Who was in your catalog, which company they buy for — and whether you can reach them
Buyers will list every tracked catalog session, resolve it to a company and a role rather than only to a person, and score it for reorder and quote intent on an explainable 0–100 scale. The same buyers will be grouped into audiences you can hand to Callio or a rep — “Viewed price list, no quote”, “Compared SKUs across a category”, “Returning buyer, no reorder in 60 days” — before the intent goes cold. It is on the roadmap, with no date announced.
- Roadmap product — no date announced, no B2B numbers on this page
- Built on shipped session identity resolution and the 28-field segment builder
- Consent-aware reachability, exactly as it works for shoppers today
Buyer sessions
Who was in the catalog, resolved to the company and the role behind it — buyer, orders, account value and how you can reach them.
Reorder & quote intent
Every buyer session in view, bucketed by its explainable intent score.
Score from catalog views, price-list opens, spec-sheet downloads, quote-form starts and the resolved account.
Concept preview of a roadmap product — figures illustrative.
0–100
Explainable intent today
Product views, dwell time, cart, checkout and resolved identity — every signal shown with its points, never a black-box probability
56
Pre-made audiences today
Scored against the store’s own customers and 90 days of tracked behaviour — the model B2B audiences will follow
28
Fields in the segment builder
Nested AND/OR groups three levels deep, twelve operators — the same builder B2B fields would be added to
≈
Recoverable stays modelled
Counted reach, modelled revenue: segment reach is counted from real rows, recoverable revenue is always marked ≈
These figures describe Comver AI Commerce for Shopify DTC as it ships today. Buyers is a roadmap product — nothing here is a B2B measurement, screenshot or price, and the mock-ups are concept previews with illustrative figures.
A session is a person today. It will also be a company.
The Unified Segments engine already resolves a storefront session to the person behind it — a Shopify customer id captured while they were logged in, a synchronised customer record, or an email — and says honestly how it matched, or that it could not. The B2B twin adds one level: through the buyer to the company they order for, and the role they hold on it. Every buyer session then belongs to an account sitting in exactly one pipeline stage, so the row you are reading always has a board position behind it.
- Resolved by the buyer’s login on the catalog, a company domain on the email, or an invitation from the account portal — never guessed
- Unresolved sessions stay unresolved: the row will read “company not matched yet”, the way anonymous sessions read today
- Role matters as much as identity — purchasing, owner or assistant changes both the script and who should call
Accounts & buyers pipeline
ComingEvery company the store knows will sit in exactly one stage — worked out from orders, open quotes and catalog behaviour, never typed in.
- 1Prospect accountA company seen in the catalog — no buyer reached yet.
- 2Identified buyerA buyer with a name, a role and a company behind them.
- 3Quote requestedA quote is open — chase it before it goes cold. Chase the quote
- 4First POBought once — the reorder is the whole game.
- 5ReorderingOrdering on a rhythm and still active.
- 6Key accountThe revenue this catalog cannot afford to lose.
- 7LapsingOrdered before, gone quiet — call the buyer now. Call the buyer
A next best action per account, and a Callio campaign on any whole column — the way the Customers board already works for shoppers.
An intent score you can read like a receipt
Today the score is built from product views, dwell time, event count, cart opened, added to cart, checkout started and whether the visitor is known — each signal shown with its points, and detected friction named. The B2B twin keeps the arithmetic visible and changes the signals to the ones that predict a reorder or a quote.
Catalog depth
How many SKUs in a category were opened, and how long each was read — the B2B equivalent of repeat product views.
Price list opened
A buyer who opens the tier table is pricing a job. It is the strongest pre-quote signal a catalog has.
Spec sheet downloaded
Specification, safety and warranty documents downloaded — the sign that the technical objection is being worked through.
SKUs compared
Two or more SKUs in the same category opened in one session, or a comparison table used. Procurement compares before it commits.
Quote form started
The form opened, part-filled and abandoned — the B2B abandoned checkout, and the highest-scoring stall in the funnel.
Reorder overdue
The account’s usual rhythm passed without a purchase order — an account-level signal that lifts every buyer on it.
Resolved buyer
A session tied to a known company and role scores higher and is worth a call, because there is someone to call.
Detected stall
Rage clicks on a tier table, repeated returns to lead times, a quote form opened three times — named as friction rather than hidden in the number.
The audiences a wholesale catalog actually needs
Comver ships 56 pre-made intent audiences for DTC, each scored against the store’s own data, with counted reach by email and SMS and a recoverable figure that is always marked ≈. These are the B2B ones being designed — definitions fixed, sized from your own rows, and activatable in one click.
Viewed price list, no quote
Opened the tier table in the last 30 days and never started a quote. The single largest stall in the planned quote funnel — and a call Callio can make the same day.
Compared SKUs across a category
Two or more SKUs in one category opened in a session, or the comparison table used, with no order. A rep with the right spec sheet closes these.
Returning buyer, no reorder in 60 days
Ordered before, still browsing, but the rhythm has broken. The B2B twin of “Slipping away”, worked with a call rather than a discount.
Quote started, never submitted
Part-filled quote forms, grouped by account. The buyer already told you what they wanted — the form simply did not finish.
New account, first PO not placed
A company registered on the portal or approved for terms that has not yet ordered. Onboarding proof, not a promotion.
Ordered below the next price tier
Accounts whose last order sat just under a volume break — the clearest case for a volume-bundle offer or a call from the rep.
Or describe the buyers yourself — 28 fields today
When the catalog does not have the audience, the builder already lets you construct it: nested AND/OR groups up to three levels deep over 28 fields across customer, purchase, source, product, cart, visitor and predicted signals, with twelve operators, a live preview of how many match and one save that makes it a Segment, an Audience or a Personalization segment. The B2B work is to add the fields that only exist for accounts — company, price tier, terms, MOQ, quote state, days since last purchase order — to the same engine, so a segment keeps meaning the same thing everywhere.
- Planned B2B fields: account stage, price tier, payment terms, open quote value and age, days since last PO, category ordered, buyer role
- One definition, reused by Accounts, Callio campaigns and catalog personalization — never three drifting copies of the same rule
- Preview will count companies, buyers and unresolved sessions separately, rather than blending them into one flattering number
Custom segment builder
Build a segment with nested AND/OR rules over customers and visitors. Save it once and reuse it across Segments, Audiences and Personalization.
Segment name
High-intent Facebook visitorsSave as
SegmentThe visitor intelligence that ships today for DTC
These are the shipped Comver AI Commerce screens the buyer view will be built from: sessions resolved to people with consent-aware reachability, an intent score itemised point by point, and a library of audiences scored against the store’s own data.
Every session resolved to the person behind it
Visitor, when, source, session, intent, contactable, purchases and lifetime value — with unknowns honestly labelled “Anonymous visitor #a1b2c3” rather than guessed. Buyers adds the company and role columns to this table.
Known customer
Anna Kowalska
Known customer · Matched to a synchronized customer record
Customer ID: 8213004119
anna.k@…
Consent to contact
Phone
+48 600 ••• 118
No SMS consent
Purchased before
Yes
3 orders
Lifetime value
$412
Last order 4 Aug
Visitor history
4 sessions
Segment intent 68/100
Acquisition source
Google Ads
Churn risk 21/100
An audience is only useful if somebody calls it
A segment in Comver is not a report you export. Today one click turns it into a Callio campaign, a journey, an on-site experience or a personalization audience, and the reach counts tell you in advance how many of them can actually be emailed or texted. For buyers, the same click should point a campaign at the accounts behind the audience — call the buyers who opened the price list and never quoted — or hand the list to the rep who owns those companies.
- Callio calls first, texts when the call cannot connect and emails when the phone cannot reach at all — from numbers and domains your store owns
- Reach is counted from real identifiers and consent; recoverable revenue stays modelled and marked ≈, and overlapping audiences are never summed
- Anything the agent should not settle — a price exception, a disputed invoice — goes to the rep who owns the account, transcript attached
Key account gone quiet · calling the buyer from a number your store owns
“Hi Marta, it’s Callio. It has been a while since your last order — shall I raise the usual quantities and text you the order link, or would you rather your rep called back?”
Account campaigns · planned presets
Every preset calls first, texts when the call cannot connect and emails when the phone cannot reach — the cascade Callio runs for shoppers today.
Counted reach, modelled revenue — and nothing until real sessions arrive
The rules that make the shipped product trustworthy carry over to buyers unchanged: nothing is shown before real sessions exist, unknowns stay unknown, and a modelled figure always says so.
0
Sample sessions
Nothing is shown until real tracked visits arrive — no demo buyers, no seeded accounts
≈
Recoverable / mo
Always modelled from the store’s own rates, never counted, and never summed across overlapping audiences
“unknown”
What an unresolved session says
A session that cannot be matched to a company will say so, rather than being attributed to the nearest one
0
Invented B2B figures
No B2B measurement exists yet. Everything here describes the shipped DTC engine or what the twin will do
Reachability is consent-aware today — Email and SMS read yes, no consent, none or unknown, from the same engine the audience reach counts use — and buyer reachability will be computed the same way.
Frequently asked questions
No. What ships today is Visitors: every tracked storefront session resolved to a person, scored for intent and grouped into 56 audiences. Buyers — the same view resolved to companies and roles, scored for reorder and quote intent — is on the Comver AI B2B Commerce roadmap and no date has been announced.
Through the buyer. The identity engine already binds a session to a person when an event carries a customer id or an email; the B2B layer adds the company that person orders for, from the account portal login, the company record or an approved domain. When none of those exist, the session stays unresolved and says so.
Catalog depth, price-list opens, spec-sheet downloads, SKUs compared, quote-form starts, whether a reorder is overdue on the account and whether the buyer is resolved — with each signal shown with its points and any detected stall named. It keeps the property that matters today: you can read the number line by line instead of trusting it.
The 56 DTC audiences are. The B2B audiences described here — “Viewed price list, no quote”, “Compared SKUs across a category”, “Returning buyer, no reorder in 60 days”, “Quote started, never submitted”, “Ordered below the next price tier” — are designs on the roadmap, and their reach and recoverable figures will only ever be computed from your own rows.
That is the plan: the same engine, with account fields added — account stage, price tier, payment terms, open quote value and age, days since last purchase order, buyer role. Today it has 28 fields across customer, purchase, source, product, cart, visitor and predicted signals, nested AND/OR groups three levels deep and twelve operators.
Reachability is consent-aware and will stay that way: Email and SMS read yes, no consent, none or unknown, computed by the same engine that counts audience reach, so the table and the audience card can never disagree. Unresolved buyers can be worked on-site — catalog personalization — but not called.
Accounts is who the company is and what to do about it — the derived pipeline. Buyers is who was in the catalog, which company they buy for and which audience they belong to. Behaviour is what they actually did: the quote funnel, catalog heatmaps, spec-sheet and quote-form stalls, and privacy-safe replays.
No. Every B2B mock-up on this page is a concept preview, labelled as such, with illustrative figures. The proof band describes Comver AI Commerce for Shopify DTC as it ships today, and each stat carries the sentence that says how it is counted.
Keep exploring
See which company was in your catalog this morning
Buyers is on the Comver AI B2B Commerce roadmap. Join the waitlist to help decide what ships first — or start today on the DTC engine it will be built from.
14-day free trial · Live on Shopify in minutes · Checkout stays in Shopify