Marketing Attribution Models: When to Use Each One | HubSpot

Marketing attribution is the practice of assigning credit for a conversion to the marketing touchpoints that influenced it. An attribution model is the rule that decides how that credit gets split. Pick the wrong rule and you will defund the channels that actually create pipeline. Pick the right one and you can finally answer the only question your CFO cares about: which marketing dollars turned into revenue.

This guide covers every attribution model HubSpot offers, the story each one tells, the real-world scenarios where it helps (and where it lies to you), how to read the data without fooling yourself, and how HubSpot Marketing Hub Enterprise actually services the reporting.

Quick answer: the nine attribution models at a glance

Single-touch models answer “what started or finished this?” Multi-touch models answer “who all helped, and how much?” Most teams need both.

What is marketing attribution, and why does the model matter?

Marketing attribution connects revenue back to the marketing interactions that produced it — the blog post someone read, the paid ad they clicked, the webinar they attended, the sales email they replied to. Because modern B2B buyers touch a brand many times before converting (commonly cited industry benchmarks put it in the range of 8 or more touches), you cannot hand 100% of the credit to one interaction without distorting reality.

That is what an attribution model fixes. It is a deliberate, transparent rule for dividing credit across touchpoints. The model you choose changes which channels look successful, which campaigns get budget next quarter, and which teammate gets the high-five. The model is not a math curiosity — it is a budgeting decision in disguise.

There is no single “correct” model. There is only the model that best matches your sales motion, your cycle length, and the decision you are trying to make right now. A demand-gen leader and a sales-ops leader can look at the same pipeline and correctly choose two different models.

Single-touch attribution models

Single-touch models give 100% of the credit to one interaction. They are simple, easy to explain, and genuinely useful for answering narrow questions — as long as you remember they ignore everything else that happened.

First interaction (first touch)

• What it is: All credit goes to the very first interaction a contact had with you.

• The story it tells: “What introduced this customer to us in the first place?” It is a pure measure of demand creation and audience discovery.

• When to use it: - You are evaluating top-of-funnel channels — SEO, paid social, content, PR, events. - Your primary goal this quarter is net-new audience, not closing. - You want to know which channels are best at finding people who eventually buy.

• When NOT to use it: - You are trying to judge what closes deals. First touch is blind to the entire middle and end of the journey. - Your cycle is long and your first-touch data is incomplete or pre-tracking (e.g., the contact existed before you installed tracking).

• Real-world scenario: A SaaS company notices that organic blog content gets almost no last-touch credit, so a naive analyst recommends cutting the content budget. First-interaction attribution reveals that 60% of closed-won customers first discovered the brand through that same blog. The content was not closing deals — it was starting them. Cutting it would have starved the top of the funnel.

• How to interpret the data: High first-touch credit on a channel means it is an excellent opener. It does not mean that channel is good at converting — pair it with a last-touch or multi-touch view before making spend decisions.

Last interaction (last touch)

• What it is: All credit goes to the final interaction before conversion.

• The story it tells: “What sealed the deal?” It is a measure of closing power and bottom-of-funnel effectiveness.

• When to use it: - You want to optimize the channels that finish the job — demo requests, bottom-funnel paid search, sales outreach, pricing pages. - You have a short, simple buying cycle where one or two touches dominate. - You need a quick, defensible number and do not have multi-touch reporting available.

• When NOT to use it: - You want to value the awareness and nurture work that made the final touch possible. - Your buyers do heavy research before that last click — last touch will over-credit “closer” channels like branded search and under-credit everything that built the demand.

• Real-world scenario: An e-commerce brand sees branded search winning nearly all last-touch credit and pours more budget into it. Growth stalls. The problem: branded search only captures demand that paid social and influencers created. By over-funding the closer and starving the creators, they ran out of new demand to capture. Last-touch told a true-but-incomplete story.

• How to interpret the data: High last-touch credit means a channel is a strong closer, often because it intercepts buyers who are already decided. Always ask, “What created the intent this channel is harvesting?”

Multi-touch attribution models

Multi-touch models split credit across several interactions, giving you a far more honest picture of a real buying journey. They are more accurate and more complex — and they require cleaner data and a higher HubSpot tier to use well.

HubSpot attribution model selector showing First interaction, Last interaction, Full-path, Linear, and U-shaped options

Linear

• What it is: Every touchpoint in the journey gets equal credit.

• The story it tells: “Everyone helped.” It is the most democratic, no-favorites view of the customer journey.

• When to use it: - Your sales cycle is long and genuinely collaborative, with many meaningful touches. - You want a balanced baseline before committing to a more opinionated model. - You are making the case that mid-funnel nurture matters and deserves budget.

• When NOT to use it: - Some touches clearly matter more than others (they almost always do). Linear flattens that signal. - You need to optimize a specific stage — linear will not tell you whether the opener or the closer drove the result.

• Real-world scenario: A B2B services firm with a 6-month cycle uses linear attribution to prove to finance that the 12 nurture emails between first contact and close are not “fluff.” Because each email gets equal credit, the cumulative value of the nurture program becomes visible for the first time, protecting its budget.

• How to interpret the data: Linear is best read as a participation map, not an importance ranking. If a channel shows up across many journeys with steady linear credit, it is a reliable workhorse even if it never “wins” a single touch.

U-shaped (position-based)

• What it is: 40% of credit to the first touch, 40% to the lead-conversion touch, and the remaining 20% split among everything in between.

• The story it tells: “Getting found and capturing the lead are the two moments that matter most.” It rewards both the opener and the moment of hand-raise.

• When to use it: - Lead generation is your core objective and the conversion-to-lead event is your key milestone. - You want to honor demand creation and lead capture without ignoring the middle entirely. - You run a marketing-led motion where becoming a known lead is the pivotal step.

• When NOT to use it: - Your most important moment is deeper in the funnel (opportunity creation or close) — U-shaped under-weights it. - Your “lead conversion” event is poorly defined, which makes the 40% landing somewhere arbitrary.

• Real-world scenario: A marketing team running gated content and webinars uses U-shaped attribution to balance two truths: the SEO article that first found the buyer and the webinar that converted them to a lead both deserve real credit. The model stops the team from over-rewarding whichever touch happened to be last.

• How to interpret the data: Watch the two “peaks.” If one channel dominates first-touch credit and a different one dominates conversion credit, you have a healthy two-stage engine. If the same channel owns both, you may be over-reliant on a single source.

W-shaped

• What it is: 30% each to three key moments — first touch, lead conversion, and opportunity creation — with the remaining 10% spread across other touches.

• The story it tells: “Three milestones make or break a B2B deal: discovery, hand-raise, and becoming a real opportunity.”

• When to use it: - You run a defined B2B funnel with a meaningful opportunity/SQL stage. - Sales and marketing share a pipeline and you need to value the touches that move a deal from lead to opportunity. - You want multi-touch nuance without modeling the entire sales cycle.

• When NOT to use it: - You do not have a clean, consistently applied opportunity stage — the 30% has nowhere reliable to land. - Post-opportunity touches (late-stage sales enablement, proposal content) are critical to your close and you need them credited — use full-path instead.

• Real-world scenario: A company with a sales-assisted motion discovers via W-shaped attribution that a specific case study is repeatedly present at the opportunity-creation moment. That asset never wins first or last touch, so simpler models ignored it — but it is quietly the thing that turns leads into real deals. They promote it across the funnel.

• How to interpret the data: The opportunity-creation peak is the differentiator here. Assets and channels that light up at that moment are your “deal-makers” — protect and replicate them.

Full-path

• What it is: 22.5% each to four milestones — first touch, lead conversion, opportunity creation, and the closing touch — plus 10% spread across the middle.

• The story it tells: “The entire journey from stranger to customer matters, and four moments matter most.” It is the most complete sales-and-marketing-aligned view.

• When to use it: - You have a full, well-instrumented funnel and a long, considered sales cycle. - Marketing and sales are jointly accountable for revenue and need a shared model. - You want to credit late-stage, sales-influencing content (proposals, ROI calculators, security docs).

• When NOT to use it: - Your funnel stages are messy or inconsistently used — full-path needs all four milestones to be real. - You are early-stage with thin data; the model will look precise while resting on too few journeys.

• Real-world scenario: An enterprise software vendor with a 9-month cycle uses full-path to settle a perennial sales-vs-marketing argument. The model shows marketing creating the opportunity and a sales-sent technical deep-dive sealing the close — both get ~22.5%. The shared, transparent split ends the credit war.

• How to interpret the data: Read full-path as a relay race. If any of the four batons (discover → convert → opportunity → close) is consistently dropped by a weak channel, that stage is your bottleneck.

Time decay

HubSpot attribution model selector showing U-shaped, W-shaped, Time decay, J-shaped, and Inverse J-shaped options

• What it is: Every touch gets credit, but touches closer in time to the conversion get progressively more.

• The story it tells: “Recent momentum matters most.” It assumes the interactions just before a decision were the most influential.

• When to use it: - You have a short or accelerating sales cycle where late touches genuinely drive the decision. - You run nurture-heavy or retargeting motions where recency correlates with intent. - You want multi-touch fairness but still want to reward the closing stretch.

• When NOT to use it: - Your awareness/discovery work is the hard part and deserves protection — time decay will systematically undervalue it. - You have a long cycle where early education does the heavy lifting; recency bias will mislead you.

• Real-world scenario: A subscription business with a 3-week consideration window uses time decay because, in their data, the last 7 days of retargeting ads and nurture emails truly tip the buyer over. Early blog touches still get some credit, but the model correctly emphasizes the closing push.

• How to interpret the data: Time decay rewards “recency.” Before acting, confirm recency actually equals influence in your business — in long cycles it often does not.
  

J-shaped and Inverse J-shaped

• What they are: Two mirror-image models. J-shaped gives a little credit to the first touch and the majority to the conversion touch. Inverse J-shaped does the opposite — most credit to the first touch, less to the conversion.

• The story they tell: These are “opinionated tilts.” J-shaped says closing is where the value is, but discovery deserves a nod. Inverse J-shaped says discovery is where the value is, but the close deserves a nod.

• When to use them: - J-shaped: conversion-focused teams who still want to acknowledge the opener — a middle ground between last-touch and U-shaped. - Inverse J-shaped: demand-creation teams who believe finding the right audience is 80% of the battle but still want to credit the close.

• When NOT to use them: - You cannot clearly articulate why you are tilting toward discovery or closing. If you cannot defend the bias, use a balanced model (linear or U-shaped) instead. - Your stakeholders distrust “opinionated” models — these can look like you are engineering the answer you want.

• Real-world scenario: A founder-led startup convinced that “distribution is everything” uses inverse J-shaped attribution to keep the team focused on top-of-funnel reach, while still tracking which closing motions work. It encodes their strategy directly into how they measure.

• How to interpret the data: Treat J and inverse-J as strategy statements. The model you pick reveals what your team believes drives growth. Make that belief explicit and revisit it as you get more data.

How to choose the right attribution model

Match the model to the decision in front of you, not to a “best practice” list:

• Judging demand creation / top of funnel? First interaction or inverse J-shaped.
• Judging closing power / bottom of funnel? Last interaction, J-shaped, or time decay.
• Want a fair, no-favorites baseline? Linear.
• Lead-gen is the core objective? U-shaped.
• Running a defined B2B funnel with an opportunity stage? W-shaped.
• Need full sales-and-marketing alignment across a long cycle? Full-path.

Three rules that prevent expensive mistakes:

1. Never rely on a single model. Compare at least one single-touch and one multi-touch view. Where they disagree is where the insight lives.
2. Let cycle length pick the family. Short cycles tolerate single-touch and time decay; long, considered cycles demand multi-touch.
3. Your model is a hypothesis, not a verdict. Revisit it quarterly as data accumulates and the motion matures.

How HubSpot Marketing Hub Enterprise services attribution reporting

HubSpot exposes all nine models above in its attribution reporting tools, but access is gated by tier — note the lock icons in the screenshots. In practice, single-touch and basic views appear at Professional, while the full multi-touch model library (full-path, W-shaped, U-shaped, time decay, J-shaped, and inverse J-shaped) is a Marketing Hub Enterprise capability. Enterprise is where attribution stops being a single number and becomes a real reporting system.

What Marketing Hub Enterprise gives you:

• Multi-touch revenue attribution. Tie closed revenue — not just contact creation — back across the full set of models, so finance-grade reporting becomes possible.
• The complete model library. Switch the same report between first-touch, linear, U-shaped, W-shaped, full-path, time decay, and the J-shaped variants to stress-test conclusions in seconds.
• Contact-create and deal/revenue attribution. Measure what generates leads and, separately, what generates revenue — two questions that need two reports.
• Interaction-type granularity. Credit specific interaction types — page views, form submissions, email engagement, ad clicks, meetings, sales activities, conversation events — so you can see which kind of touch drove credit, not just which channel.
• Custom attribution reports and dashboards. Build reports by channel, campaign, asset, content type, or URL, then pin them to executive dashboards for always-on visibility.
• Campaign-level rollups. HubSpot Campaigns aggregate assets so attribution can be read at the campaign level, not just the individual-asset level.

The practical workflow: instrument every channel with consistent UTMs and HubSpot Campaign labels, let HubSpot stitch the cross-channel journey onto the contact and deal records, then read the same pipeline through multiple models to find the channels that consistently earn credit regardless of which model you apply. Those cross-model-durable channels are your safest bets to scale.

TWH note: Attribution is only as honest as the data underneath it. Before trusting any model, confirm tracking code is live everywhere, UTMs are standardized, lifecycle and deal stages are applied consistently, and pre-tracking contacts are excluded from first-touch reads. A beautiful model on dirty data is a confident lie.

Frequently asked questions

 There is no universally best model. The best model is the one that matches your sales-cycle length and the decision you are making. Use first-touch to judge demand creation, last-touch to judge closing, and a multi-touch model (U-shaped, W-shaped, or full-path) for a balanced, revenue-aligned view. Most mature teams compare several models rather than trusting one.

What is the difference between single-touch and multi-touch attribution? Single-touch models (first and last interaction) give 100% of the credit to one touchpoint, making them simple but incomplete. Multi-touch models (linear, U-shaped, W-shaped, full-path, time decay, J-shaped) distribute credit across several touchpoints for a more accurate but more data-hungry picture of the buying journey.

Which attribution models does HubSpot support? HubSpot supports nine: first interaction, last interaction, linear, U-shaped (position-based), W-shaped, full-path, time decay, J-shaped, and inverse J-shaped. The full multi-touch set is available in Marketing Hub Enterprise.

Do I need HubSpot Enterprise for attribution reporting? Basic and single-touch views are available at lower tiers, but the complete multi-touch model library and multi-touch revenue attribution are Marketing Hub Enterprise features. Enterprise is required to switch one report across all nine models and to attribute closed revenue rather than only contact creation.

Why do different attribution models show different results? Because each model splits credit using a different rule. First-touch favors awareness channels, last-touch favors closing channels, and multi-touch models spread credit across the journey. The disagreement between models is not a bug — it is the most useful signal you have about where each channel adds value.

How often should I change my attribution model? Treat your model choice as a quarterly hypothesis. Revisit it as your data volume grows, your sales motion matures, or your primary objective shifts (for example, from demand creation to revenue efficiency). Avoid changing it mid-quarter, which makes period-over-period comparison unreliable.