A mailer match-back audit takes the list of households you mailed, matches it against your POS history by phone, plate or name and address, and counts who came in during a set window. The audit step most reports skip is the split: sort every matched customer into truly new, lapsed (brought back) and regular (would likely have come anyway), then work out your cost per truly new customer. That number, not the response rate, tells you whether the mailer paid.
Most of what ranks for this is written by mail vendors about their own wins. That's fine, and some of those campaigns did well. But a vendor report and your own POS history can tell different stories, and as an owner of a quick lube, tire or repair shop you're the one with the POS. This guide shows how to check it yourself, at every store, the same way.
Quick answer
- Match the mailed list to POS visits in a fixed window (30–60 days) by phone, plate, or name plus address.
- Split matches three ways: new (no visit ever), lapsed (no visit in 12+ months), regular (visited in the last 12 months).
- Price only the new and lapsed: campaign cost ÷ (new + lapsed) is your real cost per customer gained.
- Hold back 10% of the list next time. The gap between mailed and not-mailed is the lift.
- Our printable mailer match-back worksheet has every step.
Why does a mailer report count regulars as new?
Mostly because of how matching works. A typical match-back counts every repair order from a mailed household during the window. It doesn't check whether that household was already a customer who was due anyway.
The more honest vendor case studies say this out loud. One dealer service case study matched a 3,585-household mailer back to the store's system over 28 days and counted 403 repair orders, about 27% of the store's service business that month. The same page notes roughly a third of those were repeat visits, and there's no group that wasn't mailed to compare against (DirectMail.io). Some of those 403 would have come in without the postcard. The report can't tell you how many.
Saturation mail makes it worse. Postcards sent to every household around the shop reach your regulars too (Cactus Mailing), so a big share of "responders" are people already on your books.
None of this means mail doesn't work. It means the result needs splitting before you decide.
How do you do a mailer match-back against your POS history?
You need two files and a window.
- The mailed list from the vendor: name, address, and phone if they have it. Ask for the exact list, not a summary.
- Your POS export for every store: customer name, phone, address, plate or VIN, visit date, ticket total, store.
- The window: pick it before the mail drops. 30 days for oil changes is common; 60 if your mail goes out early in a season. Don't let the vendor change it after the fact.
Match in this order, and record which key matched:
| Match key | Strength | Note |
|---|---|---|
| Phone number | Strong | Clean formats first (digits only) |
| Plate or VIN | Strong | Ties to the car, not the person |
| Name + street address | Medium | Normalize "St"/"Street", unit numbers |
| Address only | Weak | Could be a new tenant; count separately |
Count unique households, not visits. One household coming in twice in the window is one customer.
How do you split new, lapsed and regular customers?
For each matched household, look at its POS history before the mail date.
| Group | Rule | What it means |
|---|---|---|
| New | No visit at any store, ever | Customer the mailer likely won |
| Lapsed | Last visit more than 12 months before the mail date | Customer the mailer may have brought back |
| Regular | Visited in the 12 months before the mail date | Probably would have come anyway |
Twelve months is a reasonable cutoff for oil changes; some operators use shorter. Victory Lane, a 36-store Michigan quick lube chain, targets customers seven months past their last due date with its lapsed mailers (NOLN). Tire Review's mail guide treats "inactive" as six months to two years with no purchase (Tire Review). Pick one, write it down, and use it at every store.
Check for "new" customers who visited a different store of yours: that's a regular who switched locations, not a new customer.
Worked example: one mailer, three stores
(Made-up numbers, not real store data.)
A quick lube group mails 15,000 households around three stores. The vendor charges $0.50 per piece, so $7,500. The vendor report says 600 responders, $12.50 per response.
Matched against the POS over 30 days, by unique household:
| Store | Matched households | New | Lapsed (12+ months) | Regular | Cost share | Cost per new + lapsed |
|---|---|---|---|---|---|---|
| North | 260 | 60 | 45 | 155 | $2,500 | $23.81 |
| South | 190 | 25 | 20 | 145 | $2,500 | $55.56 |
| East | 150 | 45 | 30 | 75 | $2,500 | $33.33 |
| Total | 600 | 130 | 95 | 375 | $7,500 | $33.33 |
What this shows:
- The vendor's 600 is accurate, but 375 of them (63%) visited in the year before. Most of them were likely due anyway.
- The real cost per customer gained is $33.33, not $12.50, and $57.69 if you only count truly new.
- South mailed mostly regulars. Next time, cut its route or exclude current customers from the list.
- North won the most new customers. That's where to spend next.
Whether $33 is a good price depends on what a new customer is worth to you over a year, which is why the second-visit rate matters: a new customer who never comes back is worth one ticket.
How do you measure real lift? Use a holdout
The split tells you who came in. It doesn't tell you who came in because of the mail. For that, hold back a random 10% of the list next time: same route, same households, no postcard. Then compare the visit rate of the mailed group with the not-mailed group over the same window.
If 4.0% of mailed households came in and 3.0% of held-back ones did, the mail added 1 point. On 15,000 pieces, that's about 150 extra households (example numbers). That's your real result, and it's usually smaller than the response rate. None of the top-ranking vendor guides we read describes a holdout; Tire Review suggests splitting a mailing to test wording or offers, which is a different test (Tire Review).
What should you ask your mail vendor?
- Can I have the exact mailed list, with addresses and any phones?
- What window do you use, and was it fixed before the drop?
- Do you count households or visits?
- Which match keys do you use, and what share matched on address only?
- Do you remove current customers from the list before mailing?
- Will you mail a holdout group so we can measure lift?
A vendor who answers all six is one worth keeping. This isn't about catching anyone out; it's about running the same check at every store so you can compare campaigns fairly.
How Trilix fits
We build the matching so nobody does it in a spreadsheet. The mailed list and your POS history from every store go in, and new, lapsed and regular counts by store come out, on the same owner dashboard as the rest of your numbers. The same customer history drives the quick lube manager scorecard and car count by store.
Want every campaign matched to your POS, store by store?
We'll look at your POS export and your last mailer and show you the new, lapsed and regular split.
Before you sign anything, you’ll speak with businesses we already work with. No pitch deck, no pressure. If it’s not a fit, we’ll tell you.
