A driver just clipped a guardrail. She's pulled over on the shoulder, hands still a little shaky, scrolling through five body shops that came up in her search, calling each one to ask if you can look at it, roughly what it'll cost, and whether you work with her insurance company. Your estimator can't get to the phone — he's elbow-deep in a teardown on bay two, trying to find out how far the damage actually goes before he writes anything down. By the time he's free, she's already booked a drop-off with the shop that picked up. Here's what's safe to hand to a system about that first call and the "where's my car" calls that follow it, and what still needs a trained estimator's eyes on the vehicle.
Collision repair has two separate call-answering problems, not one. The first is the new-estimate call, usually from someone who just had an accident and is comparing two or three shops before she picks. The second comes weeks later, from that same customer or an insurance adjuster, asking for a status update while the car sits mid-repair. Both are mostly informational and safe to hand to a system: confirming whether you work with her insurer, giving a ballpark estimate range and the next available drop-off date, and relaying honest status updates on parts and approvals. What can't be automated: the actual damage assessment, any pricing tied to a teardown or a supplement, and anything involving frame or structural damage — those need a trained estimator's eyes on the car, every time.
Before anything else: when this guide says AI agent, it means a software system that looks at what's happening — an incoming call from someone who just had an accident, a text asking whether you work with a specific insurance company, an adjuster's email asking for a repair status — figures out what's actually being asked, and answers or routes it the way a well-trained service advisor would, without pulling your estimator off a teardown to do it. It doesn't assess damage or price a repair. It handles the answerable, logistical layer around the repair so your estimator's judgment goes where it actually matters: on the vehicle.
A collision shop's calls split into two very different problems. New-estimate calls come from stressed, unfamiliar customers comparison-shopping shops the same way they'd compare contractors — fast, and usually by phone or text with a photo attached. Status calls come later, from a customer or an insurance adjuster, asking about a car that's already in the shop. Industry data shows why both matter: the average total cost of repair reached over $4,730 in 2024, up roughly 3.7 to 3.8 percent year-over-year, and cycle times now run 13 days for a repair that needs no sensor recalibration, but 15.5 to more than 17 days once one or more calibrations are involved.1 "Calibration" here means realigning a vehicle's driver-assist cameras and sensors after a repair, which many manufacturers require to keep features like lane-departure warning working correctly — and it's become common enough that it now measurably extends how long a car sits on your lot, and how many "is it ready yet" calls come in while it does.
None of that is a staffing problem you fix by hiring a full-time front-desk person. A two-bay shop taking 15 to 20 new estimate calls a week, on top of ongoing status calls for every car already mid-repair, is fielding more phone volume than any one estimator can handle between teardowns — and the moments a new customer calls are usually the exact moments your estimator is under a hood, not free to pick up.
| What happens | Left alone | Automated first |
|---|---|---|
| Driver calls minutes after an accident while your estimator is mid-teardown | Rings out; she calls the next shop on her list | Answered instantly, vehicle and insurer confirmed |
| "Do you work with [insurer], and roughly what would this run?" | Caller gives up before finding out, books elsewhere | DRP status confirmed, honest ballpark range given |
| "There's a weird noise from the wheel well — is that from the crash?" | (same either way — needs an estimator's eyes, always) | Never automated — flagged straight to your estimator |
| Customer or adjuster calls mid-repair asking for a status update | Waits on hold or a callback that may not come today | Real parts/approval status relayed the same minute |
Take the free 2-minute readiness assessment
Illustrative example, not a real client: say a two-bay collision shop gets around 20 new estimate calls a week. If 5 of those land while an estimator is mid-teardown or on the phone with an adjuster and go to voicemail, and half of those callers book with the next shop on their list instead of waiting, that's roughly 10 jobs a month — at something close to the current average repair cost of just over $4,730 — going to a competitor. That's close to $47,000 a month in repair work that never even makes it onto the estimate board, on top of whatever status-update calls are quietly eating into the same estimator's day.
This is a fair worry, and it deserves an honest answer, not a sales one: nothing about frame damage, structural integrity, or a total-loss judgment should ever come from a system — that's exactly the kind of call that needs to reach your estimator directly, every time, no exceptions. What a system should do is get someone a fast, honest answer on the logistics and get out of the way the moment things get technical or emotional. Picture what a caller might actually hear, illustrative only, not a real transcript: "Hi, this is the estimate line at Ridgeline Collision Center — I want to be upfront that I'm automated. I can get a few details about the accident and your insurance, give you a ballpark based on similar repairs, and get you scheduled for a full estimate this week, or connect you with our shop manager directly if you'd rather talk it through first. Which works better?" That's not replacing an estimator's judgment about whether a strut is bent or a frame rail is compromised. It's making sure a driver who just had an accident isn't left checking her phone while she works down her list of shops.
This isn't a generic call-answering script — it's what Unmanually actually builds for a business like yours: a system set up around your real DRP relationships, your real cycle-time expectations, and clear rules for what counts as a straightforward estimate or status question versus something that needs your estimator's own eyes on the vehicle, with an immediate handoff the moment a call involves damage assessment, pricing, or anything that sounds like it needs a person right now.
We also back this with a real guarantee, not a vague promise: try Unmanually for 60 days, and if it isn't saving your business real time, whatever's left of your prepaid balance converts to account credit. That's not a cash refund on usage you've already consumed, since that reflects real infrastructure cost already spent, but it does mean you're never stuck paying for a system that isn't earning its place.
For a collision shop, the growth angle is direct: every estimate call you answer fast is a car on your schedule instead of a competitor's, and every status update you relay without making a customer or an adjuster chase you is one more reason a DRP relationship stays healthy and referrals keep coming. Catching more of that call volume without hiring a full-time front-desk person is revenue and reputation you were already positioned to capture.
Not ready to commit to anything yet? That's completely fine — leave your email on our presale waitlist and we'll let you know as soon as estimate and status-update automation for collision repair shops is live, including founding-member presale pricing before it opens to everyone else this October.
1. CCC Intelligent Solutions, "Crash Course Report, Q3 2025," reporting an average total cost of repair (TCOR) of more than $4,730 in 2024 (up roughly 3.7–3.8% year-over-year), and average keys-to-keys repair cycle times of 13 days for repairs requiring no ADAS calibration versus 15.5 to more than 17 days for repairs requiring one or more calibrations.
The full pillar guide this article belongs to.
Another auto-adjacent business where the fastest honest reply tends to win the job.
More on why the same lead often goes to whichever business calls back first.
More on the handoff pattern — where automation should stop, referenced above.
It can give a reasonable ballpark based on the vehicle, where the damage is, and whether the customer's insurance company is one you already work with, clearly labeled as an estimate. It can't give a final, bindable number, because collision repair pricing depends on what a trained estimator finds once the vehicle is actually torn down — hidden frame or structural damage, whether a part can be repaired or has to be replaced, and whether the vehicle needs sensor recalibration after the repair. Most shops already work this way even with a person answering the phone: a rough range up front, a firm number after someone actually opens the vehicle up.
Not if it's upfront about what it is, moves fast, and gets out of the way the moment things get complicated. A driver who just had an accident usually isn't looking for a long conversation — she's looking for someone to tell her you can help, roughly what it'll cost, and when she can drop the car off, so she can stop calling around. A system that's honest about being automated, quotes off real information, and hands off immediately to a person for anything emotional or unclear tends to read as fast and competent, not impersonal. The actual repair, and any conversation about frame damage or a total loss, still comes from a person.
Take our free 2-minute readiness assessment — it walks through this same test against your actual call-answering process and tells you honestly what to automate first.
Take the 2-min readiness assessment