Back to the Tykon.io blog

AI Receptionists

"Send Us Photos and We'll Call You Back": A Phone-Intake Diagnostic for Collision Shops Losing Customer-Pay Jobs

DRP work shows up whether or not you answer the phone. Customer-pay and out-of-network jobs are decided by whoever calls back first. Five checks a collision shop can run this week to find out how much of that work the phone is losing.

by Jerrod Anthraper

Tuesday, 6:15pm. A guy is standing in a grocery store parking lot looking at a crumpled rear quarter panel on a four-year-old Tacoma. He backed into a light pole. No other party, no police report, and with a $1,000 deductible on maybe $1,900 of damage, he's already decided he isn't filing a claim. He calls three body shops from the parking lot. Two ring to voicemail. The third answers and says, "Text some photos to this number and we'll call you back tomorrow." He texts photos to all three. Wednesday at 10am, the first shop to call him back with a number and a drop-off date gets the job. Your shop calls Thursday.

The leak: the jobs that are decided on the phone

Between 40% and 60% of inbound leads arrive after business hours, and most of them go unanswered. For a collision shop, the mix of who's calling at those hours matters more than the raw count. DRP assignments show up in your queue whether or not anyone picked up the phone. The customer-pay job, the out-of-network claim where the customer gets to pick the shop, the fleet manager with a dented van — those are decided by whoever answers first and says something useful. Those are also, ticket for ticket, the jobs where you set the labor rate instead of the carrier.

So the phone leak in a collision shop doesn't hit your volume evenly. It skims the highest-margin work off the top and leaves you with the assigned work, and the assigned work is what makes it feel like the phone isn't a problem.

Why it happens

The only person in the building who can put a number on a damaged quarter panel is the estimator, and the estimator is in the bay, or writing sheets in CCC or Mitchell, or on the phone with a desk adjuster about a supplement. The front desk can't quote, so the front desk defers, and "send us some photos and we'll call you back" is a deferral dressed up as a process. Nobody owns the photo inbox. Nobody has a callback deadline.

There's a second cause that's specific to insurance-heavy shops. When most of your work arrives as an assignment, the customer calling in is usually already committed to you — the carrier sent them. Over the years, that trains the front desk to treat every caller as someone who's already decided. The customer-pay caller in the parking lot hasn't decided anything, and the shop has forgotten how to talk to him.

And the timing works against you. Accidents cluster on the evening commute. The phone rings hardest between 5 and 7pm, exactly when the shop is closing and the estimator has left.

The diagnostic: five checks you can run this week

Each check takes under an hour and produces a number. You don't need to buy anything to run them.

1. Count Monday morning's voicemails and time the callbacks

Next Monday, before anyone clears the box, count the voicemails from the weekend. Then track two things for each: whether it got a callback, and how many hours later. Most shops find that a third never get called back at all, and the ones that do get called Tuesday. Write down the median callback time. That's the number a caller is competing against when they leave the same message at three shops.

Fix you can make today: a 30-minute callback rule during business hours, and whoever opens the shop clears the weekend box before 8:30am.

2. Audit the photo inbox

Go through the last 30 days of texted and emailed photos. For each one, note whether it got a reply, how long that took, and whether the reply contained a number or a drop-off date. If the typical reply is "we'll need to see it in person," you've turned a 30-second exchange into a two-visit process, and the parking-lot caller went somewhere else.

Fix: let the estimator send a rough range from photos with a stated caveat — "looks like $1,600 to $2,200 depending on what's behind the panel, and we can get you in Thursday." A range with a date beats a precise number on Friday.

3. Run the three-question test on your front desk

Ask whoever answers the phone to handle these three, without pulling the estimator: Do I need to file a claim for this? When is the soonest you can look at it? Do you have a rental or loaner arrangement? If any answer is "let me check with someone," the caller is getting a callback promise instead of an answer. Every one of those three has a written answer that fits on an index card.

Fix: write the index card. Include your drop-in estimate windows, your rental partner, and a one-line explanation of when a claim makes sense versus paying out of pocket.

4. Call your own shop at 6:30pm from a cell phone

Do it tonight. Listen to the whole greeting. Does it tell the caller when you'll call back? Does it give them a way to send photos? Does it mention that you're open Saturday morning? Or does it say "we're closed, leave a message" and hang up? Then do the same thing at 12:15pm on a weekday, when the desk is at lunch.

Fix: rewrite the greeting to include the photo number, the callback window, and your next estimate slot.

5. Compare customer-pay conversion to DRP conversion

Pull the last 60 estimates written. Split them into DRP-assigned and everything else. For each group, count how many turned into a repair order. DRP conversion is usually high because the customer arrived committed. If your non-DRP conversion is below half your DRP conversion, the phone and the callback loop are where the difference lives — not your pricing, and not your Google rating.

Fix: track this monthly. It's the single number that tells you whether the intake fixes above are working.

The proof

Leads contacted within five minutes are 21 times more likely to qualify than leads contacted later. The average business response time to an inbound lead is roughly 42 hours, which is almost exactly the "we'll call you Thursday" loop from the parking lot story. For a collision shop, the caller who wants a number from photos isn't being difficult — he's giving you the fastest possible path to a booked estimate. The shops losing that job aren't losing on price or quality. They're losing on the 42 hours.

Where Tykon fits

James is the AI sales agent Tykon deploys for shops like yours. When the 6:15pm call comes in, it answers, asks whether this is a claim or customer-pay, collects the photos by text, tells the caller about your rental arrangement, and books him into your next estimate slot — in the first couple of minutes, not the next business day. Your estimator walks in Wednesday morning to a booked appointment with photos attached instead of a voicemail. This isn't a custom build; it's a framework that's already been through weeks of rigorous testing, so the shop skips the trial-and-error phase. It also handles the review request and the referral ask after the car is delivered, which is why we call it a flywheel rather than a funnel.

The next step is check number four. Call your own shop at 6:30pm tonight. If the greeting doesn't give the caller a way to send photos and a time you'll call back, that's the conversation worth having.