# Direct mail pilot plan

A one-page plan to fill in before you mail: the list, the response rule, the costs, how big the test needs to be, and the formulas to read the result.

# Direct mail pilot plan

Fill this in before anything goes in the mail.

## 1. The question

Does a handwritten card to ______________________ (the list) get more ______________________ (one action) than sending nothing, within ____ days of the mail date?

## 2. The list

- Where it comes from:
- Number of people (aim for 400 or more):
- Cleaned on (fix names, states and ZIP codes; remove duplicates):
- Split: add a column of random numbers, sort by it, and put the top half in the card group and the rest in the holdout. Never hand-pick.

## 3. The response rule

"A response is ______________________ by anyone in either group, from the mail date to day ____."

Count everyone who does it, whether or not they mention the card or use a code.

## 4. Costs and break-even

- Cards: ____ x $____ = $____
- Setup (design, list work, your time): $____
- Total mailing cost: $____
- Contribution per response (what it brings in, minus the cost of delivering it): $____
- Break-even responses: total cost / contribution, rounded up = ____
- Break-even response rate: break-even responses / cards mailed = ____%

## 5. Keep everything else equal

Both groups get the same emails, ads and calls during the window. If the card group also gets a phone call, write that down: you're testing the card and the call together.

## 6. Read the result on day ____

- Card group rate = responses in the card group / people in the card group
- Holdout rate = responses in the holdout / people in the holdout
- Lift = card group rate minus holdout rate, in percentage points
- Extra responses = card group responses minus (holdout rate x people in the card group)
- Cost per extra response = total mailing cost / extra responses

## 7. How big the test needs to be

Our calculation: people needed in each group to detect the difference 80% of the time, at 5% significance.

| Response without the card | Response you hope the card gets | People in each group |
|---|---|---|
| 2% | 4% | about 1,140 |
| 2% | 6% | about 380 |
| 3% | 9% | about 250 |
| 5% | 10% | about 440 |
| 10% | 15% | about 690 |

A smaller pilot still checks your costs, your list and your follow-up. It can't prove a small lift.

## 8. Decide now, before the results

- Mail the rest of the list if: ______________________ (for example, cost per extra response under $____)
- Change one thing and test again if: ______________________
- Stop if: ______________________

## 9. The one-sentence report

"From ________ to ________, ____ of ____ people who got the card (____%) and ____ of ____ in the holdout (____%) ______________________. That's ____ extra responses, at $____ each."

---

From Scribble Cards: https://scribblecards.com/blog/handwritten-direct-mail

Editorial review: 2026-11-10
