Strategy
Part of Offline marketing strategy: the choices that shape everything else
The offline marketing benchmark card that keeps figures checkable
A benchmark card for offline campaigns: denominators, delivery and response counts, costs and sources, plus how to judge vendor figures and small samples.
What to take away
- Fix the denominator first. Produced, delivered, reached, responded, qualified and mature units are not interchangeable.
- Publish a benchmark card with counts, rates, cost, maturity window and a named source.
- Treat outside figures as dated ranges with a sample and a geography, never as targets.
- Keep operational failures and uncertainty beside the headline rate.
- Agree before launch what change forces a rebaseline.
Set the denominator before you compare
A response rate is a fraction, and the bottom of that fraction decides what it means. Pieces produced, pieces delivered, households reached, calls answered, codes redeemed, qualified inquiries and mature customers are different bases.
Choose one base per campaign and write it into the export, the dashboard and the review note. Produced and delivered counts often diverge, so publish both.
Add returns, refusals and address errors before anyone quotes a rate. The complete guide to offline campaign measurement covers the routing and maturity windows that hold those bases steady across vendors.
Tag a printed QR code's landing URL with UTM parameters so scans are attributed to the drop. Google's Campaign URL Builder documents the parameters.
The benchmark card
Field / What to record
- Audience
- Eligible people or firms, plus exclusions
- Geography
- Postal areas, venues, stores, placements
- Delivery
- Produced, accepted, installed, returned, removed
- Response
- Call, scan, visit, inquiry, redemption, order
- Maturity
- Cancellation, return and fulfilment window
- Economics
- Full campaign and operating cost
- Source
- Dataset, owner, date, sample, method
Record the card once per campaign and store it beside the decision. A reviewer should reproduce the reasoning without asking what the numbers meant. Publish these six figures next to the headline rate:
- Delivery and undeliverable rate
- Qualified response per delivered unit
- Mature outcome per eligible audience
- Cost per qualified and mature outcome
- Complaint, waste and accessibility failures
- Comparison against a non-contact or holdout group
The definitions on the card outrank any external target. The offline marketing strategy checklist lists the definitions worth locking before the card is filled in.
Example: a door hanger test card
Illustrative numbers for a 20,000-piece door hanger drop in one metro, a two-week test with a 30-day maturity window.
- Produced 20,000. Accepted by the distributor 19,600. Delivered 18,400. Returns 400.
- Tracked scans 460. Qualified inquiries 210. Booked jobs 46.
- Total cost $8,000$6,800 print and distribution, $1,200 tracking and staff time.
- Per delivered piece2.5% scan, 1.14% qualified inquiry, 0.25% booked job.
- Cost per qualified inquiry $38.10. Cost per booked job $173.91.
- Sourceinternal distribution file and CRM export, dated.
That card is a worked example. It shows the shape a card takes and the arithmetic a reviewer can check.
Where credible benchmark data lives
For the United States, ask the USPS for delivery and addressing data, and take the current ANA Response Rate Report, which replaced the DMA version and states its own sample and year. For Canada, Canada Post publishes Neighbourhood Mail guidance and Statistics Canada supplies the geography files behind a denominator.
Free local denominators come from census tracts, business licence lists and chambers of commerce. How Toronto agencies plan direct mail campaigns with Canada Post data shows how one market builds those denominators from postal and delivery records.
Judge vendor-supplied figures
Printers, list brokers and distribution firms quote rates constantly. Ask for the counts behind the figure, the date, the geography, and whether it counts delivered or produced units.
Until you can rebuild the arithmetic, treat the number as a sales claim, not an input. Check who collected the data and in which year before it enters a decision.
A response rate without a source, sample and date is an anecdote.
The FTC's advertising guidance covers what a business may claim about results, which matters when a vendor's figure becomes your public case. Where spending is local, a cost-per-lead comparison beats an impressions quote; chamber of commerce vs local sponsorship in Halifax runs that comparison line by line.
Small samples and uncertainty
A 40-response result from 1,000 pieces looks precise on a slide and is not. Report raw counts and a confidence interval for the proportion, using the method in the NIST handbook, and note clustering by postal route.
Small tests also fail operationally. Pause a campaign when delivery records, response routing, inventory or staff coverage break, even if early response looks strong. Resume only after the cause and the affected period are written down.
Common questions
What counts as a good direct mail response rate?
There is no single answer. Rates move with list quality, offer, format, season, geography and the denominator you chose. Recalculate any figure you see from its raw counts before you use it.
Should I report cost per scan?
Cost per scan diagnoses tracking and creative. Use cost per qualified or mature outcome for the investment decision, because scans include accidental and repeat hits.
How do I handle a small sample?
Publish counts and an interval, not a point rate. State the window and avoid claims the design cannot support.
Does a vendor change force a rebaseline?
Yes, when definitions, geography, tracking or fulfilment rules change. Record the old card, the new card and the date of the switch.
