Answers from the product page
Specifications, sizing, fabric, care, compatibility, what is in the box, delivery options and the returns window — in the shopper's own language.
Online shops lose the sale at the last question. Will it fit, can I send it back, when does it arrive, is it in the colour I actually want — the answer is usually already on the product page, the size guide or the returns policy. The shopper does not find it, and the tab closes. Adoomi is a chat window on the product page itself, trained only on the catalogue, delivery information and policies you approve, answering 24/7 in the shopper's own language. Where it cannot answer — a live stock level, somebody's own order — it says so plainly and captures the question in the shopper's own words for whoever handles orders. There is no till feed and no account access here, and the assistant does not pretend otherwise.
Fig. 01.A · Retail and E-commerceAll sub-verticals
Each sub-vertical has its own page with sub-vertical-specific FAQs, chat examples and imagery.

Fashion shoppers stall on fit, so Adoomi answers sizing-chart, true-to-size, fabric-composition, care and returns-window questions from your product page, captures embroidery, gift-wrap and custom requests with the deadline attached, and never promises free returns your published policy does not offer.
Explore
Furniture buyers need measurements before they commit, so Adoomi answers dimensions, fit-through-door guidance, fabric and cover care, assembly time, delivery windows and returns-collection questions from your product page, and sends bespoke sizing or finish briefs to your consultant rather than quoting them itself.
Explore
Jewellery buyers ask about hallmarks, metals, gemstone documentation, engraving, aftercare and the resize policy, and Adoomi answers those from your published specification — while ring sizing goes to a sizer kit or an in-person fitting, and bespoke and remodel briefs go to your goldsmith.
Explore
Pet owners ask about life-stage food formulas, ingredients and protein content, harness sizing and pausing a subscription, and Adoomi answers those from your product page and published subscription rules — while anything that sounds clinical, from symptoms to dosing, goes straight to your vet partner.
Explore
Fig. 02.A · pain quotesA shopper who cannot settle a fit question does not email about it; they close the tab and buy from whoever answers first. The same handful of questions repeat on every product, every day — does it run true to size, what is the fabric, will it fit through the door, how long have I got to send it back — and each unanswered one is a basket left behind. The awkward ones land on a Saturday night, hours after the last person who could answer went home — your Monday inbox already knows the pattern.
Specifications, sizing, fabric, care, compatibility, what is in the box, delivery options and the returns window — in the shopper's own language.
Custom finishes, engraving, gift-wrap, trade quantities and made-to-order briefs arrive as one structured enquiry with the deadline attached.
No live stock, no order lookup. It says so, takes the order number and contact details, and routes the question to whoever handles orders.
These are the questions that decide a basket across fashion, furniture, jewellery and pet retail — answered from the catalogue and policy pages you have already published.
Bespoke and bulk requests are where a generic contact form quietly loses money. Adoomi asks the things your team would ask anyway — which item, which size or finish, the deadline, whether it is a gift, the quantity for a trade order — and files one structured enquiry with the shopper's own wording attached. A stock check arrives naming the exact product and variant. An order chase arrives with the order number the shopper actually has to hand, because nothing is ever looked up on their behalf.
With Adoomi
Fig. 06.A · after-hoursYour own enquiry timestamps cluster in the evenings and at weekends, which is precisely when nobody is watching the inbox. The assistant answers from your published catalogue and policies at those hours — the returns window for a nervous first-time buyer, the delivery options for a postcode, the care instructions on a fabric, the dimensions of a sofa — so the decision happens while the shopper is still on the page. What it cannot settle is captured and queued, so Monday morning is a list of real order enquiries rather than a pile of half-finished conversations.
Specifications, sizing charts, fabric and materials, care, compatibility, what is in the box, delivery options, the returns window — the questions that repeat per product, per day, get answered from pages you already publish. What still reaches your team is the work only a person can do: the bespoke brief, the trade order, the out-of-policy return, the parcel that has not moved since Saturday. Customer service spends the day on those instead of retyping the size guide.
Fig. 08.A · controlsThe lines the assistant does not cross are the ones shoppers push hardest. It will not promise a return, a refund or a free collection beyond the policy you published, so it cannot soften a rule to keep a sale moving; out-of-policy requests are captured for a person to decide. It will not guarantee stock, because there is no live feed from your till or warehouse — it answers what your pages say about availability and restocking, then captures the specific question for the team to confirm. And it will not look up an individual's order, account or tracking, because it has no access to any of them by design. Asked where an order is, it says so, takes the order number and contact details, and routes the enquiry. You review every answer before a shopper sees it. Where a product question is really a health question — a pet shop asked about a poorly animal, a supplement enquiry that belongs with a professional — it does not advise; it points to the vet or specialist named in your own content. And the chat window itself is EU-hosted, sets no tracking cookies, and what shoppers type can be exported or deleted on request under GDPR.
Product pages, delivery information and policies. Nothing else is used to fill a gap.
Returns and refunds are answered as published. Out-of-policy requests go to a person to decide.
There is no live till or warehouse feed, so the specific stock question is captured, not answered.
Individual orders, accounts and tracking are out of reach by design. It takes the order number and routes it.
Stock answers are as current as your last content sync and never live, because Adoomi reads no till or warehouse system. It answers what your pages say about availability and restocking, then captures the specific stock question and routes it to whoever handles orders.
Order lookup sits outside what Adoomi does: there is no access to an individual's order, account or tracking, by design. Asked where an order is, the assistant says so plainly, takes the order number and contact details the shopper has, and routes the enquiry to your team.
Returns and refunds are answered strictly from the policy you published — the window, the condition an item must come back in, who covers the postage. Out-of-policy requests are captured for a person to decide, because the assistant cannot soften a rule to close a sale.
Published detail: specifications and materials, sizing and fit guidance, care and compatibility, what is included, dimensions, delivery options and timeframes. Where a fact was never published, it hands the question over rather than guessing, and the gap surfaces for you to fill.
Payments, refunds and order changes stay in your own shop systems. The assistant explains from your published policy how a refund works and where to start one, but it cannot issue a refund, take a card payment, change a delivery address or cancel an order.
Any platform that lets you add one line to your site works — Shopify, WooCommerce, WordPress, Squarespace, Wix or hand-built. There is no plugin to install and no theme change; connecting your shop, reviewing the answers and adding that line runs to ~10 min end to end.
Plans start free: £0 a month covers 50 message credits, then Local £8 for 200, Growth £20 for 500 and Scale £75 for 4,000. One credit is one AI answer, so a shopper asking about sizing, then delivery, then returns spends several. Credits are spent when the assistant answers.
Most shop chatbots are either a decision tree somebody has to maintain or a queue for a human who is offline. Adoomi is neither. It reads your product pages, delivery information and returns policy as published, so when you update the page, the next content sync carries the change into the answers, and it is candid about the two things it cannot see: live stock, and one customer's order. Connect the site, read what it proposes to say, add one line to your website — ~10 min in total.
Sizing, dimensions, materials, care and delivery answered on the product page in the shopper's own language.
Weekend and late-night questions answered from your published catalogue and policies, with enquiries queued for the morning.
Stock checks and order chases are declined plainly and captured for whoever handles orders, order number and all.
Give it your catalogue, your delivery information and your returns policy, then put the questions that actually decide a sale to it — is this in stock, where is my parcel, can I return something I have worn once. How it declines is as much a part of the shopping experience as how it answers, and every reply is yours to read and change before a shopper ever sees it. Start on the Free plan at £0 a month with 50 message credits — the pricing page has the full plan comparison, and the order-answers product page walks the order-enquiry flow end to end.
Fig. 12.A · Retail and E-commerce