Marketing analytics compares campaign spending with recorded conversions, customers and realized revenue. Work through a reconciled example and learn which definitions, files and checks make the comparison useful.
Marketing analytics connects marketing spend with recorded conversions, customers and realized revenue on consistent business definitions. It tells a team what happened after a campaign and where the evidence is incomplete. A campaign report becomes useful when its spending ties to finance and its customer outcomes can be traced to individual records.
The work follows a chain: activity, response, customer outcome, financial result. An impression is activity. A qualified inquiry is a response. A first purchase is a customer outcome. Revenue after credits is a financial result. Each answers a different question; none automatically proves that marketing caused the next step.
Keep the overview small enough to reconcile. For each campaign and period, record spending, unique conversions under a written definition, new customers, and net recorded revenue from those customers. Add contribution only when matching delivery costs are available. The detailed margin treatment belongs in marketing ROI versus ROAS.
Attribution determines which campaign receives credit. Google's attribution documentation describes attribution as assigning credit to interactions along a customer's path. That allocation is a reporting choice; preserve it with the export and use the same choice in comparisons. See marketing attribution models for the detailed allocation methods. Covirage is not affiliated with Google.
Use campaign cost exports, a prepared lead or conversion register, a customer register, and posted invoice and credit records. Keep an exception list for records that do not match. Do not silently exclude them to improve reported performance.
| Export | Fields to preserve | What it proves |
|---|---|---|
| Campaign spending | Campaign ID, channel, service period, cost, currency | The cost of activity in scope |
| Conversion register | Conversion ID, date, campaign ID, customer ID, status | Unique responses under the agreed definition |
| Customer register | Customer ID, first purchase date | Whether a customer is new |
| Finance register | Document ID, customer ID, posting date, revenue, credits | Revenue actually recorded in the selected window |
Choose the grain before combining files. Spending might have a row per campaign day, while invoices have a row per document line. Summarize each to a common campaign and cohort view before comparing totals. Otherwise a campaign cost can repeat against every invoice line.
For this overview, the measures are descriptive:
Cost per recorded conversion = Campaign spend / Unique recorded conversions
Campaign cost per new customer = Campaign spend / New customers
Net recorded revenue = Gross recorded revenue - Credits
Recorded revenue per spend dollar = Net recorded revenue / Campaign spend
A fictional US business-services company reviews September campaigns. Conversions are unique qualified inquiries dated September 1 through September 30, 2026. New customers are those inquiries that made their first purchase by October 23, 2026. Revenue includes their posted sales and credits through that same cutoff, under one exclusive campaign assignment per customer.
This is a campaign-cost view, not the company's fully loaded acquisition cost. For the wider cost pool, use customer acquisition cost.
| Channel | Spend (USD) | Inquiries | New customers | Gross revenue (USD) | Credits (USD) | Net revenue (USD) |
|---|---|---|---|---|---|---|
| Search | 6,000 | 120 | 12 | 30,000 | 2,000 | 28,000 |
| 2,000 | 80 | 8 | 18,000 | 1,000 | 17,000 | |
| Event | 4,000 | 40 | 6 | 22,000 | 2,000 | 20,000 |
| Total | 12,000 | 240 | 26 | 70,000 | 5,000 | 65,000 |
Search net revenue is 30,000 - 2,000 = $28,000. Apply the same subtraction to each row, then divide using the formulas above. Ratios below are rounded to two decimal places from unrounded calculations.
| Channel | Spend / inquiries (USD) | Spend / new customers (USD) | Net revenue / spend |
|---|---|---|---|
| Search | 50.00 | 500.00 | 4.67 |
| 25.00 | 250.00 | 8.50 | |
| Event | 100.00 | 666.67 | 5.00 |
| Total | 50.00 | 461.54 | 5.42 |
The total revenue ratio is 65,000 / 12,000 = 5.42, using the totals rather than an average of channel ratios. This is recorded revenue per spend dollar. It is not profit, return on investment, or incremental revenue.
The finance check is 70,000 - 5,000 = $65,000. Independently, 28,000 + 17,000 + 20,000 = $65,000. Both must match the selected customers' finance register, including all their in-scope credits.
The spending check is 6,000 + 2,000 + 4,000 = $12,000, tied to the approved campaign cost register. The customer check is 12 + 8 + 6 = 26 distinct customer IDs. A customer assigned to multiple channels would break that exclusive count even if each channel's export looked plausible.
The company should investigate email's customer mix before shifting funds. Existing brand familiarity, different audiences or different offer costs can explain the stronger observed ratio. Behavioral segmentation can help describe those differences without turning them into causal claims.
Marketing uses the comparison to identify campaigns worth investigating and incomplete cost records. Sales operations checks whether high inquiry counts become qualified customer relationships. Finance checks whether revenue survives credits and whether delivery costs leave enough contribution to justify further spending.
Evidence for increasing a budget is stronger when the team has a credible comparison group. NIST's guidance on completely randomized designs explains random assignment of factor levels to experimental units. For marketing, that supports designing an exposure test before observing outcomes rather than treating channel differences as proof of lift.
Prediction is a separate question: predictive marketing analytics tests supplied response estimates against later actual outcomes. Keep that validation separate from this historical reconciliation.
Bring a sample of your own campaign, customer and finance exports in a file or Google Sheet; Covirage maps the columns, and its tools compute and check the comparisons. The external AI model explains the results with every figure traced to its rows. Create a free Covirage account, then set up your business and bring your own export.
See sales insights for the supporting business context.
Start with campaign spending, dated conversion records, customer or account identifiers, and invoice and credit records. Keep campaign definitions and reporting windows with the exports. Aggregate platform totals alone cannot show whether the same customer or sale has been counted more than once.
Marketing owns campaign definitions and spending completeness. Sales operations owns lead and customer status. Finance owns recorded revenue and adjustments. Assign one person to reconcile those views and document exceptions, while each team remains responsible for the records it supplies.
Review operating activity as frequently as decisions require, but compare commercial outcomes only after the stated observation window has closed. A recent campaign with open opportunities should remain provisional. Keep the same reporting cutoff when comparing campaigns or successive periods.