The most useful oil change customer retention number is the second-visit rate: of the customers whose first visit was in a given month, what share came back within 120 days? You can pull it from a POS export by matching visits on phone or plate, and you should run it store by store, because the stores with the lowest rate are where new customers are leaking out.
Quick lube columnist Scott Hempy makes the case that the second visit matters more than the first, and that most operators overestimate their retention (NOLN). He suggests tracking last year's first-time guests and how many return within three or six months. This guide turns that into a formula, a POS export walkthrough and a store-by-store table.
Quick answer
- Second-visit rate = first-time customers from month X who returned within 120 days ÷ all first-time customers from month X.
- Match visits by phone or plate, across all your stores, not just the one they first visited.
- Leave out fleet and no-charge tickets, or they'll skew it.
- Run it by store and by month so you can see which store loses new customers and when it started.
- Fix the first visit (wait, explanation, next-visit reminder) at the stores at the bottom.
What is a second-visit rate?
It's the share of first-time customers who come back for a second paid visit within a set window. It's narrower than overall retention, which mixes in long-time regulars, and that's the point: regulars come back regardless. The second-visit rate shows whether a store is turning new customers into regulars.
Dealer service guides measure something similar over 12 months, using the share of first-visit customers who return within a year (Rework). For oil changes, a year is too slow to act on. A 120-day window is long enough for many customers to be due again and recent enough that you can still fix what went wrong.
How do you calculate it?
| Term | Definition |
|---|---|
| First-time customer | A customer (phone or plate) with no paid visit at any of your stores before this visit |
| Cohort | All first-time customers whose first visit fell in one calendar month, at one store |
| Returned | Has a second paid visit at any store within 120 days of the first |
| Second-visit rate | Returned ÷ cohort size |
Two choices change the number, so write them down and use them everywhere:
- Window. 90, 120 or 180 days. Shorter windows make every store look worse; that's fine as long as they're all measured the same.
- Any store or same store. Count a return at any of your stores as a return. Then, separately, look at how many came back to a different store. If many first-timers at Store A come back to Store B, that's a signal too.
You can only measure a cohort once its window has closed. A July cohort with a 120-day window is ready in late November.
How do you pull it from your POS export?
You need every paid ticket for at least the last 16 months, from every store: ticket date, store, customer phone, plate or VIN, ticket total, and payment type or account.
- Clean the keys. Phone numbers to digits only, plates uppercase with no spaces. Drop blank and placeholder phones (000-000-0000, the store's own number).
- Remove what isn't a retail customer: fleet and charge-account tickets, no-charge re-dos, warranty work, employee vehicles.
- Build one customer ID. Phone first, then plate if phone is missing. Where a plate shows up under two phones, treat them as one customer.
- Find each customer's first visit across all stores. Ignore anyone whose first visit is in the first 12 months of your data; you can't be sure they're new.
- Assign cohorts: first-visit month + first-visit store.
- Check for a second paid visit within 120 days at any store.
- Divide returned by cohort size, per store and month.
Step 4 is the one most people skip. Without 12 months of history before the cohort, a regular who skipped a visit looks like a new customer.
Worked example: four stores, July cohort
(Made-up numbers, not real store data.)
| Store | First-time customers (July) | Returned within 120 days | Second-visit rate | Returned to another store |
|---|---|---|---|---|
| A | 310 | 158 | 51% | 9 |
| B | 280 | 134 | 48% | 6 |
| C | 260 | 83 | 32% | 4 |
| D | 190 | 91 | 48% | 31 |
| Group | 1,040 | 466 | 45% | 50 |
How to read it:
- Store C loses about two in three new customers. If it matched the other stores' 48–51%, that's roughly 45 more regulars from this one month's cohort. Before blaming marketing, look at its first visits: wait times, how services are explained, whether anyone mentions the next visit.
- Store D looks fine at 48%, but a third of its returners (31 of 91) went to another store. Ask why: hours, wait, or a bad first experience at D that the customer didn't repeat.
- Run it monthly. One cohort is a snapshot. If Store C was at 47% in March and 32% in July, something changed in between, maybe a manager or a staffing gap.
This is also the check behind any mailer or promotion. A new customer who never comes back is worth one ticket, which is why our mailer match-back audit splits new customers out before pricing a campaign.
Why don't new oil change customers come back?
The POS can tell you that they didn't. It can't tell you why. The usual suspects, in the order we'd check them:
- The wait. Long waits on the first visit, especially at peak. Check wait times by store and hour.
- The upsell. A first visit that felt like a sales pitch. Look at add-on rate on first visits vs repeat visits by store.
- No next-visit plan. No sticker, no reminder, no mention of when to come back.
- Hours. Closing earlier than a competitor, or opening late.
- Data gaps. A store that doesn't collect phone numbers will look like it has no returners. Check the share of tickets with a valid phone or plate before reading anything else.
Hempy's own columns point at the first-visit experience: explain the services, introduce the team, thank the guest and invite them back (NOLN), and use a unique coupon code per marketing channel to see which channels work at each service center (NOLN).
What the POS can't show you
Wait time, walk-aways and how long the car sat before anyone touched it don't live in the POS. The cameras you already have can count cars entering and leaving the lot, time in bay and time waiting, by store and hour. Put next to the second-visit rate, that tells you whether Store C's problem is the wait or the conversation. See quick lube camera analytics for what that looks like.
How Trilix fits
We build this so nobody runs it by hand. POS history from every store is matched into one customer list, cohorts are built automatically, and the second-visit rate by store sits on the same owner dashboard as car count and ticket average, every morning. It sits beside the quick lube manager scorecard and the KPIs by store you already watch.
Want second-visit rate by store without the spreadsheet?
We'll look at your POS export and show you each store's July cohort.
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.
