Is multi-touch attribution completely dead?

Not entirely, but its reliability has degraded significantly for cross-device and cross-platform journeys specifically. Multi-touch attribution still has value for tracking within a single owned property where first-party data is available; it's the parts depending on third-party cookies and cross-site identity resolution that have broken down.

What's the difference between attribution and marketing mix modelling?

Attribution tries to assign credit to individual touchpoints in a specific customer's journey, which requires granular tracking that's become unreliable. Marketing mix modelling instead analyzes aggregate spend and outcome data over time to estimate each channel's contribution, without needing to track individual users — which makes it more resilient to the tracking degradation that's broken multi-touch attribution.

How long does it take to rebuild trustworthy reporting?

Building the first-party data infrastructure itself can happen within weeks, but marketing mix modelling typically needs several months of consistent data to produce reliable estimates. Most teams see a genuine, defensible reporting picture emerge within one to two quarters of starting the rebuild.

Do we need an enterprise budget to do this properly?

No — mid-market teams can run credible marketing mix modelling without enterprise tooling, using the spend and outcome data they already have. The barrier is usually methodology and data discipline, not budget size.

Resources/Use Cases

Analytics & Growth7 min read·Not a case study — a common situation, and the shape of how we'd approach it

Rebuilding Attribution Now That Cookies Don't Work

Attribution reporting used to feel reliable, and now it doesn't — channel numbers don't add up against actual revenue, the same conversion gets credited differently depending on which dashboard you check, and leadership has quietly stopped trusting the marketing reports even though nobody's officially said so. This didn't happen from one dramatic change; it happened gradually as third-party cookie loss, browser privacy defaults, and walled-garden platforms each cut off a piece of the tracking picture multi-touch attribution depended on. The old model degrades slowly and invisibly rather than breaking all at once, which is why most teams notice the symptom — reports nobody trusts — long before they diagnose the actual cause. Fixing this means rebuilding around what can still be reliably measured: first-party data and modeled attribution, not trying to patch a tracking approach the underlying infrastructure no longer supports.

Does This Sound Familiar?

Different platforms (Google Ads, Meta, your CRM) each claim credit for more conversions than your total conversion count actually allows.

Attribution reports show declining performance across most channels simultaneously, which usually signals a measurement problem rather than an actual performance problem.

Leadership has started making budget decisions based on gut feel or last year's numbers rather than current attribution data.

Nobody on the team can confidently explain how a specific reported number was actually calculated.

How This Actually Works

5-stage approach

1
Audit what's actually still measurable
Map which tracking mechanisms still function reliably in the current browser and platform environment versus which ones have quietly degraded, so the rebuild is grounded in what's real rather than patching assumptions that no longer hold.
2
Build first-party data infrastructure
Establish consented, owned data collection — email, CRM, and on-site behavioral data captured directly — as the foundation, since this is the data layer that doesn't depend on third-party cookies or cross-site tracking.
3
Introduce modeled attribution where direct tracking has gaps
Layer in marketing mix modelling or media mix modelling for the channels where direct multi-touch tracking is no longer reliable, using aggregate and incrementality-based methods instead of pretending granular per-touch tracking still works.
4
Reconcile against a single source of truth
Set actual revenue or pipeline as the reconciliation anchor, and build reporting that explains variance against that anchor rather than presenting each platform's self-reported numbers as independently correct.
5
Rebuild trust with transparent methodology
Document exactly how each reported number is calculated and what its known limitations are, so leadership can trust the report because they understand it, not because it looks precise.

1
Audit what's actually still measurable
Map which tracking mechanisms still function reliably in the current browser and platform environment versus which ones have quietly degraded, so the rebuild is grounded in what's real rather than patching assumptions that no longer hold.
2
Build first-party data infrastructure
Establish consented, owned data collection — email, CRM, and on-site behavioral data captured directly — as the foundation, since this is the data layer that doesn't depend on third-party cookies or cross-site tracking.
3
Introduce modeled attribution where direct tracking has gaps
Layer in marketing mix modelling or media mix modelling for the channels where direct multi-touch tracking is no longer reliable, using aggregate and incrementality-based methods instead of pretending granular per-touch tracking still works.
4
Reconcile against a single source of truth
Set actual revenue or pipeline as the reconciliation anchor, and build reporting that explains variance against that anchor rather than presenting each platform's self-reported numbers as independently correct.
5
Rebuild trust with transparent methodology
Document exactly how each reported number is calculated and what its known limitations are, so leadership can trust the report because they understand it, not because it looks precise.

What Changes

Before

After

Attribution built entirely on third-party, multi-touch tracking assumed to be complete.

Attribution built on owned first-party data plus modeled estimates for known measurement gaps.

Each platform's self-reported conversion numbers taken at face value.

All platform numbers reconciled against actual revenue as the single source of truth.

Reporting presented as precise, without explaining its actual limitations.

Reporting presented with transparent methodology, so leadership trusts what it does and doesn't claim.

Budget decisions made on outdated attribution or gut feel because current numbers aren't trusted.

Budget decisions made on a rebuilt model leadership actually understands and believes.

Part Of

Analytics & Growth

Know what's working. Prove it. Do more of it.

See the full service →

Is this what you're dealing with right now?

Tell us the specifics and we'll tell you honestly whether this is the fix.

Work With Us See other use cases