DTC Attribution

DTC brands face an attribution challenge that is structurally more complex than most: customers typically encounter 4–8 touchpoints before purchasing, spanning paid and organic channels, and often across multiple devices. A customer might discover a brand through a Meta video ad, research via Google, return through an email, and purchase through a direct type-in — creating four potential attribution claims for one sale.

The three layers of DTC attribution are: platform-level attribution (what each ad platform reports), analytics-level attribution (what Google Analytics or a similar tool shows), and business-level attribution (what revenue actually came in). Each tells a different story, and the gaps between them reveal where measurement is weakest.

Exactius approaches DTC attribution as a triangulation problem, not a single-source-of-truth problem. The Growth Operating System uses incrementality testing, MER, and first-party data together to build a picture of channel contribution that no single attribution model can provide alone.

Attribution errors compound in DTC because every capital allocation decision — which channel to scale, which to cut — is made based on attribution data. If Meta is over-attributed and email is under-attributed, the business scales the wrong channel and systematically defunds the one building long-term LTV.

The DTC attribution problem has worsened since iOS 14, because signal loss disproportionately affects the paid social channels that DTC brands depend on most heavily. Meta's reported ROAS is now materially less reliable than it was pre-ATT, forcing a shift toward business-level measurement.

For DTC brands optimising toward LTV:CAC, attribution accuracy is a prerequisite for knowing which channels are acquiring high-value customers. Without channel-level LTV data, the business cannot distinguish a high-ROAS channel that brings repeat buyers from one that brings one-time purchasers.

The triangulation approach

Layer 1 — MER: Total revenue ÷ total ad spend. Attribution-agnostic. Tells you if the overall system is efficient.

Layer 2 — Incrementality: Geo holdouts or platform lift studies per channel. Tells you which channels are causing revenue vs. observing it.

Layer 3 — First-party analytics: Session-level data from GA4 or a custom analytics stack, using UTM parameters and server-side tracking. Tells you directional channel volume with less attribution distortion than platform reports.

No single layer is sufficient. MER tells you the outcome; incrementality tells you what caused it; first-party analytics tells you the journey. Used together, they produce a capital allocation signal that is far more reliable than any single attribution model.

Most DTC brands are over-invested in bottom-of-funnel channels and under-invested in upper-funnel because their attribution model rewards what it can see — last-click, direct, branded search — and penalises what it cannot — display, video, influencer. The capital follows the attribution, not the actual demand creation.

Exactius rebuilds DTC attribution infrastructure in the first 30 days of every engagement: CAPI implementation, UTM discipline, first-party analytics audit, and baseline incrementality testing by channel. Until this layer is in place, scaling decisions are directional guesses, not investments.

The Growth Operating System, developed by David Manela, treats attribution infrastructure as a prerequisite for the Capital Allocation Loop. Without reliable attribution signals feeding into the loop, the loop's decisions are only as good as the worst data source it receives.

Exactius embeds growth squads that own attribution infrastructure as a technical function, not a reporting function. The goal is not to produce better reports — it is to produce better capital allocation decisions.