What marketing attribution is
A marketing attribution model is the rule that decides how much credit each channel gets for a conversion. Think of a B2B buyer: they see a LinkedIn ad, a few days later search for your brand on Google, download a guide from your newsletter and, a month after that, request a demo after a Google Ads click. Who brought in that lead? And the sale?
Each model gives a different answer, and budgets move with it. That is why multi-touch marketing attribution, which counts every touchpoint and not just the last one, matters more the longer your sales cycle is.
Types of marketing attribution models
Attribution in Google Ads and GA4: which models are left
On 6 April 2023 Google announced it was retiring the first click, linear, time decay and position-based models. In Google Ads they could no longer be selected from June 2023 and, from September 2023, conversions still using them were switched to data-driven attribution. In GA4 they stopped being available in November 2023, according to its help centre.
What is left today:
- Google Ads: data-driven, which is the default, and last click.
- GA4: data-driven, paid and organic last click, and Google paid channels last click.
Two things to bear in mind. Google Ads data-driven attribution shares credit among interactions with Google ads; it doesn't see your Meta or LinkedIn campaigns. GA4's model is cross-channel, but it only sees what passes through your tagged website: not impressions on other platforms and not what happens in the CRM. If you want a U-shaped, W-shaped or first-click model, today you have to calculate it yourself or use a CRM that does.
HubSpot marketing attribution models, and which plan you need
HubSpot has its own attribution reports and, according to its knowledge base as of October 2026, they are split like this:
- Contact create attribution: Marketing Hub or Content Hub, on Professional and Enterprise.
- Deal create and revenue attribution: Marketing Hub Enterprise only.
The available models are first interaction, last interaction, linear, U-shaped, W-shaped, time decay, J-shaped, inverse J-shaped and full path. W-shaped is only available in deal create and revenue reports, and full path (22.5% each to the first touch, lead creation, deal creation and the closing touch, with 10% for the rest) only in revenue reports.
It works well if your CRM is HubSpot, its tracking code is on your website and your ad accounts are connected. What HubSpot doesn't record, it can't attribute.
B2B attribution with your CRM: from lead to closed-won
In B2B the conversion that matters isn't the form fill but the won opportunity, which arrives weeks or months later and off the website. GA4 and Google Ads attribute leads; the business wants to know which channel brings in revenue.
That means joining the two worlds. When a lead comes in, the CRM stores its source (UTMs, gclid, fbclid) and that browser's GA4 identifier. The lead is then followed through to the opportunity, and the won amount is attributed, not the number of forms. With that data, channel spend is compared with real revenue, and CAC by channel stops being an estimate.
How to build a marketing attribution model on BigQuery with GA4 and your CRM
- 1Export GA4 to BigQueryEach event arrives with user_pseudo_id, collected_traffic_source (the event's UTMs and gclid) and session_traffic_source_last_click, the session source by last click.
- 2Store the identifier in the formWhen a lead submits, the form sends the GA4 identifier and the UTMs to the CRM. Without this step there is no reliable way to match visits to the lead.
- 3Load the CRM and costsLeads, opportunities, amounts and close dates from the CRM; daily Google Ads and Meta Ads spend from their connectors.
- 4Build the pathsOne table with every touchpoint for each lead in date order, up to when the opportunity is created or won.
- 5Calculate several models at onceSQL gives you first click, last click, linear, U-shaped, W-shaped and time decay on the same paths, so you can compare them.
- 6Put it in the dashboardAttributed revenue and spend by channel in Looker Studio or Power BI, with the model as a filter.
Marketing mix modeling: when click attribution isn't enough
The models above follow people through their clicks. Marketing mix modeling (MMM) does something else: it relates weekly spend on each channel to sales, accounting for seasonality, price or promotions, without tracking anyone. It also measures what leaves no click, such as TV, radio or brand, and doesn't rely on cookies.
- Meridian is Google's open-source MMM, open to everyone since 29 January 2025. It is Bayesian, supports Google search query volume and reach and frequency for video ads, and runs in Python. In May 2026 Google announced it is bringing it into Analytics 360, with no date yet.
- Robyn is the open-source MMM from Meta's Marketing Science team, which describes it as experimental. On GitHub its latest published release is 3.12.0, from December 2024.
An MMM needs a fair amount of spend and sales history, and it pays off when budget is spread across many channels, including ones you can't measure by clicks. For a B2B company with three or four digital channels, CRM-based attribution usually gives more back first.
Which attribution model to choose
Marketing attribution FAQ
What is a marketing attribution model?
The rule for sharing the credit for a conversion or a sale among the channels and ads a customer interacted with before buying.
Which attribution models does Google Ads offer today?
Data-driven, which is the default, and last click. First click, linear, time decay and position-based were retired in 2023.
What is multi-touch attribution?
Attribution that counts every channel on the path, not just the last click or a single platform.
Does HubSpot do revenue attribution?
Yes, but only on Marketing Hub Enterprise. Professional gives you contact create attribution.
How is marketing mix modeling different from multi-touch attribution?
Multi-touch attribution follows each person's clicks. MMM works on aggregated spend and sales data and also measures channels without clicks.
What is the best attribution model?
None is right on its own. What helps is choosing one to make decisions with, calculating it on real CRM revenue and comparing it with others.
Want to know which channel brings in sales, not just leads?
In a 30-minute call we look at what you have in GA4, your campaigns and your CRM, and whether it can be joined to attribute revenue by channel. If we can't help, we'll tell you that too.
Book the 30-minute call