All revenue goes to the last click
Brand search and retargeting take the whole result, discovery channels look unprofitable — and the first budget cut lands on the channel that brought the people in.
Five attribution models on the same data
Valkiria collects touchpoints from the UTM tags on your orders and splits revenue across channels, sources and campaigns under all five attribution models. The comparison table shows which channel holds up everywhere and which one only exists under last-click.
Why last-click quietly kills the top of the funnel
Every team has its own number
Each channel counts its own conversions, the totals across reports exceed real revenue, and the argument with the agency comes down to whose spreadsheet is righter.
There is nothing to test the conclusion against
One model gives one picture. Whether the conclusion is robust or only holds under the model you happened to pick is invisible in a single report.
How the attribution report is built
Tag your links with UTM
Orders carrying utm_source, utm_medium or utm_campaign become touchpoints; the channel is derived from the tags — paid search, social, email, affiliate, organic.
Pick a model and a period
Five models and a date range, last 30 days by default. The whole report is recalculated for the model you choose.
Compare the models
Comparison mode puts all five models in one table, so each channel shows its share of revenue under every one of them.
What five-model attribution gives you
Five models out of the box
First touch, last touch, linear, time decay and position based, all switchable on the same set of touchpoints.
Models side by side
A channel × five models table shows the spread: a channel that holds everywhere is a different thing from one that lives in a single model.
Breakdown down to the campaign
Revenue, conversions and share are counted by channel, source and campaign, not only at the top channel level.
Weights you can explain, not magic
Time decay uses a seven-day half-life; the position-based model gives 40% to the first touch, 40% to the last and 20% to the middle.
An honest empty screen
With no touchpoints the system says so and points to tracking setup; a loading error is shown separately instead of as zero revenue.
Read-only
Attribution changes nothing in your ad accounts: it explains where the revenue came from and leaves the decision to you.
In numbers
- 5
- attribution models on the same dataset
- 7
- days — the half-life in the time-decay model
- 0
- invented touchpoints: only real UTM tags count
Questions about attribution
Which attribution models are supported?
Five: first touch, last touch, linear, time decay and position based. All of them run on the same set of touchpoints, so any difference in the numbers comes from the model rather than from the data.
Where do the touchpoints come from?
From your orders carrying UTM tags: anything with utm_source, utm_medium or utm_campaign filled in is counted. The channel is derived from those tags — paid search, display, social, email, affiliate, referral and organic.
Do I need a separate SDK or a CRM integration?
No. UTM-tagged ad links plus working conversion tracking are enough for orders to arrive with their tags. A dedicated CRM integration is not required for the basic report.
Why do the models disagree?
Because they split the same revenue between touchpoints differently: last touch gives everything to the final one, linear splits it evenly, time decay favours recent touches. The spread between models is itself the useful information.
Does attribution change bids or budgets automatically?
No, the report only counts. What to do about a channel is your call, and changes are made in the campaign management modules, where anything new is created PAUSED and needs approval.
Test your conclusion with five models
The report is built from your UTM-tagged orders and changes nothing in your ad accounts. Model comparison is available as soon as the first touchpoints land.