Measure postcard campaign returns from mailing invoices, response codes and matched orders. Work from net sales to customer contribution, include print and postage, and check the result before deciding whether to mail again.
Postcard marketing ROI compares contribution from matched customer orders with the full cost of the mailing. Start with response codes and sales records, deduct refunds and delivery costs, then subtract print, postage and campaign preparation. A positive attributed return means the matched orders covered those costs; proving the mailing caused the orders requires a separate comparison.
Give the campaign a mailing date, a response deadline and an order observation deadline. Decide whether an eligible response is a submitted code, a completed inquiry or a purchase. Record each separately if the files contain all of them. A page visit is not automatically an order.
Use a unique campaign code and keep a customer or order identifier in the response export. Deduplicate repeated inquiries from the same person before reporting unique responders. Match orders using the identifiers already supplied by your business, and place unresolved matches in an exception list rather than assigning them by guesswork.
The mailing method determines what evidence exists. USPS describes Every Door Direct Mail as targeting routes through ZIP codes and demographics without requiring an address list. For a route mailing, your initial denominator can therefore be pieces mailed by route, even when you cannot identify every recipient individually. A route's total sales are not automatically postcard responses.
This page handles the mailing reconciliation. The distinction between revenue return and contribution return belongs to marketing ROI versus ROAS; broader direct campaign comparisons belong to direct marketing ROI.
Use the mail vendor invoice, postage statement, creative and preparation charges, response log, order export and refund report. The order file needs gross sales, discounts if not already deducted, refunds, and variable product or service delivery costs. Keep sales tax collected for authorities outside campaign revenue.
For this analysis, contribution means net sales less the variable costs of fulfilling the matched orders, before mailing spend. This is a management measure with an explicit cost boundary. The IRS explanation of net receipts subtracts returns and allowances from gross receipts; its gross profit calculation then subtracts cost of goods sold. Use that distinction to avoid treating a returned sale as an unchanged campaign benefit.
An illustrative US home-goods retailer mails in July 2026 and observes eligible purchases through September 30, 2026. Its invoiced costs are:
| Campaign cost | Amount (USD) |
|---|---|
| Printing | 2,000 |
| Postage | 4,000 |
| Address preparation | 500 |
| Creative work | 500 |
| Total mailing cost | 7,000 |
Both mailing groups use the same production specifications. The retailer allocates these costs by pieces mailed, giving Group A 6,000 / 10,000 of the total and Group B 4,000 / 10,000. If one group required a different format or postage service, assign its actual charges instead.
Put these inputs in columns A through H, with the first group on row 2. Gross sales are after any order discounts but before refunds. Variable fulfillment costs already reflect inventory recovered from returned products.
| Group | Mailed | Responders | Buyers | Gross sales (USD) | Refunds (USD) | Variable costs (USD) | Mailing cost (USD) |
|---|---|---|---|---|---|---|---|
| A | 6,000 | 180 | 72 | 22,500 | 900 | 12,960 | 4,200 |
| B | 4,000 | 80 | 32 | 10,200 | 600 | 5,760 | 2,800 |
| Total | 10,000 | 260 | 104 | 32,700 | 1,500 | 18,720 | 7,000 |
Add calculated columns I through L. These formulas work in Excel:
I2 = E2-F2
J2 = I2-G2
K2 = J2-H2
L2 = IF(H2>0,K2/H2,"")
I is net sales, J is order contribution, K is contribution after mailing, and L is attributed ROI. Format L as a percentage and fill down. A zero mailing cost leaves ROI undefined rather than creating a zero return.
If the orders are on a separate sheet with campaign codes in column A and net sales in column B, aggregate the matching code with:
=SUMIFS(Orders!B:B,Orders!A:A,A2)
Microsoft documents SUMIFS as adding values that meet specified criteria. Filter or flag the agreed observation window before totaling. Covirage is not affiliated with Microsoft or USPS.
The calculated results are:
| Group | Net sales (USD) | Order contribution (USD) | After mailing (USD) | Attributed ROI | Response rate |
|---|---|---|---|---|---|
| A | 21,600 | 8,640 | 4,440 | 105.7% | 3.0% |
| B | 9,600 | 3,840 | 1,040 | 37.1% | 2.0% |
| Total | 31,200 | 12,480 | 5,480 | 78.3% | 2.6% |
Response rate is =C2/B2. Total ROI is =SUM(K2:K3)/SUM(H2:H3), not the average of group percentages. Dollars are shown to whole dollars and these percentages to one decimal place; calculations use unrounded values.
The reconciliation is 32,700 - 1,500 - 18,720 - 7,000 = 5,480. Adding back the mailing cost gives 5,480 + 7,000 = 12,480 of order contribution. Tie the mailing cost to invoices, sales and refunds to the order records, and every matched order to one campaign assignment.
Contribution per matched buyer is 12,480 / 104 = $120. At that contribution, break-even requires 7,000 / 120 = 58.33 buyers, rounded up to 59 whole buyers. This is a planning threshold conditional on unchanged order economics, not a forecast of how many recipients will buy.
Use customer acquisition cost when you specifically need acquisition spending per new customer. Existing customers responding to a postcard should not silently enter that denominator. Once the records reconcile, improve marketing ROI covers the neighboring question of which changes to test.
Bring a sample of your own mailing cost, response and order exports to Covirage; it maps the columns, and its tools compute and check matched contribution against campaign spend. The external AI model explains the result with every figure traced to its rows. Create a free Covirage account, which takes about a minute, then set up your business and bring your files.
See customer profitability for the supporting business analysis.
No. A response can be an inquiry without a purchase, and a purchase can have little contribution after discounts, refunds and fulfillment costs. Use responses to diagnose the funnel, then compare matched contribution with the complete mailing cost.
State the denominator explicitly. Mailed pieces provide a count you can reconcile with the mailing record. A delivered-piece rate requires reliable delivery evidence for the same mailing. Do not switch denominators between campaigns or treat estimated delivery as a verified count.
Include repeat purchases only within the observation window agreed before the comparison. Match them to the same customer cohort and apply the same refund and delivery cost rules. Show a longer-window update separately so mature campaigns do not appear better merely because they had more time.