Causation
Measurement describes; it does not prove cause
A number rising after the measurement did does not mean measurement caused it: season, stock, price and creative all move together. Raasid shows what changed and when; why stays yours.
The methodology
How a number reaches your dashboard
Before you move a budget on a number: how it was computed, over what denominator, and where it stops — including what this measurement cannot tell you.
Every number on this panel has one written definition and one written boundary — all five, in full.
The measurement chain
Six links, and any of them can be short
The number is the output of six steps. Knowing them changes what you do about a low one: look at the weak step, not the whole figure.
- Step 01
The event arrives
A visit, a view, a cart, a checkout, an order — from your store's server, not the browser, so no blocker or closed tab decides what was recorded.
- Step 02
The identity is resolved
Scattered events gather onto one person using the identifiers available — email, phone, click ID. The more complete, the fewer sessions stay unowned.
- Step 03
The order is matched
The store's order is matched to its event. One with no counterpart stays in the denominator and visible, not dropped.
- Step 04
The ad interaction is checked
A qualified ad interaction is looked for inside the window. None means a measured order with no channel — shown as it is, never spread by guesswork.
- Step 05
The order is attributed
One channel, one campaign, so an order is never counted twice. Platforms may each claim the same one — a leading reason the numbers differ.
- Step 06
The numbers are aggregated
Attributed orders roll up into campaigns, products and periods with spend divided over them — and every aggregate opens into its rows.
What never arrived
An order that never reached us is not recovered
An order that never reached us — broken connection, unconnected channel, a period before activation — appears as a known gap, neither inferred nor estimated.
Off your store
What happens inside the platform is not visible here
Views, in-app engagement and modelled conversions happen inside the platform. We show what it reports beside our numbers, and claim no way to verify it.
The limit of identity
A new device or a private window breaks the chain
Start on one device, finish on another with no shared identifier, and one shopper arrives as two. Identifier completeness is on the dashboard because it is this step's ceiling.
How much of the measurement completed at all is a metric in its own right — identifier completeness and coverage.
The limits, in writing
What this measurement cannot tell you
Four things you would test us on, written down before you ask. A methodology that does not know its edges is not one to set a budget against.
Why they disagree
Two correct numbers, measured two different ways
Why is the platform's number bigger than my store's? Four mechanical reasons — and neither has to be wrong to differ.
One journey, seen by several platforms
Saw an ad here, clicked another there, then bought. Each credits the order to itself, so their total exceeds your order count with no one wrong.
The attribution windows are not the same
Seven days after a click here, twenty-eight there, a view with no click at a third. Your period is one; their windows are not.
Part of the figure is modelled
When signal is short, platforms estimate conversions rather than leave them at zero — stated openly, but part of the figure is an estimate, not an event in your store.
The moment of purchase is not the order's outcome
A platform records the purchase as it happens; confirmation, cancellation and return are outside its reach. Raasid reads the order in your store's state.
The connection, last on purpose
What it actually takes to run this
The store through its platform's app, each channel through its own official login — no tokens, no theme edits. Measurement starts at connection, never retroactively.
app.raasid.app/onboarding/platforms
Connect your ad platforms
Click Activate on each platform - a secure one-click login. No tokens to copy, no developer.
MetaSynced 2 min ago
SnapchatSynced 5 min ago
GoogleSynced 8 min ago
TikTok
The channels read today
Questions about the methodology
Do you promise perfectly accurate attribution?
No, and we do not know anyone who can. Attribution is an inference about a journey no one sees in full. What we do commit to is different: one written definition per number, a stated denominator, visible gaps, and the ability to walk down from any aggregate to the orders inside it.
What is the difference between an attribution tool and Raasid?
An attribution tool reads data that already arrived and splits credit between channels. Raasid starts from your store's own orders, delivers events to the platforms server-side, then shows what it could confirm beside what the platform reports. The two can run together.
Is anything before the connection measured?
No. Measurement starts at connection. Earlier periods stay with the platforms' numbers alone, and we say so rather than filling them with an estimate.
The next step
Read it on your own store's numbers
Connect your store and one ad channel, then compare what Raasid confirms against what the platform reports over a period you know well.