A8gent
Build guide - ServiceM8

ServiceM8 quote follow-up: when the add-on is enough

ServiceM8's Automation add-on sends up to five quote follow-ups on every plan. Here is how to set it up well, and the four gaps worth a custom layer.

A8gent

What you get from this guide

  • What the Automation add-on does: up to five follow-ups by email, SMS or both, on every plan
  • The documented rules that cancel a sequence, so you do not rebuild them
  • A straight answer on whether you need anything custom
  • The four gaps a small layer on the API can close
  • API key auth, job webhooks and rate limits, from ServiceM8's developer docs

What ServiceM8 actually ships

ServiceM8's built-in quote follow-up is stronger than most field service tools offer. The help centre article on Quote Follow Up Automation sets it out. With the Automation add-on active, ServiceM8 emails or texts clients about a quote you have sent. You decide how many times to follow up, to a maximum of five. For each one you choose email, SMS or both, and when it sends. Each has its own template.

The same article lists what cancels the scheduled messages: the customer replying, until you respond to or dismiss the reply; the customer accepting the quote online, which moves the job to Work Order; the job status being changed by hand to Work Order, Completed or Unsuccessful; a team member cancelling it from the job diary; and resending the quote to someone who is not a saved contact on the job.

It can also tidy up after itself. You can have jobs with unaccepted quotes move to Unsuccessful after a set time, and the article notes that a job goes back to Quote status if the customer replies afterwards. The add-on is documented as available on all ServiceM8 plans, with each SMS consuming a text credit.

The honest verdict

If you run ServiceM8 and your quotes are not being followed up, the fix is a settings screen, not a project. Five touches, two channels, stop on reply, stop on acceptance, on every plan. A custom sequence that duplicates this would cost money to build, need maintaining, and would know less about the quote than ServiceM8 does.

Do this first

Turn the add-on on and run it for a month before you consider anything else. Then look at the quotes that still went cold and ask why. Only build for a reason you can see in that list.

That is not what most pages on this topic say, because most are written by someone selling a build. We sell builds too. This one is usually not needed.

Setting the native sequence up well

Activate Automation under Settings, ServiceM8 Add-ons, then open Settings, Automation and edit Quote Follow Up. A sequence that uses all five slots sensibly looks like this.

A five-touch sequence inside the ServiceM8 add-on
Follow-upTiming after quote sentChannelJob of the message
13 hoursSMSConfirm the quote arrived and invite questions
23 daysEmailAdd something useful: availability, a scope note
37 daysSMSAsk directly whether they want to go ahead
414 daysEmailRemind them when the price expires
521 daysEmailClose out and leave the door open
Then28 daysStatusAuto-update to Unsuccessful

The first two timings mirror ServiceM8's own example. Three settings details matter. Follow-ups go to the contacts the quote was sent to, and the email or phone number has to match a saved contact on the job, so send quotes to saved contacts. Scheduled messages show in the job diary, which is where your team checks what is pending. And SMS costs credits, so use text where it works hardest, at the start and at the direct question.

Follow-up 3 - SMS, day 7

Hi {first_name}, checking in on the quote we sent for
{job_description}. Are you looking to go ahead, or has
something changed? Happy to adjust the scope or talk
it through. - {business}

Use the merge fields ServiceM8 offers in the template editor rather than the placeholder names above. The full copy set for each touch is in the general quote follow-up guide.

The four gaps worth building for

None of these is a criticism of the add-on. They are things its documentation does not describe, and that some businesses want.

What a small custom layer adds
GapWhy it mattersWhat the layer does
Large quotes get the same treatmentA big job deserves a phone call, not only templatesCreates a call task for the estimator when a quote over your threshold is sent
No won and lost report by sequenceYou cannot tune what you do not measureLogs sent, won and lost with dates to a sheet
Nothing after UnsuccessfulSome lost quotes come back months laterOne check-in at 90 days for selected job types
Replies wait for a personEvening replies sit until morningAlerts the right person by phone with the reply

An illustrative sum, with figures invented for the arithmetic: if you send 60 quotes a month, 5 of them are over 5,000, and a next-day call wins 1 more of those each month, that is 5,000 of extra work a month from one small rule. If your quotes are all similar in size, this gap does not apply to you.

What the API gives you

ServiceM8's authentication page describes two kinds of application. A private application connects to your own account with an API key, created under Settings, API Keys and sent in the X-API-Key header. No developer account is needed. A public application uses OAuth 2.0 and is the only kind that can use the Messaging API to send email and SMS through ServiceM8.

The Webhooks API overview lists nine objects that support webhooks, including Job, Job Activity, Job Payment, Note and Task. You subscribe at https://api.servicem8.com/webhook_subscriptions. The payload does not contain the record. It gives the object, the UUID, the list of changed_fields, a UTC timestamp and a resource_url to fetch. Your endpoint should return 200. A 410 unsubscribes the webhook, a 429 throttles it for 15 minutes, and other errors are retried for up to 72 hours before the webhook is cancelled.

The Jobs reference gives the fields this layer reads: status, with the valid values Quote, Work Order, Unsuccessful and Completed; quote_sent and quote_sent_stamp; and unsuccessful_date. The REST overview sets the limits at 180 requests per minute and 20,000 per day, per application and per account.

The build: a layer on top

The design rule is simple. ServiceM8 sends the messages. The layer watches, routes and counts. Written for n8n, and it maps onto Make or Zapier.

1

Create the API key

Settings, API Keys. Store the key in your automation platform's credentials. Test it with a request to the company endpoint before going further.

2

Subscribe to job changes

Register a subscription for the job object, limited to the fields you care about, pointing at a Webhook node. The example response in ServiceM8's documentation shows the shape of a subscription.

A job subscription, as ServiceM8 returns it

{
  "object": "job",
  "callback_url": "https://YOUR-N8N/webhook/servicem8-job",
  "fields": ["status", "quote_sent"],
  "active": true
}
3

Fetch the job on each webhook

Return 200 at once. Then GET the resource_url with your API key and read status, quote_sent and quote_sent_stamp. Remember the webhook timestamp is UTC while most ServiceM8 timestamps are in the account's local timezone.

4

Log the quote as sent

When quote_sent changes and the status is Quote, upsert a row by job UUID: quote date, the job total (the Jobs reference exposes a read-only total_invoice_amount), job category and outcome = open.

5

Route the large ones

If the job total is over your threshold, create a task for the estimator due the next working day, with a link that opens the job. Leave the native follow-ups running. The call is in addition to them.

6

Record the outcome

When status becomes Work Order, set the outcome to won. When it becomes Unsuccessful, set it to lost with the date. That one sheet gives you a win rate by month and by job category.

7

Optional: the 90-day check-in

A weekly sweep finds lost quotes of selected types that reached 90 days and have no newer job for the same client, then sends one email. One only. Anything more is a marketing campaign and should be run as one.

Testing and measuring it

Test the native sequence and the layer separately, in that order. Most problems turn out to be in the settings, not the code.

  1. Create a client with your own email and mobile as a saved contact. Create a job, send yourself a quote, and check the job diary shows the scheduled follow-ups.
  2. Reply to the first follow-up. Confirm the remaining messages are cancelled, then respond or dismiss the reply in ServiceM8 and confirm the automation resumes, as the documentation describes.
  3. Accept a second test quote online and confirm the job moves to Work Order and the follow-ups stop.
  4. Send a quote to an address that is not a saved contact and confirm for yourself that follow-ups do not send.
  5. With the layer running, send a quote above your threshold and confirm a call task appears once, with a working link to the job.
  6. Change a test job to Unsuccessful and confirm the log records it as lost with the date.
  7. Stop your endpoint briefly, change a job, restart, and confirm ServiceM8's retry delivers the change.

Then measure. The log gives you three figures a month: quotes sent, quotes won and quotes lost. Split them by job category and by size band. Two patterns are worth looking for. If small quotes win at a healthy rate and large ones do not, the call task is doing too little or the calls are not being made. If one job category loses far more than the others, the issue is pricing or scope for that category, and no follow-up fixes it.

Look also at when wins happen. Count how many accepted quotes were accepted after the third follow-up. If the answer is close to none, you can cut follow-ups four and five and save the text credits. If a meaningful share arrive late, the full five is earning its place. This is the kind of question the native add-on cannot answer by itself and the small log can.

Check the subscription monthly. ServiceM8's webhook documentation describes a status query parameter on the subscriptions endpoint that returns inactive subscriptions along with the last failure reason. A two-minute look tells you whether the layer is still receiving changes.

Where these builds go wrong

  • Building a custom sequence that duplicates the add-on, then double-messaging customers.
  • Expecting field values in the webhook payload. It only lists what changed.
  • Planning to send SMS through ServiceM8 from a private API key integration.
  • Sending quotes to an email address that is not a saved contact, so the native follow-ups cannot send.
  • Leaving customer replies neither answered nor dismissed, which holds the native sequence in a cancelled state.
  • Comparing a UTC webhook timestamp with local-time job fields.
  • Letting the endpoint fail for days, after which ServiceM8 cancels the webhook.
  • Texting all five touches and being surprised by the credit use.

If you run a different platform, the Jobber quote follow-up guide and the unsold estimate follow-up guide cover the fuller build. Sector context is on the HVAC and plumbing and cleaning companies pages, and what an automation build costs covers pricing.

Sources checked

FAQ

Does ServiceM8 follow up quotes automatically?
Yes. ServiceM8's help centre documents a Quote Follow Up automation inside the Automation add-on. You choose how many times to follow up, with a maximum of five, set whether each one is an email, an SMS or both, and set when each one sends. The article gives the example of an SMS three hours after the quote and an email three days after.
Which ServiceM8 plan includes quote follow-up?
The help article states that the Automation add-on is available on all ServiceM8 plans. It also notes that each SMS sent consumes a text credit under your plan, so a five-touch sequence that texts every time has a running cost. Check your current plan's text allowance before choosing SMS for every touch.
Does the native sequence stop when the customer replies?
Yes, according to the documentation. Scheduled follow-ups are cancelled when the customer replies and you have not yet responded to or dismissed the reply, when the customer accepts the quote online, when the job status is changed to Work Order, Completed or Unsuccessful, when a team member cancels the automation from the job diary, or when you resend the quote to someone who is not a saved contact on the job. Responding to or dismissing the reply resumes the automation.
So do I need a custom automation at all?
Most ServiceM8 businesses do not. Five touches by email and SMS with documented stop rules, on every plan, covers the core job. A custom layer is worth considering for four things the add-on does not describe: treating large quotes differently, prompting a phone call, a won and lost report, and contact after the quote has been marked Unsuccessful.
Does ServiceM8 have an API and webhooks?
Yes. ServiceM8 documents a REST API and a Webhooks API. Private applications authenticate with an API key in the X-API-Key header, generated under Settings, API Keys, and no developer account is needed for that. Webhooks are available for nine objects including Job, and they tell you which fields changed rather than sending the values, so you fetch the record afterwards.
What are the ServiceM8 API rate limits?
The REST API overview gives limits of 180 requests per minute and 20,000 requests per day, applied per application and per account, with an HTTP 429 response when either is reached. A quote follow-up layer makes a handful of calls per quote, so the limits are not a practical concern.
Can my automation send SMS through ServiceM8?
Only if it is a public application. ServiceM8 states that its Messaging API is available to public applications using OAuth 2.0, not to private API key integrations. A private layer either lets the native add-on do the sending, which is the simplest design, or sends through your own email and messaging provider.
What does a layer like this cost to build?
Less than a full sequence, because ServiceM8 already does the sending and the stopping. The layer is a webhook, a lookup, a task and a log. We quote a fixed price in writing per scoped workflow after a free consult, and we will tell you if the native add-on is all you need.

Done for you

Not sure whether ServiceM8's own follow-up is enough for you?

We will look at your quotes and tell you straight. If the add-on covers it, we will help you set it up well. If a layer for large quotes and win-rate reporting would pay, we build that. The audit is free.

The consult is free, and you keep whatever we build.

Keep building

The rest of the cluster