can be answered with the actual turn.
- A random session identifier with no link to a person, an account or a device.
- Usage data: what each answer cost. No conversation text at all.
Not a search platform. Not a support bot.
The category makes itself hard to compare. Discovery platforms re-index your catalog and replace your search; support bots answer tickets. This is a third thing: an advisor that sits on top of the storefront you already have.
| Discovery platformsConstructor, Algolia, Bloomreach | Support botsIntercom Fin, Zendesk AI, Gorgias | AgalitAn advisor on the page you already have | |
|---|---|---|---|
| What it replaces | Your search and category pages | Your help desk | Nothing. It sits beside your listing |
| What the shopper sees | A new results page | A chat window with sentences in it | Your own product grid, pre-filtered, with editable chips |
| Integration | Re-index the catalog, rebuild the PLP, weeks of engineering | Connect the ticketing stack | One tag and one CSP line |
| Where the answer comes from | A ranking model over their index | A language model, and hope | A deterministic search over your catalog |
Which is why the three are not alternatives to one another. A retailer already running a discovery platform keeps it: the assistant asks it for the shelf, and answers on top of what it returns.
Get started
Two ways to find out what it does on your catalog
Neither of us can know today what a shopping assistant does to your conversion rate. What we can do quickly is show you the thing answering on your own products, and write down what your storefront currently loses.
See it on your catalog
A working assistant on a slice of your own range, not a generic demo.
- Read access to your product and category data is all we need to begin.
- We index it, build the product knowledge and tune it on your categories.
- A 30-minute walkthrough on your own products, your prices, your stock.
Get a discovery audit
A written read of what your search box is doing with real shopper questions.
- We run the asks a search box loses: needs, gifts, occasions, comparisons.
- Where a shopper lands on an empty or irrelevant result, and what it costs.
- A short brief of what we found, emailed to you. No obligation either way.
The long pole is never the integration. It is catalog access. Everything after that is measured in days.