Missing is not zero.
When a source hasn’t synced, its value stays unknown and named. It never quietly becomes $0 in a total.
Build on attribution
evidence, not totals.
weive treats the outcome, the touches, the rule, and the coverage as one piece of evidence, so a product built on it never shows a cleaner number than the data supports.
RecordsEvidence
Google AdsTouches
MetaTouches
StripeOutcome
weive pixelVisitsEvidence · ord_1042$300Derived · Linear rule
OutcomeStripe
Touches3
RuleLinear
CoverageComplete
A number without its method is just a claim. Build with weive and your product receives the method too: what was observed, what was derived, and what’s still unknown.
Change the rule.Watch the credit move.
Every evidence object can be re-split under any of the six rules. This is the same computation a partner product would show.
- Google Adstouch · search click
- Metatouch · ad click
Directtouch · visit- ord_1042Outcome
Outcome · Stripe$300
Google Ads—
Meta—
Direct—
One object.The whole story.
The outcome, its touches, the rule, the evidence class, and the coverage arrive together, so nothing about the method gets lost in transit.
- Evidence class on every value
- Rule and touches included
- Coverage travels with the data
{ "outcome": { "source": "stripe", "amount": 300.00 }, "rule": "linear", "touches": [ { "channel": "google_ads", "credit": 100.00 }, { "channel": "meta", "credit": 100.00 }, { "channel": "direct", "credit": 100.00 } ], "evidence": "derived", "coverage": { "state": "complete" }}
Observed, derived,estimated, unknown.
Each value carries its class, so your product can show a platform figure, a matched figure, and an estimate without ever merging them.
- Class on every value
- Estimates labelled as estimates
- Unknown returned as unknown
Evidence classesHow weive labels values
ObservedRead directly from a connected sourceSpend, paymentsDerivedCalculated under a stated ruleMatched revenueEstimatedModelled, assumptions statedKept apart
UnknownNot recorded anywhere—
weive evidence standard
Assistants answer betterwhen they know the limits.
Hand an assistant a bare number and it will repeat it confidently. Hand it weive evidence and it can say what the number means and what it doesn’t.
- Evidence and limits in one payload
- Rule stated with every figure
- Designed for assistants and agents
Summarise last month’s attribution for the team.
Matched revenue was $48,000 under Linear, led by Google Ads search. Every source synced. Credit describes how revenue was shared, not incremental lift.
Evidence: derivedRule: linearCoverage: complete
Three promisesbehind every number.
Credit is not cause.
Attribution shows how a rule shares credit across recorded touches. It never claims an ad caused the sale.
One order, counted once.
Two platforms can claim the same sale. weive credits the payment once, and keeps each claim visible beside it.
Three stepsto building with weive.
Tell us what you’re building.
Share the use case and the places where attribution evidence will appear.
Agree the evidence standard.
Evidence class, rule, and coverage shown wherever figures appear.
Build and review together.
We review the integration with you before it reaches customers.
Worth askingbefore you start.
Is there a public API?
Not yet. We are working with partners on how attribution evidence should be shared. Contact us to discuss your integration.
Which AI tools can connect?
Tell us which tools matter to your product, and we’ll work through the integration with you.
Does attribution prove a campaign caused the sale?
No. Attribution shows how credit is shared across recorded touches. weive keeps that separate from causal lift, so credit is never presented as proof.