Home / Blog / Jobber AI Receptionist review: when a text-back is enough

Jobber AI Receptionist review: when a text-back is enough

What Jobber's AI Receptionist costs, what owners in Jobber's community say, and how it compares with a missed-call text-back we tested against mocks.

If you already run your shop on Jobber, the AI Receptionist is the obvious answer to missed calls. It's one toggle away, it lives inside the tool your office already uses, and it can book work straight onto the calendar. The question most owners ask in Jobber's community isn't whether it works. It's whether their customers will put up with it, and whether a phone answered by AI is worth more than a text that says "sorry we missed you, what's going on?"

This post compares the two honestly. We haven't run Jobber's Receptionist on a live account, so everything about it below comes from Jobber's own help pages and from owners posting in Jobber's community. The text-back side comes from our missed-call text-back recipe for Jobber shops, which we did build and test, against mock services rather than a live phone line.

What Jobber's Receptionist does

According to Jobber's help article on Receptionist, it answers calls (with an AI voice) and texts, and it can answer questions about your business, take a message as a follow-up task, create a request or book a job. You choose which of those it's allowed to do on its own. A few details that matter in practice:

  • It's included with unlimited use on the Plus plan. On select other plans it's a $29-a-month add-on, then $0.79 per conversation past the included allowance. Jobber's own pages disagree on whether that allowance is 30 or 100 conversations (more on that below). Calls tagged "spam" or "no conversation" don't count.
  • A text conversation stays open for 24 hours, so a back-and-forth over one afternoon counts once.
  • When you hit the allowance, you choose between paying overages and pausing the Receptionist until the next billing cycle.
  • You need a dedicated Jobber phone number, even if you keep your existing number in front of it.
  • An optional "answering delay" lets the line ring for 10 seconds (about two rings) before the Receptionist picks up, so you or the office can grab the call first.
  • An escalation policy lets you list words like "emergency" or "urgent" that send the conversation to a team member.
  • If a caller hangs up before the conversation starts, an "SMS send back" option texts them, at most once per 24 hours and never to someone who replied STOP.

One note on the numbers, because we couldn't confirm them. When we checked on October 28, 2026, Jobber's Receptionist help article (updated July 9, 2026) said the $29 add-on includes 30 conversations, including in the section on how it shows up on your invoice. Jobber's add-ons help article (updated August 13, 2026) said it includes 100 conversations per month. We can't tell which is current or whether it depends on your account, so ask Jobber or check your own invoice before you budget. The cost example below shows both.

What owners in Jobber's community report

The most useful reading on this is Jobber's own community, because most of the people posting are Jobber users (one reply below comes from someone tied to a competing product). Two threads stand out.

In a thread from a firewood business weighing the Receptionist, a community coach who uses it for off-hours calls describes it as "basically a better voicemail" and says that with the AI disclosure on, "hardly anyone stays in the call longer than hearing it's AI." Another community coach in the same thread says it offered rescheduling on days they had no availability. A later reply suggests letting the phone ring the team first and only falling back to the AI when nobody can pick up.

A second thread about rural customers lands in the same place: use it for overflow and after-hours, say up front that it's automated, and give callers an easy way to ask for a person. That reply comes from someone who says they're involved with a competing AI receptionist product, so read it with that in mind.

None of this is a survey. It's a handful of owners, and their customers may not look like yours. What it does show is a pattern: the Receptionist earns its keep after hours and on overflow, and the first few seconds of the greeting decide whether people stay on the line.

The other option: a text-back that doesn't pretend to be anyone

A missed-call text-back takes a different approach. Nobody answers the phone. When a call goes unanswered, the caller gets a text from your business number within a few seconds, a short back-and-forth collects what's wrong and where, and the result lands in Jobber as a client and a request. Emergencies get flagged to whoever is on call. We explain the basic idea in why a missed-call text-back matters for HVAC shops, and the build itself with n8n and Twilio goes node by node.

Our version is a 17-node n8n workflow. Read top to bottom, the nodes are (some steps pair two nodes):

  1. Twilio call status webhook
  2. Missed? Build text-back, then send the text-back
  3. Twilio inbound SMS, then update the conversation and build the request for the language model
  4. LLM: extract intake fields (what's wrong, address, urgency)
  5. Merge the fields and decide the next step
  6. Messages to send now, then send the reply
  7. Opt-out request? If so, send an opt-out confirmation
  8. Ready to book? Build clientCreate, then Jobber clientCreate
  9. Build requestCreate, then Jobber requestCreate
  10. Confirm to the customer, page on-call if needed, and send the confirmation

We don't have screenshots of this workflow running yet. The list above is taken straight from the workflow file.

How we tested the text-back (not the Receptionist)

Our testing notes describe the setup for every recipe. We imported the workflow into n8n 2.45.4 and ran it end to end against a local mock server that stands in for Twilio and Jobber, built from their official API docs. The language model was a small local model (Qwen2.5-3B-Instruct, quantized, running in Ollama), not a cloud service. No real phone, customer or Jobber account was involved. The last run passed all 11 scenarios:

Test What it checks
S1 An answered call is ignored
S2 A missed call during business hours gets a call-back text
S3 An after-hours "no cool" call becomes a Jobber request over two texts
S4 A burst pipe at 2am on a Saturday pages the on-call tech
S5 A gas smell gets safety instructions and no booking
S6 The same caller twice within 12 hours gets only one text
S7 A blocked or anonymous caller is skipped
S8 A text after the request is logged goes to the office
S9 A plain-English opt-out is honored
S10 "Won't stop" about a leak is not treated as an opt-out
S11 The carrier keyword STOP

Some of those tests exist because something broke. Our first full regression run failed every language-model scenario, and the cause turned out to be boring: the local model server had died. We restarted it and reran the suite, which passed. S9 to S11 were added after we realized people don't opt out with keywords. They write "please stop texting me," and they also write "the water won't stop," which must not unsubscribe them. We also pinned the Jobber API version the workflow calls, so a Jobber API change doesn't silently alter its behavior.

What mocks can't tell you: how real callers react, how carriers filter your texts, and how a live Jobber account behaves. Treat our results as "the logic does what we say," not "it will book more jobs."

Side by side

Jobber AI Receptionist Missed-call text-back recipe
Who answers An AI voice on the call Nobody; a text arrives after the call is missed
Disclosure question Jobber encourages telling callers it's AI but doesn't require it The text comes from your number and says who you are
Cost Included on Plus; otherwise $29/mo for 30 or 100 conversations (Jobber's pages differ), then $0.79 each Twilio number and texts plus a model; the recipe page has the itemized estimate
Books into Jobber Yes, if you allow it Creates a client and a request for the office to schedule
Emergencies Escalation keywords you choose Rule-based on-call paging, tested with burst-pipe and gas-smell cases
Setup Inside Jobber n8n, Twilio and A2P 10DLC registration
What you can inspect Transcripts and recordings in Jobber Every rule is in the workflow file

To make the cost row concrete, here's an illustrative example with made-up numbers. If the add-on handled 60 conversations in a month, the bill would be $29 + (30 × $0.79) = $52.70 with a 30-conversation allowance, or just $29 with a 100-conversation allowance. The recipe's cost breakdown puts the text-back for a similar month at roughly $7.50 to $17 plus your n8n server, before the one-time A2P registration fees. Neither number is a measurement from a real shop.

Which one fits your shop

The Receptionist is probably the better fit if you're already on Jobber's Plus plan (it's included), and if most of your calls are straightforward bookings. One community coach in the first thread said it would "probably be amazing" for a shop with straightforward pricing and a built-out price book. Several commenters in the threads above say it's at its best on calls nobody would have answered anyway.

A text-back is probably the better fit if your customers hang up on AI voices, if you want to keep every rule in a file you can read, or if you're not on Jobber Plus and don't want a per-conversation meter running. It's also easier to explain: a text from your number is a familiar thing.

They also aren't mutually exclusive. Some shops let the phone ring the team, send overflow to the Receptionist, and use the text-back only when someone hangs up. If you go that route, watch for double-texting, since Jobber's own SMS send back and a separate text-back could both fire for the same hang-up.

Before you turn either one on

Jobber's help page says it encourages you to tell callers the phone is answered by AI or that calls may be recorded, "but this isn't required," and tells you to check local and state rules about recording calls. Recording-consent rules vary by state, so that check is worth doing. Texting has its own rules too. Our post on whether a missed-call text-back is legal covers TCPA consent, opt-outs and A2P 10DLC registration. It's general information, not legal advice.

Whichever you pick, measure it. Our call scorecard recipe is one way to see which missed calls turned into booked work. If you're still working out whether missed calls are costing you enough to bother, start with what missed calls cost an HVAC or plumbing shop.

Found an error or a better source? Corrections policy. More articles on the blog.