You're about to spend $1,000 on Google Ads this month. Your agency is telling you to increase to $3,000. Angi wants $225 per lead.
Before any of that, run this $135 test. It takes 20 minutes to set up and will tell you more about what works in your market than a month of digital ad spend.
The Test: 100 Handwritten Postcards
Here's the setup:
Budget: $135 (100 postcards x $1.35 each, postage included) Time to set up: 20 minutes Time to results: 2-3 weeks
That's it.
Step 1: Pick Your 100 Addresses (5 minutes)
You have three options, ranked by expected response rate:
Option A (highest response): Past customers Pull 100 customers who haven't booked service in 12+ months. These people already trust you. They just need a nudge. Expected response rate: 4-6%
Option B (medium response): Neighbors of recent jobs Take the last 5 jobs you completed. Pull the 20 nearest addresses to each one. Expected response rate: 2-3%
Option C (new customer acquisition): Targeted cold list Pick one neighborhood. Filter for: homeowner-occupied, single-family, built 2006-2016 (systems 10-20 years old). Get addresses from county assessor data or a list provider. Expected response rate: 1.5-2%
If you're not sure, go with Option A. It's the most forgiving test because these people already know you.
Step 2: Write Your Message (10 minutes)
Keep it under 150 words. Here's a template that works:
Hey [Name],
Your [AC / furnace] needs a checkup at least once a year. Skipping it is how a $150 tune-up turns into a $2,000 emergency repair.
We're booking pre-season appointments now. $89 for a full tune-up, usually takes about an hour.
If you want to get on the schedule, call or text [Phone Number].
— [Your Name], [Company]
Customize the price, the service, the timing. But don't overthink it. The power of this test is in the handwritten format, not the copywriting.
Step 3: Send It (5 minutes)
Upload your 100 addresses to Mailbots.ai. Paste your message. Each postcard gets written with a real pen using real ink — it looks like you personally sat down and wrote 100 cards.
Total time from "I should try this" to "postcards are being written": 20 minutes.
Cards typically hit mailboxes within 5-7 business days, and you can track each one individually through the USPS delivery system.
Step 4: Wait and Track (2-3 weeks)
Responses typically come in two waves:
Wave 1 (days 1-5 after delivery): The immediate responders. These are people who opened the card, saw your offer, and called right away. Usually 60% of total responses.
Wave 2 (days 6-21): The "I'll get to it" crowd. They pinned the card to the fridge or set it on the counter. They call when they remember, or when their system acts up.
Track every response. Use a dedicated phone number if you can. At minimum, ask every caller "how did you hear about us?"
What to Expect
Based on real campaign data:
If you mailed to past customers (Option A):
- 100 cards x 4-6% response = 4-6 callbacks
- At $908 average job value = $3,632-$5,448 in revenue
- ROI: 27x-40x on a $135 investment
If you mailed to neighbors (Option B):
- 100 cards x 2-3% response = 2-3 callbacks
- At $908 average job value = $1,816-$2,724 in revenue
- ROI: 13x-20x
If you mailed cold (Option C):
- 100 cards x 1.5-2% response = 1-2 callbacks
- At $908 average job value = $908-$1,816 in revenue
- ROI: 7x-13x
Even the worst-case cold scenario — 1 response from 100 cards — gives you a 7x return. Compare that to Google Ads, where your first $135 buys you approximately 4.6 clicks and maybe 0.9 leads.
Why Most HVAC Marketing Tests Fail
Here's why you haven't tested direct mail before (or why you tried and gave up):
Problem 1: You tested printed postcards. Printed postcards get a 0.40% response rate. On 100 cards, that's 0.4 responses — basically zero. Handwritten gets 2.16%. On 100 cards, that's 2+ real callbacks.
Problem 2: You mailed to a bad list. Random addresses with no filtering = low response rate. Target homeowners with aging systems in your service radius.
Problem 3: You didn't track. You mailed 500 postcards and didn't set up a separate phone number. Calls came in, but you had no idea which ones were from the campaign. So you assumed it didn't work.
Problem 4: You quit after one try. You sent one batch and expected your phone to ring off the hook. Direct mail is a compound investment. The first drop introduces you. The second makes you familiar. The third makes you the obvious call.
The $135 Decision Framework
After your test, you'll have real data from your market. Here's how to read it:
If you got 2+ responses: Scale immediately. Send 500 cards next month. Then 1,000. The response rate holds at scale.
If you got 1 response: Profitable, but test a different list or message. Try Option A if you started with Option C.
If you got 0 responses: Unusual for handwritten at 100 cards, but it happens. Before declaring it "didn't work," check: Did the cards actually get delivered? (Mailbots shows this.) Was your phone number correct? Did you track properly? If everything checks out, try a different message or list segment.
Compare This to Your Current Spend
Let's put $135 in context:
- $135 on Google Ads = 4.6 clicks (at $29/click), maybe 1 lead
- $135 on Angi = 0.6 leads (at $225/lead)
- $135 on handwritten postcards = 100 delivered cards, 2+ leads, 1+ booked job
The handwritten postcard test costs less than a single Angi lead. And it generates exclusive leads that nobody else is calling.
You owe it to yourself — and to your ad budget — to run this test before writing another check to Google.
Mailbots.ai — 100 pen-and-ink postcards, $135, delivered with per-piece tracking. Twenty minutes to set up. The easiest marketing test you'll run all year.