The Late-Night Lockout Calls You Are Not Answering, and the Jobs They Cost You

The 9pm call you let ring out is tomorrow's lost job

Picture the owner of a six-truck mobile locksmith shop. It is a Tuesday. Her dispatcher, Marco, spent the morning on the phone with a property manager who wanted a rekey quote on eleven units, then burned forty minutes chasing three techs across town because the job board is a dry-erase whiteboard next to the coffee machine. The shop runs on Housecall Pro for invoicing and a lot of loose texting. It mostly holds together. Until the phones go quiet at 6pm.

That night a college kid locked out of his apartment calls the main line at 9:47pm. It rings four times and drops to a voicemail box that Marco clears at 7am. By then the kid has called the next locksmith on the Google map, paid $165 for a fifteen-minute open, and left that guy a five-star review. The owner never sees the call. It does not show up as a lost job in any report, because it was never a job. It was just a ring nobody answered.

Multiply that by the three or four after-hours calls the line drops every week, add the two commercial estimates sitting unread in the inbox and the follow-ups nobody made, and you have the real leak. It is not measured in bad months or angry reviews. It is measured in hours of unpaid phone tag and in jobs that quietly went to whoever picked up.

Problem: the phone stops being answered at 6pm. The lockouts do not stop.

Lockouts keep no business hours. A stranded driver at midnight, a landlord locked out on a Sunday, a bar manager who snapped a key in the back door at 1am: these are the highest-intent, highest-margin calls a locksmith gets, and they land when nobody is at the desk. The shop's answer today is voicemail or a $1.50-a-minute answering service that reads a script, cannot quote a lockout, and cannot see which tech is on call.

Housecall Pro and Jobber will sell you an online booking form. A person in a parking garage with 9% battery is not filling out a form. That is the gap the off-the-shelf tools cannot close: they assume the customer comes to a screen, when the customer is dialing a phone in a panic.

A custom build puts an AI voice agent on the main line the moment it goes after-hours. It answers in two rings, asks the three questions that route a locksmith job (residential or commercial, vehicle or building, exact location), checks the job against your service area and on-call rotation, quotes a price band, and books it. Then it texts the on-call tech the address with a one-tap accept and writes the whole thing into Housecall Pro as a booked job before the customer hangs up. The 9:47pm call becomes a dispatched ticket instead of a 7am regret.

Problem: the eleven-unit rekey quote is still in the inbox three days later

The property manager Marco quoted wanted eleven units rekeyed plus a small master key system. That estimate is worth more than a week of lockouts. It went out once, as a PDF, and nobody touched it again. By Friday the manager has three quotes and picks the one whose salesperson called back. Yours was not it.

Housecall Pro can fire one automated reminder. It does not know that a $4,000 commercial rekey deserves a different chase than a $90 house call, it never references the actual units or the master key detail, and it never escalates the big ones to a human. Estimates rot on a timer.

A custom follow-up engine watches every unaccepted estimate over a threshold you set. It sends timed, plainly written follow-ups that name the real job (the eleven units, the master key system, the day you walked the property), spaced the way a good closer would space them. When a high-value quote goes cold, it drafts a message and drops it on the owner's phone to send with one tap. It also tags why quotes stall, so you learn which prices and property types close and which never do.

Problem: your best tech finishes eight jobs a day and asks for zero reviews

The kid from the parking-lot lockout gave his five stars to a competitor. Your tech, who rekeyed a storefront and cut four keys that same night, asked for nothing, because he was already driving to the next call. Reviews are how the next stranded customer picks a locksmith off the map, and your crew is too busy to chase them.

The review prompt inside most field tools blasts a generic request at everyone on the same schedule, or it never gets switched on at all. It fires while the customer is still annoyed, or a week later when the relief has worn off.

A custom flow triggers on the invoice being paid, at the moment the customer is standing back inside their house or car and grateful. It names the tech and the job, sends happy customers to your Google profile, and routes anyone who sounds unhappy to a private inbox so the owner hears it before the internet does. Reviews stop being something a tired tech has to remember and become something the job produces on its own.

Problem: two trucks cross town while a third sits three blocks from the next call

At 2pm the whiteboard says truck three is free, but truck three carries the automotive gear and the next job is a commercial safe. So the dispatcher sends truck one across the city, past a rekey it could have knocked out on the way, and when a lockout comes in there is no clean slot for it. The day becomes a map of wasted miles.

The schedule board in ServiceTitan or Jobber shows you appointments. It does not know a lockout is urgent and a scheduled rekey is not, it does not know which truck carries automotive tools versus safe tools, and it will happily double-book the one tech who can open a safe.

A custom dispatch brain holds in real time what lives in Marco's head: each tech's location, the gear on each truck, the urgency of each job, and the gaps in the day. It slots an emergency lockout into the nearest capable truck's dead time without bumping the paid commercial install, and re-sequences the route so nobody drives past a job they could have done. That is billable hours pulled out of thin air, every day.

Problem: six years of jobs in your CRM (Customer Relationship Management) that nobody has ever asked a question

Every locksmith on Housecall Pro or ServiceTitan is sitting on a mine nobody has dug: years of customers, quotes, and closed jobs. The property manager who bought one rekey in 2022 and was never called again. The commercial accounts on a natural re-key cycle. The July spike in car lockouts. It is all in the database, and the canned dashboard shows none of it.

Off-the-shelf reporting gives you revenue by month and a job count. It cannot tell you which dormant accounts are worth a call this week or which price points on which job types actually close.

A custom build reads that history and turns it into work. It surfaces the property managers who bought once and vanished, flags the commercial accounts due for a re-key cycle, and feeds those names straight into the follow-up engine and the dispatch plan. The jobs you already earned become the cheapest pipeline you will ever run, because you paid to win those customers once already.

What this costs, honestly

These bands come from Digital Heroes delivery experience across more than 2,000 projects, not a rate card. A focused first release, usually the after-hours AI phone agent that answers and books lockouts, typically runs $50k to $120k and ships in 10 to 16 weeks. A full operations platform that also owns dispatch, quoting, follow-up, and invoicing runs $150k to $350k, phased over 6 to 12 months so you are live on the phone agent long before the rest is done.

What pushes a locksmith build toward the top of the band is specific to this trade. A shop running automotive, residential, commercial, and safe work needs four sets of dispatch rules and gear logic, not one. The voice agent has to hold up on genuinely bad audio, because your callers are outdoors, stressed, and often inside a concrete parking structure. Deep two-way sync with the Housecall Pro or ServiceTitan API, multiple locations, and real on-call rotation and SLA logic all add build. A single-trade, single-location shop lands near the floor.

When Housecall Pro is genuinely enough, and when it is not

If you run one or two trucks, do not build anything. ServiceTitan, Jobber, and Housecall Pro are good products, and at that size the invoicing, scheduling, and reminders they give you are plenty. Spending six figures to save Marco an hour a day does not pay back when there is no Marco.

The signals that it is time to build, or to layer AI on the CRM you already have, are concrete. You are missing or turning away after-hours calls. You run multiple crews and dispatch lives on a whiteboard. You have thousands of past customers nobody markets to. Your quotes close below where they should because nobody chases them. You are a private-equity-backed rollup trying to run five acquired shops on four different tools. The position: in almost every one of those cases, do not rip out Housecall Pro. Keep it as the system of record and build the phone agent, the follow-up engine, and the dispatch brain against its API. Replace the tool only when it physically cannot hold your data, not because it feels dated.

How to choose a developer for locksmith software

Ask whether they will ride along on a truck for a day before they write a line of code. A developer who has never watched a tech pick a lock at 11pm or seen a dispatcher juggle the whiteboard will build something that demos well and dies in the field. This is the best filter you have.

Make them prove the AI phone agent on a deliberately bad call before you sign. A clean demo in a quiet office proves nothing. Your real callers are panicked, outdoors, and down to half a bar in a garage, and that is the only test that counts.

Confirm they can work against the Housecall Pro or ServiceTitan API and its rate limits, not just stand up a new island that forces your office to enter every job twice. Integration is the unglamorous part, and it is where most of these projects quietly fail.

Settle ownership in writing before the project starts. You own the source code, the customer data, the phone number, and the trained voice agent, full stop. Ask a candidate directly what happens to all of it the day you walk away, and listen to how fast they answer.