Open the Mailing Report You Actually Have
Start with one mailing that’s old enough for your normal buying cycle to show. Don’t mix it with the next drop. For that one group of addresses, write down:
- how many pieces were ordered, delivered, and returned;
- where the list came from and which area it covered;
- what printing and postage cost;
- which creative version each home received;
- when the piece should have arrived; and
- when you’ll stop counting new responses.
Then follow the same group through calls, forms, QR or short-URL visits, qualified opportunities, consultations, proposals, signed work, revenue, and gross profit.
A call is a response. It isn’t automatically a qualified lead or a sale.
Give one person ownership of the report and the missing source information. Otherwise, you’ll end up comparing a printer’s delivery total with a salesperson’s memory and calling it attribution.
Define Every Response Before You Count It
Use separate fields for inbound call to the mail number, unique-page visit, form submission, QR scan, direct reply, and self-reported mail exposure. Deduplicate the person. Then mark valid lead, project fit, consultation, proposal, and signed job.
Branded search can rise after a mailing, but it’s an assisted signal unless the individual record or a controlled test supports stronger attribution. Don’t add every branded visitor in the ZIP code to the mail total.
Work the Cohort From Delivery to Gross Profit
Illustrative mailing: 5,000 pieces cost $4,000 to produce and deliver. The tracking paths receive 42 unique responses. Twenty-two are valid inquiries, 11 fit the intended work, six book consultations, three receive proposals, and one signs at $45,000. At a 35% gross margin, illustrative gross profit before sales and overhead is $15,750.
That model doesn’t predict your response or close rate. Replace it with your list, actual cost, duplicates, project mix, sales outcomes, cancellations, and margin.
Use a Holdout When the Decision Is Large Enough
If the list and budget support it, randomly hold back a comparable group of addresses. Compare tracked responses, branded demand, and qualified outcomes between mailed and unmailed groups over the same window. Avoid choosing the best neighborhood for mail and the weakest one for the holdout.
Small samples can produce noisy results. If a holdout isn’t practical, say so and use the cleanest cohort tracking you can rather than pretending the test is experimental.
Set the Decision Before the Results Arrive
Make the decision rules before the results tempt you to explain them away:
- Stop if the list is inaccurate, the work is a poor fit, or cost per signed job can’t meet your margin target after a fair test.
- Change one major variable when the response or qualification data points to a clear problem.
- Scale only when the signed-job economics work and your team can handle more conversations.
Review the mailing weekly while leads are active, then once more after your normal sales window closes.
If you want someone to run the mailing and keep this measurement connected to the sales record, that’s part of Lead Care Team’s Direct Mail service.