HubSpot attribution reporting: every model explained (with use cases)

HubSpot's report builder currently supports five attribution models: First Touch, Last Touch, Linear, Time Decay, and Empirical. HubSpot's documentation, updated August 18, 2026, states that Empirical replaces the U-shaped, W-shaped, J-shaped, and Inverse J-shaped models. A separate HubSpot knowledge base page on interactions and reporting dimensions still documents U-shaped and W-shaped as selectable options. If your portal's report builder still shows them, it hasn't synced to the newer guidance yet.

This guide covers all of them, current and legacy, with the decision each one answers well. It also covers the part that trips up most SaaS growth teams: attribution model math matters less than whether your paid, outbound, content, and event touches made it into HubSpot at all. When paid, outbound, and creative sit with separate vendors, nobody owns the handoffs between them, and those gaps show up in HubSpot as missing credit.

What HubSpot attribution reporting measures, and what it costs

Three report types, each tied to a funnel stage and a pricing tier:

• Contact Create Attribution credits the interactions that led to a new contact. Available on Marketing Hub Professional and Enterprise.
• Deal Create Attribution credits the interactions that led to a deal being created. Marketing Hub Enterprise only.
• Deal Revenue Attribution credits interactions on closed-won deals. Marketing Hub Enterprise only.

The tier gating is the first hard truth. For a 20K-100K ACV motion, the two reports a board cares about, pipeline created and revenue closed, need Marketing Hub Enterprise, priced around $3,600/month for 5 core seats. Professional, which only provides access to Contact Create Attribution, runs about $890/month for 3 core seats.

The five current models

First Touch

100% of credit goes to the contact's first recorded interaction, usually a website visit. Use it in Contact Create reports to judge paid channel efficiency on net-new names and to understand how people discover your brand. In a 3-9 month sales cycle, it ignores every nurture touch that followed. Present it as "sourced," never as the whole story.

Last Touch

100% of credit goes to the last interaction before conversion. It's a reasonable closing diagnostic in a Deal Revenue report, since it values bottom-of-funnel actions. It over-credits retargeting and late-stage content, so don't set budget off it alone.

Linear

Equal credit to every interaction on the path. It's a reasonable primary model for mid-funnel content and nurture ROI, since a trade-show meeting doesn't get the same weight as one of twenty page views that way: it just spreads credit evenly across all of them, which is its own distortion. Pair it with a Campaign or Interaction Source dimension in Deal Create when you need sales activity and marketing touches on the same chart.

Time Decay

More credit to interactions closer to conversion. HubSpot's own blog documents the formula behind it: y = 2^(-x/7), a seven-day half-life, meaning a touchpoint's credit roughly halves every week. The current attribution-reports documentation doesn't restate that half-life or offer a setting to change it.

On a six-month enterprise deal, a seven-day half-life gives almost nothing to the webinar that started the cycle. Time Decay fits short, time-boxed campaigns, like a product launch or a fixed-window promotion, better than multi-quarter enterprise cycles. Test it against Linear before you put it on a board slide.

Empirical

Empirical assigns credit this way:

• Credit is weighted by how often each interaction type appears across your conversion paths.
• Rarer interaction types get more weight than common ones.
• Individual interactions of the same type split that type's credit evenly.

HubSpot publishes no fixed weights and no minimum sample size for the weighting to stabilize. It's the current, direct replacement for the retired shaped models. Cross-check it against Linear until you trust what it's telling you.

The legacy shaped models (and their published weights)

If your saved reports or your portal still show these, here's what they do, based on HubSpot's published definitions:

• U-shaped: 40% to the first interaction, 40% to the lead-creating interaction, 20% spread across everything in between.
• W-shaped: 30% each to the first interaction, the lead-creating interaction, and the interaction that created the deal, with the remaining 10% spread across the rest. Deal Create and Revenue reports only.
• Full Path: 22.5% each to the first interaction, lead creation, deal creation, and the last interaction before closed-won, with 10% spread across the rest. Revenue Attribution only.
• J-shaped and Inverse J-shaped: weighted toward early interactions (Inverse J) or late interactions (J), respectively. HubSpot's current documentation doesn't publish an exact split for either model, and we couldn't find one published elsewhere.

HubSpot's August 18, 2026 documentation says Empirical replaces all four of these. Full Path's status is murkier: it isn't named in that replacement sentence, and it isn't listed as selectable on the current page either. HubSpot hasn't said what happens to saved reports built on any of the retired models.

Which model for which decision

Decision

Report

Primary model

Cross-check

Top-of-funnel paid efficiency

Contact Create

First Touch

Empirical

Which sources create pipeline

Deal Create

First Touch for origin

Linear or Empirical

Mid-funnel content and nurture ROI

Deal Create

Linear

Empirical

Crediting SDR outreach

Deal Create or Revenue

Last Touch (diagnostic) + Linear

Time Decay

Board pipeline-influence slide

Deal Create

Linear ("influenced") + First Touch ("sourced")

W-shaped if your portal still has it, or Empirical

Closed-won revenue reporting

Deal Revenue

Linear (Full Path on legacy portals)

First and Last Touch side by side

Run several models against different questions and read them as relative performance, not precise return.

The hard part: coverage, not the model

Most HubSpot attribution arguments inside SaaS teams are about model choice, and most of the damage happens upstream before any model runs. Seven HubSpot gates determine what the models can see:

• Deal eligibility and associations: Revenue attribution excludes deals missing Amount, Create date, or Close date, and excludes deals with no associated contact. Calls and meetings only count once they're logged to both the contact and the deal record, not just the contact.
• Sales email logging: Emails sent through external b2b outbound marketing tools like Instantly or HeyReach only get credit once replies sync back into the HubSpot contact timeline. HubSpot's own documentation confirms that email sent from outside the platform isn't credited unless it originates in HubSpot.
• Marketing email identity: Marketing email clicks count only when the email was sent to the contact's Primary email address. Clicks from an email sent to a Secondary address aren't attributed.
• External interaction tracking: Any interaction outside HubSpot needs a UTM-tagged link to a page carrying the tracking code, or it's excluded. Untagged paid traffic lands in Direct, and so does Google Ads traffic when auto-tagging is off or a redirect strips the gclid. UTM values are case-sensitive, so "linkedin" and "LinkedIn" register as two different sources.
• API and import tracking: Contacts created by API, or created before the tracking cookie existed, stay under Offline Sources, and imports don't backfill ad interactions. Routing contact creation through the Forms API with the hubspotutk cookie attached is the standard fix for sessions created outside the normal form flow.
• Buying-group coverage: Revenue Attribution is contact-level by default, but B2B buying decisions involve multiple people. Contact Create Attribution often credits whoever filled out the form first, frequently a researcher screening vendors rather than the economic buyer.
• Dark social coverage: When we compared HubSpot-tracked attribution against a mandatory self-reported "how did you hear about us" field for our dark social attribution work, HubSpot credited web search with 78% of conversions versus 12% in self-report. Dark social got close to 0% of software-tracked credit despite accounting for 98% of self-reported closed-won revenue.

Together, these seven gates determine how much of the real buyer journey any model can actually see. If your paid, outbound, and creative vendors don't share a UTM convention, a deal-association rule, or a logging habit, HubSpot doesn't throw an error. It files the touches under Offline Sources or drops the deal, and your board slide reports a fraction of what happened.

One UTM convention across every vendor is what allbound marketing coordination looks like inside HubSpot. At Understory, we start every engagement by printing the report type, the model, and a coverage percentage, meaning attributed revenue against total closed-won revenue for the same period, on every attribution report we hand a client.

Setup checklist before trusting any model

Use this checklist to close the most common tracking and association gaps before you evaluate a model. For the deeper data-hygiene playbook, lifecycle stage protection, workflow conflicts, UTM governance, see our HubSpot attribution setup guide.

1. Tracking code installed on every externally hosted page, with external domains added to report settings. Declined cookies mean no page views on the record.
2. Ad accounts connected (Meta, Google, LinkedIn, Microsoft) with auto-tracking left on. Clicks made while it's off are never attributed retroactively.
3. One lowercase UTM convention, built in HubSpot's tracking URL builder, applied by every vendor who touches a link.
4. Contact association required at deal creation. Set a workflow alerting RevOps when a deal has no contact within 48 hours.
5. A free-text "How did you hear about us?" field on your highest-intent form, copied from contact to deal with an "Edit records" workflow, and asked again on the discovery call. Pair it with software attribution and flag disagreements instead of forcing one source of truth.

Once these controls are in place, model comparisons are more likely to reflect channel performance instead of tracking gaps.

Get clean HubSpot attribution with Understory's allbound team

At Understory, we run LinkedIn, Google, and Meta ads, Instantly-powered outbound triggered on signals like a recent CRO hire, a funding round, or a tech-stack change, and on-staff creative for B2B SaaS clients under one team. That means one UTM convention, one logging standard, and deal associations that don't break Deal Revenue Attribution. That coordination is how we helped Rivial Security scale paid spend from $20K to $70K a month without losing sight of what the spend was producing, and we're glad to work alongside an existing paid media partner when outbound and creative are the gap.

Book a consultation with Understory to audit your HubSpot attribution setup and close the coverage gaps.