Define the response event
A scan, a phone call, a form submission and a sale should not be combined into one unexplained response figure. Specify the event before sending, and distinguish people from events: one person can scan several times. For account-based campaigns, report both responding contacts and responding accounts when the data allows. The account measure helps prevent a buying group with several engaged people from looking like several independent sales opportunities.
Make the denominator visible
Responses divided by pieces sent answer a different question from responses divided by confirmed deliveries. Report the denominator alongside the rate, and do not assume that a piece was delivered merely because it was not returned. If delivery cannot be confirmed, say so. Keep returned mail and address-quality problems visible as operational measures. Excluding them silently can make weak campaign preparation appear more effective than it was.
Compare like with like
A warm customer list has a different starting relationship from a cold prospect list. Consumer promotions, fundraising appeals and enterprise invitations also ask recipients to take different actions. Yotru’s discussion of published B2B direct mail statistics is useful background for reading campaign examples. It is vendor-authored interpretation, not a universal forecast. Before borrowing any number, inspect the audience, offer, sample, response definition and collection period behind it.
Show counts as well as percentages
In an illustrative test, six responding accounts out of 300 mailed accounts produce a 2% account response rate. One additional response changes that estimate, and a handful of unusually strong accounts can affect commercial results much more. Display six out of 300 alongside 2%, with the reporting cutoff. Small samples are useful for finding process failures and developing hypotheses; they rarely justify precise claims about what a much larger campaign will produce.
Follow the response into the sales process
Build a funnel that separates response, qualified conversation, meeting held, accepted opportunity and won customer. If many recipients scan but few book, examine the landing page and offer. If meetings happen but opportunities do not, inspect fit and qualification. These are possible explanations to investigate, not diagnoses established by the numbers alone. Review delivery and tracking quality before attributing a weak rate to the message.
Establish a working baseline
Preserve campaign-level records rather than pooling every mailing immediately. Compare similar segments across successive waves and record changes to creative, list source and follow-up. Use the baseline to set a decision threshold tied to economics, then revisit it as outcomes mature. A response target should help decide whether to continue, revise or stop; it should not become a headline detached from the revenue and costs it was meant to inform.
Sources are linked in the text. Examples are illustrative. About this publication.
