Attribution has a credibility problem. Marketers present a slide showing that paid social drove 42% of revenue, a finance director asks how that number was produced, and the room discovers nobody can explain the mechanism. HubSpot's attribution is genuinely good, but only when the people using it understand what it is measuring and what it cannot see.
This is the mechanism, end to end.
Step 1: the click becomes an identity
When someone clicks a Google ad, the landing page URL carries a click identifier — gclid for Google, fbclid for Meta. HubSpot's tracking code captures those parameters along with your UTM values and stores them against the visitor's cookie. The moment that visitor submits a HubSpot form, chats, books a meeting or is otherwise identified, those values are written onto the contact record.
That stored identifier is the linchpin of the entire system. It is what allows HubSpot to say, months later, "this deal traces back to this ad", and it is what allows HubSpot to push a conversion back to Google or Meta so their bidding models learn from the outcome. If tracking parameters are stripped by a redirect, a third-party form, or a landing page on a domain HubSpot is not tracking, the chain breaks at step one and no reporting configuration downstream can repair it.
Step 2: interactions accumulate on a timeline
From then on, every tracked interaction attaches to the same contact: page views, email opens and clicks, form submissions, meeting bookings, chat conversations, ad interactions from connected accounts, and sales activity logged by the team. This is the raw material attribution works from — a chronological list of touchpoints with sources, campaigns and timestamps.
HubSpot classifies each touch by type — for example ad interaction, organic search, direct traffic, email, sales activity — and by asset, so credit can later be assigned not just to Google Ads but to the specific campaign and landing page involved.
Step 3: two different attribution questions
HubSpot answers two questions with two different report families, and conflating them is the most common analytical error we see.
- Contact create attribution: which interactions were responsible for a person becoming a contact. This is a lead-generation view and it is available on lower tiers.
- Revenue attribution: which interactions were responsible for a deal closing, weighted by deal amount. This is the commercial view, and it requires deals to be properly associated with contacts.
A channel can dominate contact creation and contribute almost nothing to revenue. Cheap top-of-funnel content offers do this constantly. Report both, and make revenue the one that drives budget decisions.
Step 4: choosing the model
The model determines how credit for a deal is spread across the touches on the path. HubSpot offers six, and each encodes a different belief about how buying works.
- First touch — all credit to the interaction that introduced the buyer. Use it to evaluate demand generation.
- Last touch — all credit to the final interaction before the deal. Use it to evaluate demand capture.
- Linear — credit split evenly across every touch. The most neutral, and a good default for long journeys.
- U-shaped — 40% to first touch, 40% to lead conversion, 20% spread between. Good when acquisition and conversion are the two moments you manage.
- W-shaped — 30% each to first touch, lead conversion and deal creation, 10% spread. Suits considered B2B sales with a defined mid-funnel.
- Time decay — credit weighted towards interactions nearest the close. Suits short, fast cycles.
Pick two models, not one. A channel that looks strong under first touch and weak under last touch is not underperforming — it is doing a different job.
Step 5: closing the loop back to the ad platforms
Attribution reporting is the visible half. The half that actually changes performance is offline conversion sync. When a contact reaches a lifecycle stage you nominate, or a deal reaches a chosen stage, HubSpot sends that event — with the click identifier and, for deals, a value — back to Google or Meta.
The platforms then optimise towards the campaigns, keywords, audiences and creatives that produced those events rather than the ones that produced the most form fills. This is where measurement stops being reporting and starts being a performance lever, and it is the single highest-return thing most advertisers using HubSpot have not switched on.
Setting it up so the numbers hold
- 01Install the HubSpot tracking code on every domain in the journey, including any separately hosted landing pages.
- 02Connect Google Ads, Meta and LinkedIn as ad accounts inside HubSpot, and confirm click identifiers are appearing on new contacts.
- 03Standardise UTM parameters with a written convention — lower case, no spaces, one taxonomy — and use auto-tagging for Google.
- 04Define lifecycle stages in writing with the sales team, including the exact criteria for MQL and SQL, and enforce them with automation rather than judgement.
- 05Make sure deals are always associated with the originating contact; unassociated deals are invisible to revenue attribution.
- 06Turn on offline conversion sync for the stage that best balances quality and volume.
- 07Build one dashboard per audience: a channel-performance view for marketing, a pipeline view for sales, a revenue view for the board.
What attribution cannot see
Be candid about the blind spots, because credibility depends on it. Dark social — a link forwarded in WhatsApp, a recommendation in a private community — arrives as direct traffic. Consent settings and tracking prevention remove some paths entirely. View-through influence, where somebody watched your video and never clicked, is invisible by design. Offline conversations, referrals and word of mouth only appear if someone logs them.
This is why we treat attribution as one input among three: HubSpot's models for path-level detail, self-reported attribution on the form asking how the buyer found you, and periodic incrementality tests for causal truth. Where all three agree, you can act with confidence. Where they disagree, you have found the question worth investigating.
Done properly, HubSpot attribution ends the quarterly argument about which channel deserves credit and replaces it with a defensible view of what produced revenue — and, more usefully, a signal loop that makes the next quarter's media buy smarter.

Written by
Danielle Magbanua
HubSpot Specialist
Danielle architects CRM and automation inside HubSpot — lifecycle stages, pipelines, lead routing and reporting that sales teams actually trust and use.


