Skip to content

Attribution for performance marketers

Every channel reports that it is working. They cannot all be right.

You are moving budget weekly on numbers that overlap, and the platform reporting the spend is also the one grading it. Soran grades it from your orders instead.

  • Campaign
  • Ad set A
  • Ad 1
  • Ad 2
  • Ad set B
  • Ad 3

The channel slipped. Which ad did it? Four levels in one pass.

Channel, campaign, ad set, and ad are tabs in one hierarchy. Tick rows on one and everything below narrows instantly, because the whole report already arrived.

  • Last touch
  • Direct
  • Time decay
  • Linear
  • Position based
  • Participation
  • First touch
  • least creditmost

One model, and no way to stress it. Disagreement is the signal.

Strong on last touch and weak on first touch means a channel is harvesting demand, not creating it. The gap tells you which budget is safe to cut.

MetaTikTokEmailDirect

Revenue with no spend against it. Shown, not quietly absorbed.

A channel earning with nothing spent is almost always an ad account nobody connected. You see the gap instead of a flattering ROAS.

#1042 1.00
Meta
0.50
Email
0.30
Search
0.20

The platform reports the spend and grades it. Your orders do the grading.

Attribution is calculated from the sessions that led to each order, against the spend Soran pulled from the accounts. Neither half of that comes from the party being marked.

None of this works on data Soran cannot see. It reads the platforms you already run.

See all integrations

Stop grading the platforms on their own homework.

Stop guessing which ads worked. Start seeing it.

  • No credit card to start
  • Every report and all ten models from day one
  • Your data exports whenever you want it
  1. 01

    Connect your store

    Shopify authorizes in a click and checkout tracking installs itself.

  2. 02

    Pick a goal

    The conversion you actually get paid for. Soran credits every touch that led to it.

  3. 03

    Read your first report

    Spend against the revenue it earned, on all ten attribution models.