The One Record Rule: a field list for auto repair CRM setup

This is the field list from our video on auto repair CRM automation. It follows the video's three parts:

1. One record. Pick where the customer record lives, usually your shop system, and make every other tool read from it.
2. Cheap jobs automated. Tags, tasks, routing and reminders can run on their own. Names, phone numbers and prices wait for a person.
3. Fresh data checked. Ask how often each tool syncs, because a reminder built on hour-old data can text about an appointment that was just cancelled.

The usual reason CRM automation misfires is two systems editing the same customer. The rule fixes that before anything is switched on.

Why the record drifts

Your shop management system is the record of work: customers, vehicles, repair orders, parts, labor and payments. A CRM is the record of the relationship: every contact, what was promised, when to reach out next. Shop systems have added messaging and marketing, and shops add outside tools on top. Tekmetric's integrations page lists more than 70 integrations, and each tool you connect is one more place a customer record can drift.

Drift looks small. A phone number with two digits swapped, a duplicate customer under a first initial, a car with no mileage since its last visit. Then an automation texts the wrong number, or marks a customer inactive who was in last month.

Part 1. One record

Write down, for your shop, which system is the source of truth for each kind of data. For most shops it's the shop system for almost everything.

Data	Source of truth	Everyone else
Customer name and contact	shop system	read only
Vehicle, VIN and mileage	shop system	read only
Repair orders, prices, payments	shop system	read only
Marketing consent and opt-outs	wherever STOP is recorded, synced to all	must honor it
Notes from calls and texts	the tool that captured them, linked to the customer	read
Tags and follow-up tasks	the CRM or automation tool	read

Then clean before you automate. Run a duplicate report and a missing-field report (no phone, no email, no mileage) and fix them by hand. Automation on a messy list makes the mess faster.

Part 2. Cheap jobs automated, the field list

This is the list to copy. Start with every field read only, then open the "automation may write" column one rule at a time.

Automation may write (low risk, easy to undo):

• Tags by what happened: service type, declined recommendation, complaint, fleet customer.
• Tasks with an owner and a date: "declined front brakes, call Tuesday," "text unanswered for a day."
• Routing: fleet inquiries to the owner, warranty questions to the manager, everything else to an advisor. Shopmonkey says its Concierge assistant creates follow-up tasks from calls, which is the same idea.
• Reminders: next service due by date or mileage, sent from approved templates.
• Flags: duplicate suspected, phone missing, mileage missing.

A person changes (high risk, hard to undo):

• Customer name, phone number and email.
• Prices, discounts and anything on a repair order.
• Merging two customer records.
• Marking a customer inactive or lost.
• Removing an opt-out.

A useful test: if the automation got it wrong, could someone undo it in one click without the customer noticing? If not, a person makes that change.

Part 3. Fresh data checked

This is the part vendors rarely explain. Systems exchange data in one of two ways: they check on a schedule, or they get told when something changes. One Tekmetric partner's help page says Tekmetric offers webhooks for appointment and repair order updates, added by an admin under Settings, then Integrations, and that without them the partner's account data syncs once per hour. That's one vendor's description of its own connection, not a rule for every tool, but it shows why the question matters.

Ask every tool these questions before a reminder rule goes live:

1. How often does it pull data from the shop system, and can the shop system push changes to it?
2. What happens to a queued reminder if the appointment is cancelled after the last sync?
3. Does it show when each record was last updated?
4. When a field conflicts between two systems, which one wins?

Then set a rule of thumb: no appointment reminder fires from data older than your sync interval. If a tool can't tell you its interval, assume the worst.

A setup order that works

1. Write the source-of-truth table.
2. Clean duplicates and missing fields by hand.
3. Turn on one automation rule, for example a task for every declined safety item.
4. Run it for two weeks and read every task it created.
5. Add the next rule only when the first one is boring.
6. Give every automation an owner who notices when it breaks.

How to tell it's working

After a month, your duplicate report should be short, every task should have an owner and a date, and nobody should be texting customers about appointments that no longer exist. Watch the reasons shops give for holding back on AI. In the 2026 Ratchet+Wrench survey, shops not using AI cited not knowing what's trustworthy (23%), cost (20%) and lack of time (19%). A small, read-only first step answers all three.

Where RevvOS fits

RevvOS sits alongside the system you already have. It reads from your shop, customer and marketing systems, shows where each item came from and when it was last updated, says when data is missing, and prepares next steps for a named person to approve. It doesn't replace your shop system and isn't a standalone CRM. See auto repair CRM automation and the integrations page.