Northbeam: marketing attribution and measurement

Northbeam AI answers account questions inside the dashboard, such as which campaigns declined week over week or how spend and revenue break down by platform. Teams can add promotion calendars, budget plans and operating guidance to Context Center so those documents are available in conversations on that dashboard. It can change the requesting user’s default attribution or accounting settings and file a confirmed support ticket, but it does not edit orders, spend or platform connections.

Northbeam MCP lets an external assistant such as ChatGPT or Claude query dashboard figures and orders through the operator’s existing Northbeam login. An operator can compare Meta ROAS between two weeks or request a platform-level spending breakdown; anyone with multiple dashboards must identify the account to query. The connection is read-only and follows the user’s existing access, sending bounded answers to specific requests rather than synchronizing a full account in the background. The receiving AI provider’s data terms apply.

Overview is the team’s starting dashboard for revenue, spend, customer acquisition cost and transactions. Its data combines website activity captured by the Northbeam pixel and tagged links with backend orders and connected advertising records; backend orders also capture purchases that did not fire a web pixel. Operators can save shared KPI views and add conversion-lag charts to see how revenue arrives after the marketing interaction.

Metrics Explorer compares how two marketing metrics move together over time, such as paid-social spend and branded-search revenue. Operators choose the metrics, dates and daily, weekly or monthly detail, then save or export the view. The relationship can identify a channel interaction worth investigating before a budget cut; the correlation itself does not establish that one metric caused the other.

Multi-Touch Attribution divides an order’s revenue across recorded marketing interactions, giving the team a common basis for comparing campaigns. Clicks-Only emphasizes acquisition: when an upper-funnel touchpoint exists, it excludes direct visits, organic search, branded paid search and email or SMS from the allocation. Last Non-Direct Touch instead credits the last non-direct visit and helps inspect closing channels; Linear distributes credit equally. The choice changes the question answered, so a low email value under Clicks-Only should not be read as evidence that email closes no orders.

Clicks + Deterministic Views (C+DV) adds platform-confirmed ad views to the click-based customer journey. An administrator enables the required sharing, then Northbeam matches the available view data to orders; this helps inspect exposure-driven campaigns that a click-only view misses. Initial processing can take 24–48 hours and the model updates three times daily. Platform matching and privacy thresholds determine which details are available.

Creative Analytics puts ad cards beside spend, acquisition cost and ROAS, with charts for comparing up to six ads. Product Analytics lets the operator select a SKU and follow it through platform, campaign and ad views to see which ads sell it; its CAC and ROAS indexes compare the account’s own assets, not an industry benchmark. Subscription Analytics separates first subscription orders from renewals using order tags. These views support different decisions: which creative to retest, which product an ad actually sells, and which campaigns bring new subscribers.

Media Mix Modeling+ supports decisions about how to distribute a larger budget across channels over weeks or months. It models historical channel spend and business results, including factors such as promotions and seasonality, so the team can compare alternative allocations. The analysis works at channel and business level and needs sufficient history; individual ad and campaign decisions use the more detailed attribution reports.

Incrementality tests whether sales change when a group is withheld from advertising, with geographic comparisons available when a randomized holdout is impractical. Northbeam reports incremental return on ad spend and calibrated ROAS, which can inform the interpretation of MTA and MMM. This is useful when the team needs to decide whether a channel is creating additional sales before committing more budget; test scope and eligibility are arranged for the account.

Apex sends Northbeam’s attributed order signals to supported advertising platforms so their algorithms can use the team’s selected measure of performance. Setup specifies new, returning or blended revenue, an attribution model and a 1-, 3- or 7-day window in Accrual mode. On Meta, enabling data sharing and selecting Custom Attribution at ad-set level are separate steps, with optimization subject to access. Changing the selected North Star signal restarts an approximately two-week calibration period; existing pixels and conversion APIs remain in use.