Skip to main content

Quote follow-up · Australia

How do you automate quote follow-up without annoying customers?

The safest useful workflow is not “send three reminders.” It knows why the customer is in the sequence, what happened to the quote, when a person should step in and exactly what stops the next message.
By Ajay DabhiPublished 28 July 2026Updated 28 July 2026

01 / DECISION

Map every exit before writing the first message.

  1. 01Choose the system that owns quote status and customer consent.
  2. 02Define sent, viewed, replied, accepted, declined, expired and stopped states.
  3. 03Route questions, complaints and scope changes to a named person.
  4. 04Measure accepted and completed work without claiming every response was caused by automation.

02 / MAP

What information should start the workflow?

Start only when the quoting system confirms that a real quote was sent to the intended customer and the contact path is permitted for that purpose.
The workflow needs a quote identifier, customer contact, service, owner, send time, current status and next permitted action. If staff can send quotes outside the system, decide how those records enter the same queue or remain deliberately manual.
This is different from a missed-call workflow. A caller may still need qualification, while a quoted customer already has scope and price context. Use the focused missed-call calculator when unanswered demand is the earlier leak.
  • Require one stable quote ID so duplicate records cannot start parallel sequences.
  • Store the staff owner who will receive replies or exceptions.
  • Keep the service and quoted scope available for a human, not exposed unnecessarily in every message.
  • Record the source and lawful communication basis appropriate to the business.

03 / STOP

Which events must stop the sequence?

The automation must stop on a decision, a reply, an opt-out, an error or a manual hold. Continuing after any of those events is both poor service and a control failure.
Accepted and declined quotes are obvious exits. Less obvious stops include a customer asking a technical question, changing scope, disputing price, reporting a complaint or saying they already booked through another channel. The workflow should pause and show the full conversation to the responsible person.
An AI classifier can help route common replies, but it should have a low-confidence path. When intent is unclear, “send another reminder” is the wrong default. The broader AI automation service starts with these operating rules before choosing a tool.
  • Accepted, booked or paid according to the business's real process.
  • Declined, cancelled, expired or outside scope.
  • Any reply that needs pricing, technical judgment or complaint handling.
  • Unsubscribe, invalid destination, delivery failure or staff-applied hold.

04 / WRITE

What should the follow-up message say?

Identify the business, refer to the quote without exposing unnecessary detail, give one useful next action and make it easy for the customer to reply or stop.
The tone should match the conversation that already happened. A customer waiting on an urgent repair needs a different message from someone comparing a larger planned project. Avoid manufactured deadlines, vague “just checking in” copy and promises the operating team cannot keep.
Australia’s ACMA says commercial electronic messages must meet consent, sender identification and unsubscribe requirements. It also states that a marketing message cannot be sent merely to ask for consent. OAIC guidance separately explains when APP 7 applies to use or disclosure of personal information for direct marketing. This guide is operational information, not legal advice.
  • Say who the message is from and why the customer is receiving it.
  • Use one action: ask a question, accept, request a call or decline.
  • Do not include sensitive service details that are unnecessary for the reminder.
  • Make the opt-out path functional and propagate it to every connected tool.

05 / MEASURE

How do you know the workflow is useful?

Reconcile quote states before and after the workflow, then inspect response quality, accepted value, cancellations and complaints. Message delivery alone is not the outcome.
Track how many eligible quotes entered, how many stopped for each reason, how many required human help and how many reached the business’s accepted and completed definitions. Keep a control period or comparison group where volume permits, and label small samples honestly.
Do not assign every later booking to automation. Customers may have accepted without it, received a phone call, or returned through another channel. The useful report shows the workflow’s touched records and the final outcome without overstating causation.
  • Eligible quotes and duplicate prevention failures.
  • Replies by reason, including questions and complaints.
  • Accepted, declined, expired and unknown outcomes.
  • Unsubscribe requests, delivery failures and incorrect sends.

LABELLED WORKFLOW DEMO

A quote state machine, not a timer chain.

This is a process model. Timing, messages and legal controls still need to be set for the actual service and systems.

01

Quote sent

A valid quote ID, owner, customer and permitted channel enter the workflow.

TRIGGER
02

Follow-up due

Re-read the live status before sending anything.

CHECK
03

Customer responds

Accept, decline, answer or send uncertain intent to a person.

ROUTE
04

Decision recorded

End every connected sequence and preserve the outcome.

STOP

This is not a ready-made message sequence or legal determination. The business must validate consent, message purpose, privacy handling, unsubscribe behaviour and staff ownership for its circumstances.

SOURCES

What this guide relies on

Official sources establish the platform or Australian compliance facts. The operating method is my practical interpretation, not a promise of a particular result.

ACMA: avoid sending spam

Explains the Australian rules ACMA administers for commercial electronic messages, including consent, sender identification and unsubscribe controls.

OAIC: direct marketing

Explains how APP 7 and related Australian direct-marketing obligations can apply to use or disclosure of personal information.

QUESTIONS

Questions worth answering before you start.

How soon should a quote follow-up be sent?
There is no universal delay. Urgent repair work, planned projects and professional-service proposals have different decision cycles. Set the timing from how the service is sold, what the customer was told and how quickly the team can respond to a reply.
How many automated follow-ups should I send?
Use the fewest contacts needed to support the decision and respect the person's expectations. A fixed sequence copied from another business may be too slow, too aggressive or legally unsuitable. Every workflow needs a final stop condition.
Can I send quote follow-ups by SMS in Australia?
The content and circumstances matter. Commercial electronic messages can be subject to Australia's Spam Act, including consent, sender identification and unsubscribe requirements. Get advice for your situation and build those controls into the workflow rather than treating compliance as copy at the end.
Should AI write every follow-up reply?
No. AI can classify routine responses or prepare a draft, but price disputes, complaints, unusual scope, sensitive information and uncertain intent should route to a person. The business remains responsible for the communication.
What should stop the automation?
At minimum: accepted quote, declined quote, customer reply requiring a person, unsubscribe request, invalid contact, complaint, job already booked or a manual stop by staff. The exact list should match the quoting and CRM system.

YOUR NEXT MOVE

Show me how a quote moves today.

I will map the triggers, stop rules, handoffs and records first, then tell you whether automation is useful and where a person must stay in control.