--- title: "Single Source of Truth for Marketing Data — Sealmetrics" description: "Why ad platforms, GA4, the CRM and finance report different numbers, and how to reconcile them to one total every team accepts. With measured cases." summary: "A single source of truth for marketing and finance is one measured total that every team accepts because it can be checked against something that really happened: the orders or bookings the business recorded. Marketing numbers disagree because each tool counts something different. Ad platforms report the conversions…" canonical_url: "https://sealmetrics.com/use-cases/single-source-of-truth/" lang: "en" author: "Sealmetrics" author_url: "https://sealmetrics.com/" content_type: "product" owner: "web" llm_priority: "useful" last_verified: "2026-09-15" source: https://sealmetrics.com/use-cases/single-source-of-truth/ publisher: Sealmetrics --- Use case · Marketing, agency, analytics, CFO # Marketing and finance are reading *different numbers.* The ad platform claims one figure, GA4 shows another, the order system records a third and finance books a fourth. Sealmetrics gives marketing, agencies, analytics and the CFO one measured total — checked against the orders that actually happened — so the meeting can move on to the decision. Checked against real orders · last click per session · no cookies · EU-hosted in Dublin Same week · four sources Four definitions 01 · **Ad platform** · **Conversions it credits itself** 02 · **GA4** · **Consented visits + estimates** 03 · **Order system** · **Orders that happened** 04 · **Finance** · **Revenue net of refunds** None of them is lying · only one is anchored in fact · none carries the channel Quick answer A single source of truth for marketing and finance is one measured total that every team accepts because it can be checked against something that really happened: the orders or bookings the business recorded. Marketing numbers disagree because each tool counts something different. Ad platforms report the conversions they credit to themselves, which can include modelled and view-through conversions. GA4 only records visitors who accept the banner and are not blocked, and fills part of the gap with Consent Mode estimates. The order system records what was sold, without a dependable channel. Finance books revenue after refunds, taxes and shipping. Sealmetrics sits in the middle: it counts sessions without cookies, attributes revenue to the last click of each session, and is reconciled against the order total first. On Incapto's Shopify store it recorded 96% of real orders and 97% of revenue over 48 days. It does not replace the finance ledger or match individual orders to people. Why the numbers never match ## Nobody is lying. *Everyone counts differently.* Each system answers a different question with its own definition of a conversion. The disagreement is structural, which is why another reconciliation spreadsheet never ends it. | Source | What it counts | Why it drifts | Who defends it | | --- | --- | --- | --- | | Ad platforms (Meta, Google Ads) | Conversions the platform credits to its own ads | Its own attribution window; modelled and view-through conversions, depending on settings; two platforms can claim the same sale | The agency or performance team | | GA4 with Consent Mode | Sessions from visitors who accepted the banner and were not blocked | Visitors who reject the banner are not recorded; part of the gap is estimated; traffic with no usable source lands in direct or unassigned | Analytics | | Order system (Shopify, CRM, PMS) | Orders and bookings that were placed | Accurate totals, but no dependable channel behind each order | eCommerce and sales | | Finance ledger | Recognised revenue | Net of refunds, taxes, shipping and cancellations, often booked after the order | The CFO | The measured cost of that gap: at [Palladium Hotel Group](https://sealmetrics.com/case-studies/palladium-hotel-group.md), 40% of inbound traffic had no source or medium and 35% of the bookings GA4 recorded had no channel. At [Incapto](https://sealmetrics.com/case-studies/incapto.md), 14 of every 100 GA4 visits had no origin anyone could act on, against 0.3% in Sealmetrics. Why consent-based tools lose that traffic is covered under [Consent Mode v2](https://sealmetrics.com/glossary/consent-mode-v2.md) and [data loss in analytics](https://sealmetrics.com/glossary/data-loss-in-analytics.md). What the disagreement costs ## The budget follows *whichever number wins.* When no figure is trusted, the loudest report sets the plan. The cost shows up in three places. 01 ### Meetings spent on the data Brand, departments and agencies arrive with different numbers and different incentives. The review ends as a debate about which figure is right, and the decision waits. That was Palladium Hotel Group's starting point. 02 ### Budget moved on the wrong base Consent loss is not even across channels. At Incapto, GA4 showed paid campaigns as 50% of traffic; measured without consent loss they were 62%. Twelve points of difference in the line that decides media allocation. 03 ### Agencies optimising for their own report Each platform credits itself, so every partner can defend its results. Without a neutral layer there is no fair way to compare partners, placements and audiences on the same terms. The reconciliation method ## Anchor to the till. *Then read the gap.* Nothing here depends on trusting one vendor over another. You anchor every tool to a number none of them produces — the orders the business actually took — and only then compare. Why decisions skew on incomplete data is the argument behind [complete data](https://sealmetrics.com/complete-data.md). 1. Take the real order total Use the order system's own figures for the period: Shopify, your CRM or the PMS. Leave out subscriptions, in-store and manual orders, which have no web visit behind them. 2. Measure in parallel Keep your current analytics and run Sealmetrics on the same site over the same days. Nothing needs to be switched off. 3. Reconcile before you compare Check each tool against the order total first. A tool that cannot match the till is not a reference for anything else. At Incapto, Sealmetrics recorded 96% of real orders and 97% of revenue. 4. Read the difference by channel Once the total holds, break the gap down by channel, because that is where budget decisions change. Treat the remaining gap as a quality signal, as Dreamplace Hotels does against its CRM total. 5. Agree the number every team reports from Marketing, agencies, analytics and finance take the reconciled figure as the reference. Ad platforms keep optimising in their own interfaces; the budget conversation happens on the neutral layer. One number, four readers ## The same total, *read four ways.* Every team keeps the tools it needs. What changes is the figure the decision is made on. ### CMO / Marketing Defend the budget with a number finance does not dispute. Channel revenue without consent loss, reconciled against orders before it reaches the board deck. [Analytics for CMOs](https://sealmetrics.com/for/cmo.md) ### Agency Show results on a figure the client does not have to take on trust. Agency staff can be members of each client's organization and switch between them. The platform still optimises bids; the result is judged on the neutral layer. [Analytics for agencies](https://sealmetrics.com/for/agencies.md) ### Analytics / BI Work from full-resolution data, not a sampled export. A BigQuery connector, the REST API and an MCP server read the same dataset the dashboard shows. [Integrations](https://sealmetrics.com/integrations.md) ### CFO / Finance Tie marketing revenue to the ledger. Aggregate reconciliation against the order total, EU-only processing in Dublin and a signed DPA. It does not replace revenue recognition. [Revenue attribution](https://sealmetrics.com/use-cases/revenue-attribution.md) Measured in practice > “Today every player is happy. The data is neutral, there's no black box, and everyone has accepted these numbers as the reference.” > > Toni Andújar · Digital & Direct Sales Director · Palladium Hotel Group Palladium Hotel Group moved the final discussion with its brand, departments and agencies onto one neutral measurement layer. Agencies still optimise their platforms; the decisions are checked in one place. **+165%** Display Cost-per-Search after rebalancing on the neutral model Palladium Hotel Group [Read the case](https://sealmetrics.com/case-studies/palladium-hotel-group.md) **15–20%** more sales attributed, with the CRM total as the reconciliation point Dreamplace Hotels [Read the case](https://sealmetrics.com/case-studies/dreamplace-hotels.md) **96%** of real Shopify orders recorded, reconciled before any comparison Incapto [Read the case](https://sealmetrics.com/case-studies/incapto.md) What it does not fix ## A shared number *has edges.* Stating the limits is what makes the number defensible in front of finance. Attribution is [last click](https://sealmetrics.com/glossary/last-click-attribution.md) within each session, by design. ### — It does not replace the finance ledger Refunds, cancellations and revenue recognition happen after the order. Reconcile to the order total; finance still closes the books. ### — It does not match orders to people Sealmetrics does not store order IDs or identify visitors, so reconciliation is on totals and channels, never row by row. ### — It counts what fires It does not deduplicate or validate orders. Fire the conversion once per confirmed order; on Shopify it comes from the orders/create webhook. ### — Ad platforms will still report more They credit themselves and can count modelled or view-through conversions. Keep them for bidding; decide the budget on the reconciled layer. ### — Last click per session, nothing more No multi-touch model, no view-through, no journeys across sessions. Earlier influence belongs in a marketing-mix model. ### — Orders without a web visit stay outside Phone, in-store and manual orders have no session to attribute. Questions teams ask ## Before the next *numbers meeting.* ### Why don't marketing and finance numbers match? Because each system counts something different. Ad platforms report the conversions they credit to their own ads, GA4 records only visitors who accepted cookies and were not blocked, the order system records orders without a dependable channel, and finance books revenue after refunds, taxes and shipping. The fix is not a better spreadsheet but one measured total anchored to the orders that actually happened. ### Why does GA4 show fewer conversions than Shopify? GA4 does not record visitors who reject the consent banner or block the script, and Consent Mode only estimates part of that gap. When Incapto ran GA4 and Sealmetrics side by side on its Shopify store for 48 days, GA4 did not record 29% of visits, while Sealmetrics recorded 96% of the store's real orders. ### Why do Meta Ads and Google Ads report more conversions than my CRM? Each platform credits conversions to its own ads within its own attribution window, and depending on settings it can include modelled and view-through conversions. When two platforms touch the same sale, both can claim it. Their reports are useful for bidding, not as the shared total. ### What is a single source of truth in marketing analytics? One measured total that marketing, agencies, analytics and finance all accept as the reference, because it reconciles with recorded orders and nobody in the room produced it to defend their own results. Palladium Hotel Group uses Sealmetrics this way across brand, departments and agencies. ### Can Sealmetrics match each order in my CRM to the channel that drove it? No. Sealmetrics does not store order IDs or identify visitors, so reconciliation happens on totals and by channel, not row by row. It is the same design that lets it measure without cookies. ### Do we have to stop using GA4 or the ad platforms' reports? No. Most teams keep GA4 and the ad platforms running for Google Ads and Meta optimisation. What changes is the number the budget decision is made on: the reconciled Sealmetrics total. ### How long does it take to know whether the numbers reconcile? Run both tools over a full commercial cycle. Incapto's reconciliation covered 48 days; most teams run a parallel test of at least 30 days before moving budget decisions. Reconciliation review ## Bring your order total. *We bring the other side.* Thirty minutes: we compare what your analytics and ad platforms report with the orders your business recorded, and show where the channel picture changes.