The First 90 Seconds of a Locksmith Call (2026): A Second-by-Second Handling Standard

The First 90 Seconds of a Locksmith Call (2026): A Second-by-Second Handling Standard

Most locksmith shops do not have a call standard. They have a person, and the call goes however that person is feeling. On a good day that produces a booked job in seventy seconds. On a bad day it produces a hold, a vague range, a promise to call back, and a customer who is already dialing the next result.

As of August 2026, the opening of an inbound call is the highest-leverage ninety seconds in a locksmith business, and it is completely standardizable. Everything a caller needs to hear in order to book fits inside that window, in a fixed order, with a rough time budget for each part. This piece is that standard: five segments, what must happen in each, the four things that must never happen, and how to measure whether your shop actually does it. It is written to be printed and taped next to the phone.

The five segments at a glance

Segment	Budget	What must happen	Fail signal
Greet	0 to 5 seconds	Live human-sounding answer with the business name, no menu, no hold	Rings past four, a menu plays, or the greeting does not name the business
Classify	5 to 20 seconds	Establish vehicle, residential, commercial, or existing customer before anything else	Quoting or scheduling starts before the situation type is known
Capture	20 to 45 seconds	Job-defining data in fixed order: what, where, key status	Data collected in random order, or the caller has to repeat something
Quote	45 to 70 seconds	One complete all-in number with trip charge and any surcharge inside it	A range, a starting-at price, or a base that grows on arrival
Close	70 to 90 seconds	Concrete arrival window offered, job booked, callback number confirmed	Call ends without a booked window or without the number read back

Ninety seconds. Every segment feeds the next one, which is why order matters more than speed.

0 to 5 seconds: answer live, name the business

The first five seconds are not about information. They are about proving a human-scale interaction is happening.

The rules are short. Answer within three or four rings. Lead with the business name, then a name, then an offer to help. Never open with a menu, never open with a hold, and never open with a recorded greeting that runs longer than a breath.

The reason this window is so unforgiving is that the emergency locksmith caller is not evaluating you. They are triaging. Someone standing next to a locked car at 9 PM has a results page open and three numbers queued. They are not going to listen to options one through four, and they are not going to hold. A menu at the front of an emergency call is a hang-up with extra steps.

Speed to answer is measurable and it correlates directly with whether the call becomes a job, which is broken down in speed to lead for locksmiths. It also matters in paid local channels where answer rate is an operational input rather than a vanity metric, covered in Local Services Ads answer rate.

One more rule for this segment: the greeting is the same every time. Not a script the person resents, just a fixed opening line. Callers form a judgment about competence in the first sentence, and improvised greetings deliver a different level of competence depending on who is holding the phone.

5 to 20 seconds: classify before anything else

This is the segment shops skip, and skipping it is why calls go long.

Before a single question about price, address, or timing, establish which of four situations you are in.

Vehicle. Car key, fob, lockout, all keys lost, ignition. Different data, different pricing, different tooling, and by far the most parts-sensitive category.

Residential. Lockout, rekey, lock change, broken key extraction. Property type and lock type drive everything downstream.

Commercial. Access control, master keying, panic hardware, storefront. Frequently a property manager rather than the end user, frequently an account, frequently a purchase order rather than a card.

Existing customer. A callback about a job in progress, a warranty issue, a redo, or a question about an appointment. This one has to be caught early because it changes the entire call. Nothing about the standard below applies to a customer who already has an open job, and running an existing customer through a fresh intake is one of the most irritating experiences a service business can deliver.

Classification is one question and about ten seconds. It is worth the ten seconds because everything after it is different depending on the answer, and getting it wrong means collecting the wrong data and having to start over at second forty.

The other thing this segment does is triage urgency. A person locked out on a highway shoulder and a person who wants a spare key made next week are both valid calls with completely different handling. Sorting emergency from scheduled early is what keeps the emergency from waiting behind the appointment, and the mechanics of that split are in emergency versus scheduled triage.

20 to 45 seconds: capture the job-defining data in a fixed order

Twenty-five seconds, three items, always in the same sequence. The order is not arbitrary; each item narrows what the next one needs to be.

First: what the job is, in enough detail to price it.

For a vehicle: year, make, model, and key type. Key type means smart key or push-to-start, remote head, transponder, or plain mechanical. Most callers do not know the term, so ask what they do with it: do you push a button to start it, or turn a key. That question gets a usable answer from everyone.

For residential or commercial: property type, lock type, and how many. A rekey of four exterior locks and a single deadbolt change are not the same job and should not enter the pipeline the same way.

Second: where.

Full address, and for a vehicle in a parking lot, the business name and which side of the building. Spoken street addresses are the most error-prone data on any locksmith call, so confirm it back before moving on. If the caller genuinely does not know where they are, which happens constantly in mall and hospital lots, sending them a location link to tap is faster and more accurate than any amount of describing.

Third: whether they still have a key.

This single question changes the job more than any other. A customer who has one working key is in a duplication or spare-key job. A customer with no working key at all is in an all-keys-lost job, which is a different procedure, a different price band, and often a different part. Ask it every time, even when it seems obvious, because a caller who says they are locked out sometimes means they have a key sitting on the passenger seat, and that is a completely different job.

Two rules govern this segment. Ask in the same order every call, because a fixed order is what makes it fast and what stops items from being forgotten on a busy night. And never make the caller repeat something. Repetition is the clearest signal to a customer that nobody is writing anything down, and it is the moment they start wondering whether the tech will show up at the right place.

45 to 70 seconds: give one complete all-in number

This is the segment where jobs are won and lost, and the rule is a single sentence: say one number, and have everything already inside it.

Everything means the service price, the trip charge, and any after-hours or weekend surcharge. Not mentioned as separate items to be added later. Inside the number. The caller should hear a total that will match the invoice.

Say it plainly, then say what it includes, then stop talking. Silence after a price is not awkward, it is the close working.

Three things are forbidden here and each deserves naming.

A range is not a price. Somewhere between one number and another is heard by every caller as the lower number, and it will be quoted back to you when the tech arrives with the higher one. A range also signals that you do not know your own costs, which is not a good look on the call where the customer is deciding whether to trust you with their vehicle.

A starting-at price is a range wearing a costume. Same problem, same result, plus it primes the customer to expect the real number to be higher, which reduces the chance they book at all.

A base price that grows on arrival is the worst outcome available. It converts a booked job into an argument in a driveway, and then into a review that specifically says you quoted one price and charged another. That review sits in front of every future emergency caller comparing three shops. If your pricing genuinely has a fork, such as a model year that straddles two immobilizer systems, name the fork on the call and give both numbers before the truck moves. A named contingency is fine. A surprise is not.

There is one legitimate contingency and it must be stated out loud during this segment, not discovered later: when the key type cannot be confirmed until a tech sees the vehicle, the caller hears the firm price for the identified key type and the alternative price if it turns out to be the other one. That is honest and customers accept it. Silence followed by a higher number in a driveway is not.

70 to 90 seconds: close with a window and a number

The last twenty seconds are the close, and most calls fail here by simply not doing it.

Offer a concrete arrival window. Not soon, not within the hour if you cannot be, not a range so wide it means nothing. A window a customer can act on is roughly an hour wide with a real start time. Nine to five is not a window, it is a shrug.

Book it on this call. The appointment gets created before the customer hangs up, not after somebody checks something. A caller who hangs up unbooked is a caller who is still in the market, and the shops that are still in the market with them are the ones answering their phones right now.

Confirm the callback number and read it back. Read the digits back for confirmation rather than assuming caller ID is right. A meaningful share of callers are on a phone that is not theirs, or calling from a business line that nobody answers, and a job with a wrong callback number is a job you cannot rescue when anything changes.

Send the confirmation immediately. Text with the window, the address you captured, the price you quoted, and the tech or shop name. This is your last chance to catch a wrong address before it costs a truck roll, and it is your written record of the price.

The four things that must never happen in the first 90 seconds

Everything above is what to do. These four are what to never do, and each one is a job-loser on its own.

Never put an inbound call on hold. The hold is the single most expensive button in a locksmith shop. An emergency caller with three tabs open does not wait, and a caller who was going to book does not stay through a hold to find out. There is no version of the message on hold that is more valuable than a live conversation. If you are on another call and cannot answer this one, the answer is not a hold. It is coverage, which is a structural problem rather than a manners problem, addressed in simultaneous calls without a busy signal.

Never promise a callback. Let me get a price and call you right back is functionally the same sentence as goodbye. By the time the callback happens, the caller has spoken to two other shops and one of them gave a number. A callback promise also tends to be made specifically on the calls where you do not know the price, which are exactly the calls worth the most attention. If you truly cannot price it, say what you do know, book a window contingent on tech confirmation, and keep the customer inside your process rather than releasing them back into the results page.

Never quote a range. Covered above, worth repeating because it is the most common failure of the four. Every range is heard as its bottom number and remembered as a promise.

Never raise the price after the caller has agreed to it. Whatever the reason, whatever the justification, a price that goes up between the call and the driveway is the fastest route to a chargeback, a bad review, and a customer who tells everyone. If a genuine contingency exists, disclose it during the quote segment and give both numbers. There is no acceptable version of this discovered on site.

A fifth, softer failure worth noting: an unanswered call handled by an automatic text-back is not a save. It converts a lost call into a maybe, which is better than nothing and nowhere near an answer. The honest comparison is laid out in missed-call text-back versus an AI receptionist, and the money at stake per missed call can be run through the missed call cost calculator.

Measuring adherence: you cannot fix what you cannot hear

A standard that is never checked is a poster. Within a month, everyone is back to improvising, and the shop has no idea because there is no evidence either way.

The prerequisite is that calls are reviewable. That means recorded, with the customer notified as your jurisdiction requires, retained long enough to be useful, and searchable by date and number so pulling a specific call takes a minute rather than an afternoon. Consent and notification requirements vary by state, so confirm your obligations locally before turning recording on. The practical setup, including what to listen for and how to run reviews without it feeling like surveillance, is in call recording and quality review.

Then score a small sample on a schedule. Ten calls a week is enough to see the pattern. Score each against six binary questions, because binary questions produce a usable number and opinion-based scoring does not.

Was it answered live within four rings, with the business name. Was the situation type established before anything else. Were the three capture items collected in order without the caller repeating themselves. Was a single complete all-in price given, with no range. Was a concrete window offered and the job booked on the call. Was the callback number read back and a confirmation sent.

Six yes-or-no answers on ten calls gives you sixty data points a week and immediately shows which segment is failing. In most shops the failure clusters in one place, usually the quote segment or the close, and once you can see it you can fix it with fifteen minutes of coaching rather than a general exhortation to do better on the phone.

Track the outcome number alongside it: what share of pricing-eligible inbound calls become booked jobs. That is the number the standard exists to move, and reasonable expectations for it are laid out in call booking conversion benchmarks.

How an AI receptionist executes this on the fourth simultaneous call at 3 AM

Everything above is achievable by a well-trained person on a normal Tuesday. The problem is that locksmith call volume is not distributed like a normal Tuesday. It arrives in bursts, at night, on weekends, three and four calls deep, and the standard degrades exactly when it matters most, because the person holding the phone is tired, is already on another call, or is under a vehicle with a flashlight in their teeth.

The structural advantage of an automated answer is not that it is smarter. It is that the ninetieth call of the week is handled identically to the first. There is no fourth-call degradation and no 3 AM version of the greeting. Concretely, that means the phone is answered on the first or second ring on every simultaneous line at once, so a burst never produces a busy signal or a hold. Classification happens in the same order every time. The three capture items are collected in the fixed sequence and written to the job record as they are captured, so nothing has to be repeated. The price comes from the shop's own price book with the trip charge and any after-hours surcharge already inside the number, so the range and the on-arrival increase are structurally impossible rather than merely discouraged. And the close actually happens, because booking the window and confirming the callback number are steps in the process rather than things a busy person might get to.

Two honest limits. An AI receptionist executes the standard the shop wrote; if the price book is wrong or the service boundaries were never defined, it will apply the wrong rules very consistently. And there are calls that should leave the standard entirely and go to a human immediately, which is a design decision worth making deliberately rather than discovering, covered in what an AI receptionist cannot do. The mechanics of the call layer itself are in AI call handling.

The bottom line

The first ninety seconds of an inbound locksmith call are fully standardizable and almost never standardized. Five segments, in order: a live answer with the business name inside five seconds, situation type established by twenty, the three job-defining items captured in a fixed sequence by forty-five, one complete all-in price with the trip charge and any surcharge already inside it by seventy, and a concrete window booked with the callback number read back by ninety. Everything else about a locksmith call happens after the customer has decided to trust you, and that decision is made inside this window.

The four prohibitions matter as much as the sequence. No hold, no callback promise, no range, and no price that grows after the caller has agreed to it. Each one converts a winnable call into a lost one in a documented way. Write the standard down, make calls reviewable so adherence can be scored on ten calls a week against six binary questions, and fix the segment that is actually failing rather than exhorting everyone to be better on the phone. Then solve the part that discipline cannot solve, which is that the standard degrades on the fourth simultaneous call at 3 AM precisely when the highest-value calls arrive.

Frequently asked questions

What should happen in the first 10 seconds of a locksmith call?

The call should be answered live within three or four rings with the business name, and the situation type should already be getting established. No menu, no hold, and no recorded greeting longer than a breath. Emergency locksmith callers are triaging with two other numbers queued, so a menu at the front of the call is a hang-up with extra steps.

In what order should a locksmith collect information from a caller?

Always the same three items in the same sequence: what the job is in enough detail to price it, then the location, then whether the caller still has a working key. For a vehicle that means year, make, model and key type, asked as whether they push a button or turn a key, because most callers do not know the terminology. A fixed order keeps the call fast and stops items from being forgotten on a busy night.

Why should a locksmith never quote a price range over the phone?

Because every range is heard as its bottom number and remembered as a promise. When the tech arrives with the higher figure the customer believes the price was raised on them, which produces disputes, chargebacks and reviews that specifically say you quoted one price and charged another. Give one complete all-in number with the trip charge and any after-hours surcharge already inside it.

Is it ever acceptable to put a locksmith call on hold?

No, not on an inbound first-contact call. The caller has a results page open and other numbers queued, and a hold is the point at which they dial the next one. If you cannot answer because you are already on another call, that is a coverage problem rather than a manners problem, and the fix is being able to answer simultaneous calls rather than parking one of them.

How do you measure whether a call standard is actually being followed?

Record calls, following your state notification requirements, then score ten calls a week against six binary questions: answered live with the business name, situation type established first, three capture items in order without repetition, one complete all-in price, concrete window booked on the call, and callback number read back with a confirmation sent. Binary scoring produces a usable number where opinion-based review does not, and the failures almost always cluster in one segment you can then fix with fifteen minutes of coaching.

How much does TheKeyBot cost for a locksmith shop?

TheKeyBot's Core plan is $500 per month with 500 AI minutes and 45 cents per minute overage, Pro is $750 per month with 1,000 minutes at 40 cents overage, and Elite is $1,200 per month with 2,500 minutes at 35 cents overage. Every plan includes bilingual 24/7 answering, quoting from your own price book, calendar booking, GPS-aware dispatch, payment links, and automated review requests, with no per-seat fees. Plan details are at https://www.thekeybot.com/pricing.

Sources

1. Federal Communications Commission - rules and consumer guidance covering telephone calls, recording notification, and business messaging: https://www.fcc.gov/
2. Federal Trade Commission - guidance on truthful pricing and disclosure practices for service businesses: https://www.ftc.gov/
3. Associated Locksmiths of America - professional standards and business practice guidance for locksmith operations: https://www.aloa.org/
4. Google Local Services Ads Help - documentation on lead handling, answer rate, and advertiser requirements: https://support.google.com/localservices/