DMS Outage Plan: How to Keep Selling and Servicing When the DMS Goes Down

A DMS outage plan is a printed playbook that lets a dealership sell cars, write repair orders, move parts, and pay people on paper when the dealer management system is down, with one named person deciding when to switch.

Last updated: October 16, 2026

It’s a Wednesday morning in late June. A couple has picked out the SUV, agreed on the number, and they’re sitting at the desk waiting on paperwork. The salesperson clicks to pull up the deal. Nothing. They click again. Then the service drive calls the sales desk to ask if anybody else’s screens are frozen, and the parts counter is already on the phone with somebody about the same thing.

That’s roughly how June 19, 2024, started at a lot of stores. CDK Global’s systems went down after a cyberattack, and about 15,000 dealerships across North America lost their DMS for close to two weeks.

Some stores kept selling. Some sent people home.

The difference wasn’t luck, and it usually wasn’t the size of the group either. It was whether anybody had decided, ahead of time, what the store does on paper. That’s what this guide covers: the paper deal and repair order packs, the workarounds for payroll and parts, what to print now, and who makes the call. If you’re in the middle of a DMS move, including a Tekion enablement project, this belongs in the plan before go-live, not after.

We manage the technology so dealers can sell and service cars and heavy trucks. An outage is the day that promise gets tested.

What a DMS Outage Plan Actually Is

A DMS outage plan is a short, printed set of procedures that tells each department how to keep working without the dealer management system, which forms to use, where they’re kept, who switches the store to paper, and how the paper gets entered back in once the system returns. It’s an operations document, not an IT one.

That last part matters more than it sounds. Your incident response plan tells IT and leadership how to contain an attack, who calls the insurance carrier, and who talks to lawyers. The outage plan tells the F&I manager what to hand the customer at 10:15 a.m. Different readers. Different binder.

And it doesn’t care why the DMS is down. Ransomware at the vendor, a fiber cut on your street, a bad update, a storm that takes out the building’s power for half a day. The sales floor experiences all of them the same way.

The screen doesn’t load. Now what?

The plan is the answer to that, written down before anyone needs it.

What June 2024 Cost the Stores That Had to Improvise

Anderson Economic Group put a number on it. Their revised July 2024 estimate came to $1.02 billion in direct losses to franchised dealers over the three weeks from June 19 through July 5, including roughly 56,200 new-vehicle sales that didn’t happen, lost earnings on used cars, parts, and service, extra staffing and IT costs, and floor plan interest on cars that sat. That’s real money. By July 2, CDK told dealers that “substantially all dealer connections are live” on the DMS, as Automotive News reported.

Two weeks is actually on the short side. CDK’s own 2023 State of Cybersecurity in the Dealership study put average downtime after a dealership cyberattack at 3.4 weeks. Call it a month.

What hurt most wasn’t always the selling. Plenty of stores were writing deals on paper within days, by most accounts that summer. The pain showed up in the places nobody had thought about: how to run payroll when the hours and flat-rate times live in the DMS, how to sell a part when nobody can look up the price or the bin, and what to do with three weeks of paper once the system came back. I’d bet a lot of service departments spent July catching up on June.

We’ve written about the security lessons from that summer, including why some dealers still haven’t changed anything a year later and why CDK wasn’t a one-time wake-up call. This piece is the other half. The playbook.

Improvising is expensive.

The Outage Kit: What to Print and Keep Before You Need It

Everything below goes in one binder per department, plus a master copy in the general manager’s office and one more copy somewhere off-site. CISA’s #StopRansomware Guide says it plainly: keep “a hard copy of the plan and an offline version” available. A plan that lives only on the network you just lost isn’t a plan.

The exports are the part stores forget. A form is easy to print once and stick in a drawer. A parts price list from six months ago is a parts counter giving away margin, or arguing with every customer, and neither one is great on a day you’re already behind.

Somebody owns the refresh, by name.

Selling Cars With the DMS Down

When I started on the sales floor in the late ’90s, a deal jacket was mostly paper anyway. Everybody in the building knew how to work one. That’s not true anymore, and it shouldn’t be. Your newest salesperson has probably never seen a carbon-copy buyer’s order. So the paper deal pack has to do the remembering for them.

Put every form for a complete retail deal in one pre-assembled jacket, in signing order, with a checklist stapled to the front. Number the jackets before the outage, 001 through 100, and log each one out by number. When the system comes back, that log is how accounting knows nothing went missing.

Then find out which of your lender and credit tools run outside the DMS. Lender portals such as RouteOne and Dealertrack’s credit application network usually run separately from the DMS itself. Which of yours still work if the DMS is down? Ask now, on a slow Tuesday, not during the outage.

Two more decisions belong in writing. Who’s allowed to approve a deal on paper, since the desk can’t see the usual numbers on a screen? And what’s the floor for a paper deal: do you hold anything you can’t fully document, like a trade with an open payoff you can’t verify? The store that decided this in advance doesn’t have a sales manager making it up at 7:40 p.m. with a customer waiting.

The deal still closes. It just closes on paper you already printed.

Service and Parts on Paper

Service usually hurts first. Appointments are on the screen, the customer history is on the screen, and so is the warranty claim. If the appointment list wasn’t printed the night before, the advisors are working from memory and whatever is in the parking lot.

So on the service side, the thing I’d want first is tomorrow’s appointment list printed every night at close and left on the counter, which sounds a little silly right up until the morning you need it.

The repair orders are the bigger deal, honestly. They have to be just as complete on paper as they would be on the screen, and I know that’s a lot to ask with a full drive, but your flat-rate techs get paid off the hours on that RO, and if those hours don’t make it onto the paper, you’re going to have your office manager and a tech sitting there two weeks later rebuilding a paycheck from memory, and the tech is usually going to remember it a little more generously than the office does.

Parts is more about trust. The counter needs a price it believes, which is what that weekly export is for, and every part that crosses the counter needs a numbered ticket or a line on an RO, or your inventory won’t tie out for a month after you’re back up. And while things are quiet I’d check whether your OEM’s parts ordering actually runs outside the DMS, which it usually does, and whether more than one person in the department has a password that works.

Warranty can wait a week. The customer in the drive with a check engine light can’t.

Payroll and Accounting Workarounds

Payroll is the one that keeps controllers up at night. It caught a lot of stores flat-footed in 2024, because the hours, the flat-rate times, and the commission math all live in the same system that just went dark, and Friday shows up on schedule whether the DMS does or not.

People still expect a check. Hourly staff still clock in. Somebody still has to figure commissions.

Decide the rule before you need it. A common approach is to pay each employee the same amount as the last pay period, then true up the difference once the DMS and time clock data come back. Whatever rule you pick, your payroll provider should know about it in advance, and so should your employees, in writing. Keep a printed copy of the last pay register in the accounting binder every cycle so the rule can actually be carried out.

For cash, checks, and card payments, a manual receipts log with numbered entries does the job. Deposits still go to the bank. Every dollar gets a line.

The controller should be able to run this from a binder.

Who Decides, and What Happens in the First Hour

Every store needs one named outage lead, usually the general manager, and a named backup for the days the GM is off. When the DMS goes down at 9:05 a.m., somebody has to answer “do we go to paper?” by about 9:35, and a group text with 12 managers in it has never answered anything that fast.

I’d also write down the trigger, something like: if the DMS is unreachable for 30 minutes during business hours and IT can’t say when it’s coming back, the store goes to paper. Your number might be 20 minutes or an hour. What matters is that nobody has to argue about it that morning.

When the first hour goes well, it tends to look like this. Whoever notices first calls the IT help desk instead of the person at the next desk, because IT needs to know right away whether it’s one PC, one store, or the whole group, and whether it’s the vendor, your internet line, or something worse in your own building. Once IT has a read on it, the outage lead makes the paper call and starts the call tree by phone and text, on the assumption that email is down too. (CISA calls that out-of-band communication, and it’s in their guidance for a reason.) Department heads pull their binders, and the receptionist and BDC get the customer script so every caller hears the same calm answer.

And nobody plugs anything back in or “tries something” until IT clears it. If it’s a cyberattack, a well-meaning manager rebooting the server can make things a lot worse.

That’s where a good IT partner earns the monthly fee. When a store calls our help desk, a live agent picks up in 82 seconds on average, and on an outage morning that’s the difference between a decision at 9:20 and a guess at 11:00.

One person decides. Everybody else knows who.

Getting the Paper Back Into the DMS

Nobody plans the re-entry. It’s the most predictable part of the whole outage, and it’s where the hidden cost lives.

Every paper deal and repair order has to go into the system, in order, accurately, while the store is also trying to run a normal day. It’s tedious work. Staff it like a project:

• Pull one or two people per department off their normal work.
• Enter the oldest numbered packs first, and check each one against its log as you go.
• Accounting reconciles parts tickets against inventory, and receipts against the bank, before month-end rather than during it.

Keep the paper. File it by number and hold onto it until the deals fund and the warranty claims pay. If something entered wrong, the paper is your source of truth. Don’t shred it early.

Plan for re-entry to take longer than the outage did.

Cloud DMS Platforms, DMS Moves, and the Internet Line

Moving to a cloud-native DMS doesn’t make this plan optional. It changes what the most likely outage looks like. Usually it’s the internet. With a cloud platform, the system can be perfectly healthy and still unreachable from your building if the store’s internet connection goes down. That’s why a backup internet connection, on a different carrier and a different physical path into the building, is part of an outage plan and not a separate IT wish-list item.

If you’re switching platforms, this is the right time to write the plan, since you’re already documenting every workflow. Our DMS conversion checklist covers the IT side of a switch, and if you’re working with our DMS transition services team, ask for the downtime procedures to be part of the cutover plan. For stores moving to Tekion, Helion stays on after go-live, extending the momentum Tekion’s own team builds, and keeping the outage binder current is part of that work.

Then test it. Pick a slow weekday afternoon, tell one department the DMS is “down” for two hours, and watch what happens. You’ll find the missing form, the expired login, and the person who didn’t know where the binder was. That’s the point. Whenever something like this doesn’t go the way I want, I ask myself the same question: what am I not doing? A drill answers it cheaply.

A plan nobody has practiced is a document.

What Dealers Say When We Bring This Up

“CDK was a one-time thing. They’re back up.” They are. But outages come from a lot of places, including your own internet line, a storm, or an attack on your store directly rather than your vendor. The plan isn’t about one vendor. It’s about your store.

“If it happens, we’ll just close for a few days.” You could. Payroll doesn’t close, though, and neither does floor plan interest. AEG’s estimate above is what closing looked like across the industry. It wasn’t cheap.

“Our DMS provider has a backup.” They likely do, for their data. Their backup doesn’t print your repair orders or tell your controller how to run Friday’s payroll. Those decisions are yours, and so is backup and disaster recovery for everything the DMS doesn’t hold.

“We’re too small for this.” A single rooftop needs a thinner binder, not no binder. Two pages per department is fine.

Things Dealers Ask About DMS Downtime

The Binder Is the Plan

Nobody wants to spend a Tuesday afternoon printing repair orders for a day that might not come. I get it. But we work with 2,000-plus dealerships, and the pattern I’d put money on is simple: the stores that get through an outage in decent shape aren’t the ones with the best software. They’re the ones where somebody had already decided what happens on paper, and everybody knew who that somebody was.

That’s peace of mind you can actually put on a shelf.

If you want a second set of eyes on where your store stands, from backup internet to who holds the binder, start with a complimentary IT and cybersecurity assessment. We’ll go through it with you and tell you what we’d fix first.

Print it before you need it.