Employing AI: A Different Kind of Tool Box

by Alana Quartuccio

AI. It’s everywhere. Just ask Google:

AI is deeply integrated into daily life. Over 77 percent to 99 percent of people use AI-powered services weekly, though many don’t realize it. Globally, well over 1.5 billion people use standalone AI tools monthly, alongside constant, invisible interactions with algorithms running in the background.

Let that AI-powered answer be a reminder of what this widely-used technology actually is: a tool.

And it can be employed in just about every corner of the body shop.

Have no fear. The robots are not coming – well, not just yet. AI won’t replace your trusted CSR, repair planner, painter or technician. Artificial intelligence is only as good as the human who puts it to work.

“AI is everywhere and the reality is that we are accelerating almost at breakneck speed,” observes Ryan Taylor (Body Shop Booster). “It’s going so fast that it’s exciting and scary at the same time. When we think about how AI can help shops, we think about how it can help them right now. The industry has slowed down. We’re seeing more vehicles being totaled. Shops are pressed for work. They are down and they need help.”

Taylor focused on the seven core areas where shops get work – customer pay, DRPs, agents, fleets, B2B referrals, dealerships and OEMs – to identify how AI can assist shops get more work in all those areas. “One of the things we found is that there are two major leaks in the bucket, some low-hanging fruit where AI can actually help. Number one is lead capture.”

Triage/Lead Generation

“Shops lose a lot of customers on first contact and they don’t even realize it,” Taylor reports. “We work with a company called Canary Calls which has data that shows 43 percent of customers are abandoning the shop after first contact. This means after they have a conversation with someone at the shop, they opt to find an alternative.”

This is where AI communication tools can step in.

“If the CSR appears rushed or busy on the phone, offering no more than a suggestion to come in for an estimate and hang up, the customer won’t come to the shop. AI is really good at talking to customers, collecting the lead information, sending the customer the damage assessment tool and setting up an appointment.”

According to James Spears (Tractable), AI has advanced to the point where shops can begin to assist a customer while the customer is still at home. A tool called Instant Quote can be displayed right on the home page of a website. “They scan the QR code, upload four photos of the vehicle and two more photos of the damage and it will provide an overall quote based on the damage it can see. It links the customer to your shop.”

Spears notes the significance of supplying a customer with the ability to get a price quote instantly at this triage stage. “Before they even report it to their insurance company, they are answering a couple of basic questions that carriers fail to do. When a policy holder calls their carrier, the carrier immediately opens a claim. At this early stage, all the customers want to know is whether the damage will exceed their deductible and how much they may have to pay. They also may be able to determine if it’s a total loss. Offering this to them via your website is very sweet.”

AI provides a way to “staff” your shop after hours or on a weekend when no one is in the building. “AI plays a big, big role in making sure that customers get what they want, which is really faster answers,” adds Spears. “They want speed. They want convenience. They want the answers they are looking for right now and if they are able to get them, they are willing to book with the shop.”

Taylor points to an MIT study which unveiled that if a shop does not respond to a customer’s call or text within five minutes, the chances of them going elsewhere are 78 to 80 percent. He researched it further and found that the window of expectation is indeed five minutes. As he explains,  AI phone systems and texting can help with this, as well as a feature he calls a “shot clock.”

“If a CSR does not respond to a customer after four minutes and 30 seconds, AI will jump in and take over the conversation to get that customer committed to an appointment. Response time matters.”

“We’ve found that it’s not the first shop that responds; it’s the first shop that answers their questions that gets the job,” Spears explains. “AI can help with that.”

AI has had a hand in photo estimating for some time, but Tractable’s LumaScanner is taking it to the next level by generating fast damage appraisal via a drive-thru scanner.

“We’re installing three to four of those per week at dealerships and body shops around the country,” Spears reports. “You drive your car through the LumaScanner and it provides an estimate of damage within 30 seconds, giving you the information you need to have a discussion with the customer. It’s a very fast solution.”

Repair Planning and Estimating

Once you have that customer, AI can be employed to help shops get the most out of every job.

“In a soft market, you do not fix profitability with more cars; you fix it with more complete cars,” says Sergio Palma (CARMIC AI). “Three to six hundred dollars of legitimate, OEM-required work falls off an ordinary bumper estimate, and that is pure margin walking out the door on every repair.”

Repairable claim volume dropped nearly 10 percent last year while severity kept rising, and total losses hit a record 23 percent of claims, he points out. In order to leverage higher, a shop has to capture as much as they can out of every estimate and handle total-loss customers better than any other shop. When volume is down, it boils down to how complete each repair order is.

Palma suggests, “Every operation you leave off is margin you will never get back. Shortfall is the gap between what the estimate pays and what the repair actually takes. It hides in stuff that got forgotten, and in the not-included operations the estimating systems will never add for you because they are not in the labor time. That second bucket is where shops bleed quietly, every day. And the stuff that goes missing is boring, which is exactly why a machine catches it and a tired estimator at 5pm does not. [Things like] post-repair scans. A radar calibration after the bumper comes off. One-time-use clips reused on paper. Blend time, color sand and buff. This is not a fringe issue, either; 28 percent of repairable estimates now carry a calibration, and those get missed constantly.”

“AI doesn’t raise your prices; it stops you from leaving money on the table,” he continues. “It catches the calibration, the one-time-use bolt, the weld-through primer the manufacturer requires and cites the page so you actually get paid. That’s found money on work you’re already doing.”

Palma suggests a shop doesn’t have to buy anything to start using AI. “Drop the estimate, photos and repair plan into Claude/ChatGPT/Grok, feed it the OEM procedures and position statements, and ask it to list the not-included operations this specific damage on this specific car probably needs. You will get a not-included checklist built for the car in the bay instead of a laminated card on the wall.”

Once the customer is secured, AI can also be used to accelerate the sales process. Spears references tools which can help identify ADAS information, and determine what parts need to be repaired or replaced.

These tools can also get the ball rolling with parts ordering. “You can use AI type tools to order those parts for you as soon as you recognize what needs to be repaired or replaced, and what needs to be recalibrated. You can get those on early order.”

If any parts are on backorder, AI tools can also search outside the sphere. Spears says shops normally order parts from distinct delivery areas, but if a part is in jeopardy, these tools can strategically search other sources for the part, ultimately helping the customer by being able to complete the repairs faster once the part arrives.

Failing to sell estimates and a decline in estimate capture rate represent the other hole in the bucket, Taylor identifies. The national average for estimate capture is 56 percent. Capture rate for some shops is as low as 33 percent. Why are shops not closing the deal? As it turns out, they’re not following up on estimates, but AI can be put to the task, he reports.

“Estimate follow up works,” he stresses. “The problem is that shops find it to be time consuming and staff members don’t want to do it. On average, it takes 13.4 days for a customer with drivable damage to make a buying decision. That has gone up from 11 days because customers now have more access to information than ever before. Back in 2009, a consumer would look at three data points – what their insurance company says, their dealership and a friend or colleague. Today, they can look at 12 factors thanks to the ability to research online. They use AI tools like ChatGPT to search. The good news is that they want to make an informed decision.”

Taylor created a tool he calls the 12 Days of Follow-Up that works with any estimating system. “So 24 hours after they receive an estimate, it’ll send a text message. It can also send an email, or leave a voice message that lands on their phone without it ringing. It provides a follow up with a very strategically designed cadence.”

It moves the needle just enough to help shops achieve a five-percent increase in capture rate over a three-month period.

AI can help repair planners get the most out of every sheet. “Manufacturers spell out what is actually required to put the car back safely, and most of it never makes it onto an estimate because no one has time to read 40 pages per repair,” Palma points out.  Procedures such as weld-through primer on a specific seam, or a bracket that is replace-only and cannot be straightened tend to be left off repair plans. “These are not gray-area upsells. They are written requirements for a safe and proper repair, and that is exactly what an AI model is good at: pulling the must-do line out of the procedure and citing the page it came from. Now the operation is not just on the estimate, it is defensible, because you can point to where the manufacturer said to do it. When insurers are pushing back on every line, the citation is what gets it paid.”

Palma emphasizes how much this really helps a shop meet their bottom line. “The estimating systems will never add the not-included operations for you, because they are not in the labor time. That gap is not a rounding error. Multiply a few hundred dollars across every RO and it is the difference between a profitable quarter and a flat one.”

Customer Service

AI virtual assistants have been utilized for some time now, but things have advanced rather quickly!

Want to know how fast? Call or text (432) 465-2978 to interact with “AiMe” of Demo Collision –  an AI tool that can talk to customers, give them an assessment and book an estimate appointment. Taylor says it speaks 32 different languages and can text in 50 different languages. It can also be used to update customers if their vehicle is already at the shop.

AI can be taught everything it needs to know about the shop to deliver information to customers, just like a human would. It can even engage and empathize. “We are seeing customers really connect with AI as long as it’s intelligent, intuitive and well trained. They are enjoying the experience,” Taylor shares. His studies have found there is an 80-percent adoption rate where people are willing to deal with AI, and 20 percent who have not. (Ten percent of the “not-yet-adopted” group does not want to interact with AI at all, while the other 10 percent are on the fence and will if necessary.)

Marketing and Management

In addition to assisting with profitability under the hood, AI can sometimes serve as a capable marketing assistant. “Across 40 different makes and models, we’ve found that 40 percent of the vehicles that drive through the LumaScanner have repairable damage that can be fixed by the body shop. That’s an incredible marketing tool,” Spears notes. “It can immediately print out an appraisal the service advisor can show the customer.” 

AI can even be recruited to help manage workflows and get customers into rental cars.

“Companies like Entegral are using AI tools to help assist with coordinating rental cars, and entering information into shop management platforms,” Spears relays. “You can better manage time and you don’t have to re-key information across multiple platforms.”

Humans and AI Working Together

No matter how far AI advances, the technology won’t replace the human element in the body shop. Spears suggests that AI can carry administrative burdens so the front office staff can deliver the personal care customers deserve. 

The human touch is necessary. Machines can perhaps be more accurate in assessing damage, but it will always need a human to step in and do the blueprinting.

“I don’t believe we will ever get past the point of having a human in the loop to accurately assess damage,” Spears believes.

It may feel like AI has reached its maximum, but according to technology experts, we are still at the bottom tier. Spears considers how AI is used in ride sharing apps, which has completely changed the way one feels when they leave a building and get into a vehicle. “That’s what we need to start to use AI for in this industry. How do people feel when they get into a collision and then how can these tools really help?”

“I equate where we are in history as being no more advanced than cave men painting on walls,” he continues. “I know we think that AI is advanced, but it’s not.”

Taylor agrees it’s just the beginning. “We are just scratching the surface of what we’re working on.”

As for the robots? Taylor admits he has put robots in service lanes before but has found they are not nearly as effective as tools like the LumaScanner. “But maybe in three to five years…”

Want more? Check out the July 2026 issue of Hammer & Dolly!