Automotive CRM Onboarding Checklist

The order matters. Get opt outs and lead sources in before anything else, and the rest of the setup stops fighting you.

There are two documents you need when you change CRM, and stores usually only make one. The implementation plan sets the phases, the owners and the review dates. This checklist is the other one: the actual settings, in the order that avoids redoing work.

The order is not arbitrary. Three items have to come before anything else, because everything downstream depends on them. Opt outs and your blacklist come first, because a customer who unsubscribed at the old system and then gets a message from the new one is a compliance problem rather than an inconvenience. Lead sources come second, because their naming determines whether your attribution reporting will ever mean anything and splitting them retroactively after a year of history is genuinely painful. Permissions come third, because retrofitting roles onto a live system with sixty users is far harder than designing them once.

Everything else can be adjusted later at low cost. Templates can be rough at launch. Drip content can be improved in week six. Calendar preferences can be argued about after go live. Do not let a debate over text message wording delay a launch date, and do not let a decision about lead source naming be made casually.

Print this, assign a name to each block, and put a date next to it. A checklist with no owner is a wish list.

Start with the structure, because every later step attaches to it.

1. Company settings. Store name exactly as customers know it, address, hours, time zone, and the phone number you want customers to see. Time zone in particular: a store that leaves it wrong spends the next month wondering why scheduled messages go out at odd hours.
2. User list. Every person who will touch the system, with a real email address rather than a shared one. Shared logins destroy activity reporting and make the login log useless the first time you need it.
3. Roles and permissions. Decide what a salesperson sees versus a manager, whether gross is visible, who can export, and who can change settings. Write the model down before you build it.
4. The administrator. One named person owns configuration going forward. Not a committee, because committees produce setups nobody can explain six months later.
5. Departing users. Agree now how access is removed when somebody leaves. The login log exists for the week a salesperson goes to a competitor and someone asks what they could reach.

Our CRM security checklist goes deeper on access control and what to review quarterly.

This block is boring, it is usually rushed, and it is the reason a lot of new CRM launches produce a bad first week. Give it real attention.

On the phone side, provision your numbers and decide what customers see on caller ID, then set up call forwarding rules so nothing rings into a void. If you are running inbound volume through AutoMail, build the IVR now: the greeting, the menu, the routing to departments, and what happens after hours. Record the greeting with a real voice rather than leaving a synthetic one in place, because the first impression of your store is frequently this recording. Set up voicemail drop messages so agents can leave a prepared message in three seconds instead of forty.

On the email side, complete email domain authentication before you send anything to anyone. Sending campaign email from an unauthenticated domain is how a store lands in spam folders for months and then blames the software. Then run your list through the email validator before the first campaign, and run your phone list through the phone validator before your team spends an afternoon dialing disconnects. Both tools exist precisely because migrated dealership data is usually older and dirtier than the dealer believes.

Then the item that goes first in practice even though it sits in this block: import your opt outs and your blacklist. Every unsubscribe, every do not call, every customer who asked your old system to stop. Bring the blacklist across before you bring the contacts across, so there is no window in which a message could go to someone who opted out. See texting compliance for the obligations behind this.

This is the block that determines whether your reporting is worth reading in twelve months.

1. List every source that produces a lead today. Website forms, third party lead providers, marketplace listings, chat, inbound phone, walk in, referral, campaign landing pages. Every one of them, including the ones nobody has thought about since 2021.
2. Configure lead feeds and custom source feeds for each, and be more granular than feels necessary. Merging two sources later is easy. Splitting one retroactively is not.
3. Name them in a convention you can live with. Agree the convention with whoever reads the reports, not with whoever builds the setup.
4. Set distribution rules. Who receives what, in what order, with what escalation when nobody responds. Add distribution lists for the cases a rule cannot express. Global lead history will record where each lead went, which is what settles the inevitable argument about whether something was received.
5. Define your territory if you are on a lead program, using the Leads Manager zone editor, so you are buying the geography you can actually serve rather than a default radius. Groups should draw zones deliberately so two of your own rooftops are not competing for the same shopper.
6. Test every single source with a real submission and watch it land. Not a sample record. A real one, submitted the way a customer would, from a phone.

That last step catches more problems than any other item on this page, and it is the one most often skipped because it feels redundant.

Getting the export is a contract question. Getting it clean is a discipline question, and the second one is where the value is.

Ask your outgoing vendor exactly what you can export, in what format, whether notes and communication history are included, who performs the extraction, what it costs and how long it takes. Get it in writing while you are still a customer in good standing, because leverage evaporates the moment notice is given. Our data export checklist covers the questions to ask.

Then clean before you load rather than after. The recurring problems are duplicate customer records created by every source that ever posted, phone numbers stored in six different formats, dead email addresses accumulated over years, and vehicle or deal history attached to the wrong person. All of these are cheaper to fix in a spreadsheet than inside a live system where your team is already working. Our pages on CRM data cleanup and duplicate lead management cover the mechanics.

Decide deliberately what you are not bringing. Records with no valid contact method and no activity in four years are not an asset, they are a drag on every report and every campaign you run afterward. Archive them rather than importing them.

Then import contacts, verify a sample by hand against the old system, and only then let anyone start working. Verifying a sample takes twenty minutes and it is the difference between finding a field mapping error on day one and finding it in month three.

A CRM with no cadence running is a database, and databases do not change behavior. Get this live in week one rather than deferring it to month two.

Build the message library first: text templates and email templates for every touch in your follow up schedule, so an agent is editing rather than composing. Composing from scratch is where the fifteen minutes per lead disappears, and it is the main reason agents drift back to personal phones.

Then build the follow-up processes themselves. Scheduled sends cover the automated touches. Task automation creates the calls that need a human, assigned to a named person with a due time rather than a due day. Drip campaigns with the drip editor handle the longer tail out to twelve weeks. Reminders put the human touches on a calendar with an owner attached. Our page on follow-up cadence covers what to build and in what order.

Set the branches while you are in there, because a flat cadence is only a skeleton. Appointment set starts a confirmation sequence. No show starts a same day recovery sequence. Not ready reschedules to a real date rather than dropping into a generic drip. Bad contact stops the sequence. Opt out stops everything, everywhere, permanently.

Finally configure appointments and calendar settings: who owns the calendar, what an appointment confirmation looks like, and what the reminder schedule is. Appointments that are set but never confirmed are the most common source of a show rate problem, and it is a configuration issue as much as a coaching one.

Not everything belongs in week one. These are the pieces to add once lead handling and follow up are genuinely running, usually somewhere around day thirty to sixty.

Desking. Configure your state tax treatment, fees, lender programs and lease methods before anyone quotes a customer with it. The engine covers loan and lease with a fifty state tax matrix, semimonthly payment frequency, trade credit caps and three lease tax methods, which matters most if you sell across state lines. Then decide whether customer facing deal pages are on, since sending real numbers to a shopper's phone changes how your desk works. See the desking tool page.

Lead pages and salesperson sites. Build one landing page properly rather than six roughly, and give your salespeople their personal sites once there is something worth sending traffic to.

Live chat. Put the widget on your site only once you have decided who answers it and how fast. An unanswered chat widget is worse than none.

Campaigns. SMS campaigns, email campaigns and bulk email, after validation has run on the list. Not before.

Mobile and document tools. Get your team onto the mobile workspace, and turn on document reading and VIN scanning so a driver licence or a VIN can be captured from a phone camera instead of a copier.

Train by role rather than by feature. A salesperson needs six things and will never use the seventh. A BDC agent needs the queue, the templates, the dialer and the disposition flow. A manager needs the reporting and the coaching tools. Teaching everyone everything guarantees that nobody remembers anything.

Keep sessions to forty five minutes, run them in the room where people work rather than in a conference room, and use real records rather than demo data. The video training library covers the follow up questions that always arrive on day three, which is the day the initial training has faded and the real usage has started.

On go live day, do these five things. Confirm every lead source is delivering, with a live test from a phone. Confirm distribution is putting leads in front of named people. Confirm the follow up processes are switched on rather than built but idle. Confirm opt outs came across correctly. And state the one rule out loud in a sales meeting: customer communication happens inside the platform, no exceptions, including for the top producer. That last clause gets tested in week two, and the answer your managers give decides whether this worked.

Then run a five minute stand up every morning for the first two weeks, looking at untouched leads and yesterday's response times by name. Not to punish. Attention is what makes a change real.

Pricing is month to month with no long term contract, from $199 on CRM Only and $799 for programs including exclusive local leads. Details on the pricing page. We cannot guarantee results, but a store that works this checklist in order rarely has a bad first month, and a store that skips blocks two and three almost always does.

What has to be done before anything else?

Opt outs and your blacklist, then lead source configuration, then permissions. Those three are expensive to fix later. Templates, drip content and calendar preferences can be rough at launch and refined afterward without cost.

How long does onboarding take?

Two to four weeks to configured and live for most stores. The variable is rarely the software. It is how clean your outgoing data is and how quickly your team stops working outside the system.

Should we import every old record?

No. Records with no valid contact method and no activity in several years drag on every report and campaign afterward. Archive them instead. Clean duplicates and phone formats in a spreadsheet before loading rather than inside a live system.

Why does email domain authentication matter so much?

Sending campaign email from an unauthenticated domain is how a store lands in spam folders for months. Complete authentication before the first send, and run the list through the email validator so dead addresses do not damage your reputation for the good ones.

Do we need a DMS integration or an inventory feed to start?

No. Neither is required to operate, which shortens onboarding considerably and is why brokers and individual salespeople can run on the platform. If you do have a feed it can be ingested and published separately.

What is the difference between this and an implementation plan?

This is the task list: settings, imports, templates and tests, in the order that avoids rework. The implementation plan is the project structure around it: phases, named owners, the pilot, and the thirty, sixty and ninety day reviews.

We will sit with your administrator, configure the blocks in order, and test every lead source with a live submission before you go live. Month to month, no long term contract.