The average B2B journey now spans 88 touchpoints. Last-click credits one.
Only 21% of B2B marketers say they trust their own attribution. We select and build the model — linear, U-shaped, W-shaped, or data-driven — that actually matches your sales cycle, and account for the dark-funnel activity tracking alone can't see.
What does attribution modeling actually involve?
Attribution modeling involves selecting and building the specific framework — first-touch, linear, time-decay, U-shaped, W-shaped, or data-driven — that assigns revenue credit across every touchpoint in a buyer's journey, matched to the actual sales cycle length and complexity rather than a default that happened to come pre-installed.
The confidence number that should worry every budget owner: only about 21% of B2B marketers say they're confident in their marketing attribution. Close to 80% of teams are deciding where next quarter's budget goes using numbers they don't fully trust — and the average B2B buyer journey has stretched to roughly 88 touchpoints across 10 stakeholders, which most default models were never built to capture.
- Cycle first
- Sales cycle length decides the model, not the other way round.
- Dark funnel counted
- Self-reported fields plus MMM recover what tracking misses.
- Two models, in parallel
- Disagreement between them is where a single model misleads.
- Ends in a budget shift
- Output translated into a recommendation, not just a report.
Seven models. Each answers a different question.
The right one depends on sales cycle length and how many touchpoints a deal actually takes.
| Model | Credit distribution | Best fit |
|---|---|---|
| First-Touch | 100% to the first interaction. | Auditing top-of-funnel and awareness spend only. |
| Last-Touch | 100% to the final interaction. | Short cycles under 90 days; misleading beyond 6 touchpoints. |
| Linear | Equal credit across every touchpoint. | Ensuring no channel goes invisible, 30–90 day cycles. |
| Time-Decay | More credit closer to conversion. | Sales-enablement-heavy journeys where late touches decide outcomes. |
| U-Shaped | ~40% first touch, ~40% last touch, 20% middle. | Balancing demand creation with the close — the most common B2B fit. |
| W-Shaped | ~30% each at first touch, lead creation, opportunity creation. | Long, milestone-based cycles, 90 days to a year. |
| Data-Driven | Algorithmic, based on statistical correlation with closed deals. | High conversion volume, strong CRM data, technical resources. |
No model can credit a touchpoint it can't see.
The dark funnel, and which layer of the measurement stack this service actually is.
A median 38% of B2B pipeline leaves no tracking signal
Private Slack and LinkedIn messages, podcast mentions, peer communities, and review-site research leave no digital tracking signal, yet drive a median 38% of B2B pipeline — rising to 51% for product-led growth motions. Self-reported attribution (a simple “how did you hear about us” field) and marketing mix modeling are the two practical ways to recover that visibility.
Audit, then infrastructure, then modeling
A GA4 audit diagnoses errors in an existing property. Marketing analytics builds the broader tracking, dashboard, and general attribution infrastructure. Attribution modeling is choosing and building the credit-assignment model itself. Most engagements run in exactly that order, since each layer depends on the one before it — building a model on broken tracking just compounds the problem.
Attribution work businesses bring us.
Matched to your actual cycle, with the dark funnel treated as a measurable layer rather than an excuse.
Model selection & build
Matching the credit-assignment model — linear, U-shaped, W-shaped, or data-driven — to your actual sales cycle.
Dark-funnel measurement
Self-reported attribution fields and signal correlation to recover pipeline that tracking alone misses.
Marketing mix modeling
Aggregate statistical modeling measuring channel impact over time, including offline and dark-funnel effects.
CRM milestone mapping
Connecting marketing touchpoints to CRM stages so W-shaped and full-path models have accurate data to run on.
Multi-model validation
Running two models in parallel to see where they disagree, since disagreement is where a single model would mislead.
Budget reallocation reporting
Translating model output into a specific budget shift recommendation, not just a report nobody acts on.
A clear path from default model to matched model.
Four stages, recalibrated as the sales cycle itself shifts.
Sales cycle & data audit
We map actual sales cycle length, stakeholder count, and CRM data quality before touching a model.
1–2 weeks · AuditModel selection & build
We select and configure the model matched to your cycle — U-shaped, W-shaped, or data-driven.
2–4 weeks · BuildDark-funnel layer
We add self-reported attribution fields and, where volume supports it, a lightweight MMM layer.
2–3 weeks · ExtendBudget review & iterate
We report model output against actual budget decisions quarterly and recalibrate as the sales cycle shifts.
Ongoing · IterateThe tools we use for attribution modeling.
B2B attribution platforms, CRM data, and warehouse tooling behind every model we build.
HubSpot
Looker StudioExplore more digital marketing services.
Attribution modeling works best paired with these.
Marketing Strategy Consulting
The full strategy service this sits under.
ExploreMarketing Analytics
The measurement infrastructure underneath the model.
ExploreGoogle Analytics Audit
Confirming the tracking is accurate first.
ExploreGoogle Analytics 4 Setup
Building the data layer the model runs on.
ExploreInbound Marketing
The funnel the model is measuring.
ExploreSalesforce CRM Integration
Milestone data for W-shaped and full-path models.
ExploreA/B Testing
Causal proof alongside correlational credit.
ExploreSaaS SEO
A long-cycle channel default models under-credit.
ExploreAttribution modeling questions
The things clients ask us most before starting an attribution project.
Marketing attribution modeling involves selecting and building the specific framework — first-touch, linear, time-decay, U-shaped, W-shaped, or data-driven — that assigns revenue credit across every touchpoint in a buyer's journey, matched to the actual sales cycle length and complexity rather than a default that happened to come pre-installed.
No. Only about 21% of B2B marketers say they're confident in their marketing attribution, meaning close to 80% of teams are making budget and channel decisions on data they don't fully trust. The average B2B buyer journey now spans roughly 88 touchpoints across 10 stakeholders over 272 days, which is why a single-touch default model leaves most of that journey invisible.
It depends on sales cycle length. Deals closing in 30 to 90 days generally fit linear or U-shaped models for full-funnel visibility, while deals running 90 days to a year fit W-shaped or full-path models that map to milestone-based CRM stages. Last-touch — still the most common default because it's built into most CRMs — actively misleads on anything with 6 or more touchpoints, which describes most mid-market and enterprise deals.
The dark funnel is the share of the B2B buyer journey that happens in untrackable places — private Slack and LinkedIn messages, podcasts, peer communities, and review-site research — that leaves no digital tracking signal. It accounts for a median 38% of B2B pipeline, rising to 51% for product-led growth motions, and Gartner puts total dark-funnel influence on B2B buying as high as 70%.
A U-shaped model credits first touch and last touch heavily, often 40% each, with the remaining 20% split across middle touchpoints. A W-shaped model adds a third anchor point, typically opportunity or lead creation, giving roughly 30% credit to each of three key milestones with 10% spread across the rest — built for longer B2B cycles with clearly defined funnel stages.
No, they're complementary, not competing. Multi-touch attribution tracks individual click-level paths and works for day-to-day channel decisions, while marketing mix modeling uses aggregate statistical analysis to measure channel impact over time, including dark-funnel and offline effects MTA can't observe. MMM adoption has tripled from 9% to 26% since 2023, driven by cookie deprecation, but most B2B teams still haven't adopted it.
Marketing analytics builds the broader measurement infrastructure — tracking, dashboards, and a general attribution setup. Attribution modeling is the deeper statistical discipline of choosing, weighting, and validating the specific credit-assignment model itself, work that typically follows once the underlying tracking has been confirmed accurate.
A simple "how did you hear about us" field at the lead-form or checkout stage captures dark-funnel influence tracking can't see — podcast mentions, peer recommendations, community discussion — and combining that self-reported data with tracked multi-touch data gives a more complete picture than either source alone.