Blog · Territory, capacity and quota planning
Find the highest-return channels in your own exports using contribution, full marketing costs, and equally mature acquisition cohorts. Follow the spreadsheet steps and reconcile the channel ranking to the company total.
The highest ROI marketing channels are the ones with the strongest contribution return on your own full marketing costs, measured across comparable cohorts. There is no reliable universal ranking that can replace that calculation. Build a channel table from CRM, revenue, delivery-cost, and marketing-spend exports, then check the totals before changing the budget.
Choose the acquisition cohort, observation window, customer type, and credit rule. For example, compare customers first acquired during January 2026 using each customer's first 90 days of observed revenue. Keep existing-customer promotions separate from new-customer acquisition.
Record the reporting cutoff and exclude customers without the full observation window from the mature ranking. Recent cohorts can appear separately as incomplete. Google's conversion-window documentation explains that the window determines which conversions can be recorded after an ad interaction. Different windows can therefore produce different channel totals even before costs enter the comparison.
Use one supplied attribution rule. The marketing attribution guide covers the credit choices; do not merge independently credited channel reports as though their revenue were additive. Covirage is not affiliated with Google.
Start with eligible customer or order references, acquisition dates, channel credits, net revenue, and associated delivery costs. Match spend records to the acquisition cohort rather than taking marketing invoices from whichever month revenue closed.
Allocate shared agency, creative, and marketing labor costs using a documented basis. Keep an unallocated row when a defensible assignment is unavailable. Unmatched revenue also needs its own row; silently distributing it changes the apparent winners.
The following illustrative US company has already reconciled credit and costs. All channels use the same cohort and observation window. Every monetary column is USD.
| Channel | Net revenue (USD) | Delivery costs (USD) | Media cost (USD) | Other marketing cost (USD) |
|---|---|---|---|---|
| Paid search | 180,000 | 108,000 | 20,000 | 10,000 |
| 120,000 | 72,000 | 2,000 | 13,000 | |
| Events | 240,000 | 144,000 | 35,000 | 25,000 |
| Total | 540,000 | 324,000 | 57,000 | 48,000 |
For email, the other marketing cost includes production and campaign labor. It is not a costless channel because its media bill is small.
Use the table columns as A through E, with the first channel on row 2. Add contribution in F, full marketing cost in G, after-marketing contribution in H, and ROI in I. Enter these formulas in F2, G2, H2, and I2 respectively:
=B2-C2
=D2+E2
=F2-G2
=IF(G2=0,"Not comparable",H2/G2)
Fill down to row 4 and format I as a percentage. For the underlying margin distinction, use marketing ROI versus ROAS.
| Channel | Contribution (USD) | Full marketing cost (USD) | After-marketing contribution (USD) | ROI | ROI rank |
|---|---|---|---|---|---|
| Paid search | 72,000 | 30,000 | 42,000 | 140.0% | 2 |
| 48,000 | 15,000 | 33,000 | 220.0% | 1 | |
| Events | 96,000 | 60,000 | 36,000 | 60.0% | 3 |
| Total | 216,000 | 105,000 | 111,000 | 105.7% |
Percentages are rounded to one decimal place. For numeric ROI rows, rank descending with:
=RANK.EQ(I2,$I$2:$I$4,0)
Email leads on ROI, while paid search contributes the most dollars after marketing. Events contribute more dollars than email while using much more marketing spending. Those are different decisions, and the table preserves both.
The net revenue total must match the eligible cohort's sales after returns and credits. The delivery-cost total must cover those same sales. Marketing costs must reconcile to the cohort cost schedule, including accruals and shared allocations.
The arithmetic check is 540,000 - 324,000 - 57,000 - 48,000 = 111,000, which equals the sum of after-marketing contribution. Total ROI divides total H by total G:
=SUM(H2:H4)/SUM(G2:G4)
It returns 105.7%, rounded. Do not average channel percentages. If a channel has no recorded marketing cost, investigate missing costs before reporting an unlimited return.
Marketing owns channel mapping and the proposed budget change. Finance owns cost scope and the reconciliation. Sales confirms whether additional leads can receive the same follow-up and whether customer mix is changing.
Choose a limited spending change with an agreed measurement window. Track additional contribution and additional cost separately from the historical averages. The guide to improving marketing ROI covers the operational changes once the baseline is comparable.
Attribution still does not prove causation. Google's measurement guidance distinguishes attributed conversions from incremental conversions measured against a control group. An experiment may be needed before calling a channel's observed advantage incremental impact.
Bring a sample of your own CRM cohort summary, revenue and delivery-cost export, and campaign cost schedule. Covirage maps the columns; its tools compute and check the channel returns, and the external AI model explains the result with every figure traced to its rows. Create a free Covirage account and start with your own files.
See sales insights for the shared view of channel credit and recorded sales.
There is no universal winner. Channel results depend on audience, customer mix, margins, acquisition costs, and how revenue credit is assigned. Rank your own comparable cohorts after delivery costs and full marketing costs, then examine the additional return from a controlled increase in spending.
Use both. ROI shows return relative to the spending required, while after-marketing contribution shows dollars available toward overhead and profit. A small channel can have the highest percentage but less total contribution, and its reachable audience may limit further expansion.
Choose an observation window that fits your sales cycle and apply it equally to each acquisition cohort. Review the distribution of days from first contact to purchase. Mark recent cohorts as immature rather than comparing their partial outcomes with fully observed older cohorts.
No. An attributed return describes observed sales under a credit rule. Additional spending can reach different customers at different costs, and some credited customers might have bought anyway. Use the ranking to propose a measured spending test with agreed cost and contribution outcomes.