You Already Paid for the MPI

If I walked into your dealership this week, I would not buy another service tool.

I would sit in the lane.

Then I would listen to one advisor call a customer about an MPI.

If the conversation starts with:

You need...

I already know where the problem is.

You need a control arm.

You need a coolant service.

You need front brakes.

You need an alignment.

You need a cabin filter.

The technician may be completely right.

The advisor may be completely right.

But the customer is not a technician.

They do not think in control arms.

They do not think in millimeters of brake pad.

They do not think in coolant seepage.

They think:

Is my car safe?

Why does this matter?

Can I wait?

What happens if I do?

Are you trying to sell me something?

That gap between what the technician found and what the customer understands may be one of the biggest opportunities sitting inside fixed operations.

And most dealerships have already paid for the tools needed to close it.

Cox Automotive released its 2026 Fixed Operations and Ownership Study in April.

The study included 500 dealership fixed operations decision makers and 2,500 consumers who had a service visit during the previous year.

One number immediately stood out to me.

Customers who reported receiving photos or videos of recommended work also reported spending about $230 more per repair order than customers who did not.

Cox's underlying data puts the averages at roughly $640 versus $410.

Nearly half of consumers said photos or videos made them more likely to approve recommended service. Cox also found that 65% of dealership service customers said photos and videos helped build trust with the provider.

Think about what dealerships already own.

Digital MPI.

Tablet.

Camera.

Texting platform.

Electronic estimate.

Online approval.

Technician notes.

The infrastructure is already sitting there.

Yet plenty of stores still finish the process the same way they did years ago.

The technician finds something.

The advisor calls.

Then starts speaking shop.

I do not think the next breakthrough in the service lane requires another piece of software.

I think it requires better translation.

Dealers have solved the inspection problem better than they have solved the communication problem.

The technician knows what they are seeing.

The advisor usually knows what the technician means.

The customer often does not.

That creates friction.

Take a technician note like:

RF lower control arm bushing torn. Excessive movement. Recommend replacement.

Inside a dealership, everyone understands that sentence.

To the customer, it sounds like another repair recommendation.

Now translate the same finding:

The rubber bushing that helps keep your right front wheel stable is torn and has too much movement. It should be replaced because continued wear can affect how the vehicle handles and can lead to uneven tire wear.

Same finding.

No additional diagnosis.

No scare tactic.

No invented damage.

Just English.

That is where I think AI can become extremely useful in service right now.

Not diagnosing the vehicle.

Not replacing the technician.

Not autonomously selling repairs.

Translating technical information into language a vehicle owner can understand.

That is a much smaller use of AI.

It may also be one of the easiest to implement.

Photos and videos matter because they change the conversation.

An advisor saying:

Your control arm bushing is torn

requires trust.

A customer seeing the torn bushing while someone explains what it does requires less imagination.

The evidence is sitting in front of them.

That matters because one of the biggest obstacles in service is not necessarily price.

It is suspicion.

Cox found that 13% of frustrated service customers specifically felt they were being pushed toward additional services.

At the same time, customers who received visual information reported greater trust and greater willingness to approve the work.

That should tell us something.

The customer does not necessarily want fewer recommendations.

They want to understand why the recommendation exists.

A picture without an explanation can still confuse them.

An explanation without a picture still asks them to trust us.

Put the two together.

Now you have something powerful.

Show me what you found.

Tell me what it means.

Tell me what happens if I wait.

Then let me make the decision.

That is a much better service conversation.

This is where I would keep the process simple.

I would not give an AI system unlimited access to customers tomorrow.

I would use it behind the advisor.

Technician writes the finding.

Advisor or approved dealership workflow passes only the necessary non identifying information into an approved AI environment.

The AI translates the technical note.

Advisor reads it.

Advisor checks it against the technician's actual recommendation.

Advisor edits anything that sounds wrong.

Then the approved explanation goes to the customer alongside the photo or video.

The advisor still owns the conversation.

That matters.

AI can write a sentence that sounds confident while still misunderstanding an unclear technician note.

So the system should never be allowed to invent.

If the technician says:

Coolant seep at water pump

the AI should not suddenly tell the customer the water pump has failed.

If the note does not explain severity, the AI should not manufacture urgency.

If information is missing, it should say the information is missing.

The technician owns the diagnosis.

The advisor owns the customer.

AI owns the translation.

That is the lane I would start with.

I would not overengineer this.

Start with ten declined MPI recommendations from yesterday.

Remove all customer identifying information.

No customer name.

No phone number.

No email.

No full VIN.

No repair order number.

Use only the technical finding and the basic vehicle context needed to understand it.

Then run this:

Rewrite this multi point inspection finding for a vehicle owner. Use eighth grade English. Explain whether it relates primarily to safety, maintenance or normal wear. In one sentence explain what could happen if the customer waits. Do not add parts, prices, failures, damage or urgency that are not contained in the technician note. If the note does not contain enough information, state what information is missing.

Then make one rule mandatory:

A human approves every explanation before it reaches the customer.

That is it.

No six month implementation.

No new platform required to test the idea.

Take ten dead recommendations.

Translate them.

Match each one with the inspection photo.

Then see whether the customer response changes.

There is another reason I would start now.

According to Cox, 16% of consumers already used an AI website or tool during their most recent service journey.

They used AI to research providers, compare options and better understand potential service needs.

That number is only going higher.

Think about what happens when the advisor sends a customer this:

Recommend RF lower control arm due to failed bushing. $847.

What does the customer do?

Increasingly, they copy it into ChatGPT.

What is a control arm?

Is this dangerous?

Does this really cost $847?

Can I drive another month?

Can an independent shop do it cheaper?

The customer is translating the repair without us.

That is a bad place for the dealership to surrender control of the explanation.

We should translate it first.

Clearly.

Accurately.

Without exaggeration.

Then if the customer wants to research it further, fine.

At least the first explanation came from the people who actually inspected the vehicle.

Here is the part of the Cox study that should bother fixed operations leaders.

Average dealership service and parts revenue reached approximately $9.23 million per dealership in 2025, up 33% over eight years.

Sounds great.

Except dealership share of service visits fell from 33% to 29%.

General repair facilities and other service alternatives are taking share.

And the customer perception does not completely match the economics.

Cox found average customer spend of $261 at dealerships versus $275 at general repair facilities.

Yet many customers still perceive the dealership as the more expensive option.

That is not simply a pricing problem.

That is a perception problem.

And perception is heavily influenced by communication.

If the independent shop says:

Come look at this. Your tire is wearing through the inside edge because this part is loose. If we leave it, you are probably going to keep eating tires.

And we say:

Your right front control arm bushing failed. Do you want to authorize $847?

Who sounds more trustworthy?

Maybe both shops diagnosed the exact same issue.

One simply translated it better.

This gets bigger than today's repair order.

Cox found that 80% of new vehicle buyers say they are likely to return to the selling dealership for service.

But only about a quarter actually leave the purchase process with their first service appointment scheduled.

Among buyers overall, only 23% had that first appointment booked.

That is an incredible gap.

The customer is basically telling us:

I am willing to come back.

And the dealership responds:

Perfect. Hopefully we remember to call you.

Then consider what happens if we lose them.

Cox estimates losing a service customer can represent more than $12,000 in potential lifetime service spend.

But even that understates the importance of the lane.

Customers who return to the dealership for service are 30 percentage points more likely to buy their next vehicle from that dealer.

Cox reports 74% repurchase likelihood among customers who returned for service versus 44% among those who did not.

And 88% of consumers say their service experience affects whether they would buy another vehicle from that dealership.

That makes the service advisor part of the next sales transaction whether we treat them that way or not.

The repair order is today's revenue.

Trust is tomorrow's car deal.

If I were running fixed ops this week, I would hold two meetings.

Bring ten declined MPI recommendations from yesterday.

Not the easy approvals.

The ones that died.

Look at the technician note.

Look at the photo.

Then look at what the customer actually received.

Could a normal vehicle owner understand it in fifteen seconds?

Does it explain why the recommendation matters?

Does it distinguish maintenance from safety?

Does the photo actually show what we are talking about?

Then rewrite the explanation in plain English.

Use AI if it helps.

Advisor approves it.

Technician confirms anything unclear.

Send the photo first.

Then explain what the customer is seeing.

Track the result.

Ten MPIs.

Approved or declined.

Then do it again tomorrow.

No new vendor.

Look at every vehicle delivered last week.

How many customers left with their first service appointment already scheduled?

Cox says 80% of new vehicle buyers are inclined to return to the selling dealer for service.

Why are we letting willing customers leave without putting a date on the calendar?

The first service appointment should be part of delivery.

Not something service hopes to recover three months later.

Sales gets the customer.

Service keeps the customer.

Then service helps bring them back to sales.

That loop should be designed.

If I ran the store, I would make five changes.

For any recommendation the customer cannot physically see or easily understand, show the evidence.

No unexplained technician language in customer communication.

Translate it.

Do not dumb it down.

Make it understandable.

Use AI to help translate technical findings, not replace the technician or advisor.

Human review stays mandatory.

Do not just track what was declined.

Review why.

Was it price?

Timing?

Trust?

Poor explanation?

No visual evidence?

A recommendation the customer never understood should not automatically be treated as price resistance.

The first repair order starts in the sales department.

Book it while the relationship is strongest.

I think the next competitive battle in fixed operations will be around clarity.

Dealerships already have technical expertise.

They already have equipment.

They already have OEM information.

They already have trained technicians.

They already have customer data.

The problem is translating all of that expertise into something a normal person understands quickly enough to make a confident decision.

AI makes that translation layer dramatically easier.

Eventually, the service agent may do much more.

Read the MPI.

Understand the vehicle history.

Identify previous declined work.

Draft the explanation.

Attach the correct photo.

Know whether the issue is maintenance or immediate attention.

Prepare the estimate.

Answer basic customer questions.

Flag complicated conversations for the advisor.

But I would not start there.

Start with one sentence.

Make the technician understandable.

That alone could change the repair order.

You already paid for the camera.

You already paid for the MPI.

You already paid for the tablet.

You probably already paid for the texting platform.

I would not buy another service tool until I knew we were actually using the tools we already own to make the customer understand their vehicle.

Cox's data gives us a pretty clear signal.

Customers who receive photos and videos report spending more.

They are more likely to approve recommended work.

They report greater trust.

Meanwhile, dealerships are still losing service visit share.

The opportunity is sitting there.

The technician finds the problem.

The camera proves it.

AI can help translate it.

The advisor creates the trust.

And the customer still makes the decision.

That is how I would use AI in the lane today.

Not to replace the advisor.

To make the advisor easier to understand.

Because the customer should never have to open ChatGPT after your phone call just to figure out what you were trying to sell them.

The photo goes first.

AI helps write the English.

The advisor still owns the yes.

Pull ten declined MPI recommendations from yesterday.

How many would the average customer actually understand without calling someone else?

That is the number I would start with.

What are you seeing in your lane? Are photos and videos actually changing approval rates, or are we still sending customers technical information and hoping they understand it?

Follow Rooftop Insider on LinkedIn and join the conversation.

I'm Kamil Grzych. I work inside automotive retail, with experience across dealership sales, BDC operations, digital retail, customer experience and technology implementation across Toyota, CDJR and Lexus dealerships.

I started Rooftop Insider to examine what is changing inside automotive retail from the perspective of people actually working inside dealerships.

Rooftop Insider is an independent automotive retail publication covering AI, technology, inventory, pricing, digital retail, customer behavior and dealership operations.

The goal is simple.

Separate the signal from the noise and explain what changes actually mean for the people running dealerships.

Rooftop Insider articles are based on independent analysis and real world dealership experience. No company referenced in this article paid for or approved its inclusion.

Source: Cox Automotive 2026 Fixed Operations and Ownership Study, published April 9, 2026.