Paid media · Field noteGTMhub method

Attribution needs a qualified outcome

An attribution model divides credit. It cannot decide whether a conversion was useful to sales. Define that outcome before changing how credit is assigned.

The source and the change

Google’s December 2022 announcement described data-driven attribution using account history to distribute conversion credit across ad interactions. It also introduced more ways to inspect the model’s effects. That is the source event covered here, not the date this GTMhub analysis was first published. Original Google announcement.

The decision for a B2B team

Start with the outcome that receives the credit. A model trained against completed forms can give a sophisticated account of how those forms happened while saying little about qualified pipeline. Attribution and qualification solve different problems. The first assigns credit; the second determines whether the business event belongs in the number you are optimising.

Our recommendation is to trace one accepted opportunity back through the measurement chain before changing an attribution setting. Check the account match, accepted stage, value, event time, conversion action and reporting period. Repeat with an inquiry sales rejected. The two should remain distinguishable all the way to the campaign decision.

What to inspect before changing budget

QuestionUseful evidenceDecision it supports
What counts as qualified?A sales-accepted stage with an owner and explicit exclusionsWhich event should inform optimisation
Are outcomes mature?Cohorts grouped by acquisition date and expected sales lagWhether a recent campaign can be compared fairly
Is value meaningful?Won-deal values or a documented stage-value modelWhether value-based decisions are justified
Are events distinct?Stable event IDs and reconciliation countsWhether reported growth is actually duplication

Use it in your own account

Write down one pipeline definition with sales and finance. Map the qualified event into the measurement contract. Keep the current campaign history and changes that already work. Compare a bounded change against a mature cohort rather than treating a before-and-after platform screenshot as causal proof.

A missing outcome is not a zero-value outcome. If the CRM does not yet contain the result, keep it unknown and report the lag. Otherwise the newest cohort may look worse simply because sales has not had time to work it.

Current implementation note

For a new Google Ads integration, use current platform instructions rather than the API setup implied by an old announcement. Google changed offline-import access for new adopters in June 2026; our offline conversion workflow explains the current path. The durable lesson from the 2022 source is to agree the commercial event before arguing about its share of credit.

Publication and source record

Published on GTMhub: . Last reviewed: .

Original source: Turning data into results with data-driven attribution, . This is a retrospective GTMhub analysis of that source.

How we run this

Put it to work in your company.

Our paid media lane plugs into your company’s GTM brain. Start with one lane; services can run in parallel.

Book a call
  1. Week 1Your GTM brain

    Monday kickoff. A 45-minute review Friday to confirm the context.

  2. Weeks 2–3Connect the lane plugin

    Wire it into your existing tools and test the work with your team.

  3. Week 4Live, then improve

    About three hours of your time in month one, then 15 minutes a week.

The first term is three months. The brain and lane plugin are handed over in full after it. Your accounts and data stay yours.