Northbeam attribution

Attribution and incrementality answer different questions. Attribution: 'which ad got credit for this conversion?' Incrementality: 'would this conversion have happened anyway if the ad hadn't run?' The two answers can diverge dramatically, especially on retargeting and brand campaigns where the user was already going to convert and the ad just got in front of them at the right moment.

Most attribution tools skip incrementality testing because it's operationally hard. A real geo-holdout experiment requires you to deliberately pause spend in some markets for 4-6 weeks, which marketing teams resist because it 'wastes' near-term revenue. The methodology requires careful experiment design (matched-market selection, power analysis, lift calculation) that's outside most attribution-tool teams' expertise. The output is uncomfortable - incrementality tests routinely reveal that retargeting is generating less incremental revenue than the attribution dashboard claims, which forces hard budget-reallocation conversations.

Northbeam built the incrementality layer because attribution alone systematically over-credits the bottom of the funnel and under-credits the top. Brands that have been over-spending on retargeting because the last-click attribution looked great are exactly the case incrementality catches. The first incrementality experiment on a Northbeam implementation typically pays for the annual subscription many times over by revealing where attribution and reality disagree.

Hyros defaults to a custom multi-touch model with first-click weighting - the founders' belief is that top-of-funnel acquisition gets systematically under-credited in last-click models, so the default explicitly over-credits first touchpoint. The model is opinionated; the math reflects the founders' point of view.

Northbeam takes a different approach: rather than picking an opinionated default, it lets you run multiple attribution models side-by-side and see how credit shifts under each. The default is a time-decay model (recent touchpoints weighted higher) but the platform encourages you to compare against linear, first-click, last-click, and U-shape simultaneously.

This is more analytically honest but more operationally demanding. You have to interpret the spread between models and decide which makes sense for your business - which requires analytics literacy that not every brand has. Hyros's opinionated default is friendlier for operators who want one answer; Northbeam's model-comparison approach is better for analytics teams that want to see the methodological sensitivity.

Where they converge: both tools acknowledge that incrementality testing is the only way to validate whether any attribution model is correct. Hyros doesn't build incrementality natively; Northbeam does. For brands at $5M+ where attribution-vs-reality gaps move real budget, the incrementality layer is the deciding factor.

Northbeam's CAPI pushback uses the same identity-resolution layer that drives attribution, which produces cleaner events than what platform-native server-side integrations send on their own. The Meta Conversions API gets deduplicated conversion events with full click-ID matching; Google Offline Conversion Import gets gclid-matched events; TikTok Events API gets ttclid-matched events; equivalent for Snap and Pinterest.

The operational benefit is consistency. When Meta's ads manager, Google's ads manager, and Northbeam's attribution dashboard agree on what happened, the marketing team trusts the numbers and acts on them. When they disagree, the team spends weeks in reconciliation conversations instead of optimization conversations. Northbeam's CAPI pushback minimizes the disagreement window by making sure the platform-native data and the Northbeam data come from the same underlying events.

Comparison to Hyros: Hyros's CAPI pushback is comparable in mechanism and quality; the difference is what's around it. Hyros pairs CAPI with the info-product-specific funnel-stage tracking; Northbeam pairs CAPI with cohort LTV analytics and the incrementality layer. Both are good at CAPI; the surrounding capabilities differ by target customer.

Pays back for: brands at $5M+ where budget allocation across channels matters as much as per-creative optimization, where finance and leadership want defensible methodology for ad-spend justification, and where there's an analytics function capable of interpreting incrementality results and acting on them. For these buyers, Northbeam attribution is the highest-fidelity option in the category.

Doesn't pay back for: sub-$5M brands where the absolute dollar impact of methodological precision is smaller than the subscription cost, info-product / coaching businesses where the customer journey doesn't fit Northbeam's DTC-shaped attribution model, and teams without dedicated analytics resourcing where the depth doesn't get extracted. For these buyers, simpler tools (Triple Whale, Hyros depending on workflow) deliver more value per dollar.

The deciding question: how much would better budget allocation across channels move your business? If the answer is 'meaningfully more than the Northbeam subscription cost', you're a fit. If the answer is 'we don't have the scale for it to matter', you're not.

Internal: northbeam, northbeam-analytics, northbeam-pricing.