Following up quotes and appointment requests: a workflow

A trades office can support the follow-up of quotes and the arranging of appointments with a fixed workflow. A list remembers when each quote is due for a follow-up, a daily run then creates a draft, and the office clerk sends it or lets it be. The core needs no AI. Whether and how you may follow up depends on section 7 of the German Act against Unfair Competition (UWG) and on the GDPR, and the individual case belongs with your lawyer. The workflow saves hardly any time here; its benefit would be that follow-ups happen more often.

What is meant here

A workflow is a fixed sequence of steps started by an event or a time of day. What that means in a small business is explained in What process automation in a small business is. Following up here means asking, a few days after a quote has gone out, whether any questions are open. The workflow reminds and prepares, and the decisions stay in the office. An appointment request is a wish for a site visit, a measurement or a servicing appointment.

The workflow for following up

The basis is a list, for example a spreadsheet. Each row is a quote with customer, contact, quote number, date sent, validity period, follow-up date and status. The clerk adds the row when sending the quote. That minute is new effort and appears in the calculation below.

A schedule starts the workflow on working days in the morning. According to the documentation, the n8n Schedule Trigger runs a workflow at fixed times, much like cron, a scheduling tool on servers. Two points from it matter for an office. First, the node follows the configured timezone, and the default on self-hosted instances is New York. Second, by default the node discards missed runs, for instance if the server was down that morning.

So the run does not check a waiting time but the list: has the follow-up date been reached or passed, and is the status still “open”? Then a run the next day catches up on what was missed. The alternative is the Wait node, which waits for weeks for each quote and, according to the documentation, offloads the execution data to the database while it does so. Every waiting quote is then a paused run; in my assessment, a list the office reads itself is easier to keep track of for small volumes.

For each quote that is due, the workflow creates a draft from a fixed template, with salutation, quote number and validity period. For Gmail the manufacturer lists the operation “Create a draft”, separate from sending. The clerk reads the draft, changes or discards it, sends it personally and sets the status. This matters because the workflow does not know whether the customer has accepted by telephone in the meantime. The pattern behind this is described in AI in a workflow: one AI step, fixed rules, human approval.

1. Enter the quote Office adds customer, deadline, follow-up date to the list
2. Daily run Working days in the morning: has the follow-up date arrived?
3. Create a draft Fixed template with quote number and validity period
4. Office decides Check the draft, change it, discard it or send it personally
5. Record the status Note reply, order or refusal in the list

Text version of the graphic

• 1. Enter the quote: Office adds customer, deadline, follow-up date to the list
• 2. Daily run: Working days in the morning: has the follow-up date arrived?
• 3. Create a draft: Fixed template with quote number and validity period
• 4. Office decides: Check the draft, change it, discard it or send it personally
• 5. Record the status: Note reply, order or refusal in the list

AI is only an optional extra here: a language model could adapt the template to the occasion. It is not necessary, and every reply from a customer remains untrusted text, as Sorting enquiries in the trades office shows. My suggestion for practice: at most one follow-up per quote, and the validity period stated in the quote. Under section 148 of the German Civil Code (BGB), an offer can be accepted only within a period set by the person making it.

Appointment requests

An appointment request arrives by form, e-mail or telephone note. The office offers two or three time slots. According to its documentation, the n8n Google Calendar node has the operation “Availability”, which checks whether a time slot in a calendar is free. You give the start and end of the slot, and the output says whether appointments overlap. The documentation does not describe the node finding a free slot itself. The office therefore sets the slots.

What a calendar does not know: travel, duration, who can take which job, and whether a customer should get an appointment that day. That is discretion, and why such tasks are not automated is covered in When not to automate. The confirmation is again produced as a draft and sent by hand.

The documentation describes Gmail and Google Calendar here. Which customer data may go into your e-mail and calendar service is covered in Data zones: which data may go into which tool?.

What the law says

This is a general description, not legal advice. Section 7(1) UWG declares unlawful a commercial practice “by which a market participant is unreasonably harassed” (my translation), in particular advertising where it is evident that the person addressed does not want it.

Subsection 2 lists cases where an unreasonable nuisance is “always” to be assumed. No. 1, in my translation: advertising by telephone call to a consumer without their prior express consent, or to another market participant without at least their presumed consent. No. 2: advertising using an automatic calling machine, a fax machine or electronic mail without the addressee’s prior express consent.

Subsection 3 makes an exception for e-mail when four conditions are met together: the business obtained the address from the customer in connection with the sale of a product or service, uses it for its own similar products or services, the customer has not objected, and the customer is told clearly, when the address is collected and each time it is used, that they can object at any time.

Section 7 itself does not explain the word “advertising”. Whether a query about a quote the customer has requested falls under it is therefore a question of interpretation in the individual case. Your lawyer clarifies that.

The GDPR is the second check. It does not ask how you approach someone but whether you may process their data for that purpose. Article 6(1) lists, among others, point (b): processing is necessary “for taking steps prior to entering into a contract” (“zur Durchführung vorvertraglicher Maßnahmen”) at the request of the data subject. Point (f) names legitimate interests. Recital 47 adds that processing for direct marketing purposes “may be regarded as carried out for a legitimate interest”. Article 21(2) gives the data subject the right to object at any time to processing for direct marketing, and paragraph 4 requires an explicit notice of that right at the latest at the first communication. Which basis supports your follow-up is for your data protection officer to clarify.

Organisationally, this means for the workflow: the list has an “objection” column, and the office sends the draft only when nothing is entered there.

Which figures exist and which do not

I have found no quote or conversion rate for the trades, neither in the ZDH economic reports (most recently the third quarter of 2026) nor in a search for official publications.

Worked example with invented assumptions

All values are invented and not measured. Assume 30 quotes sent a month. Today 12 of them are followed up, at six minutes each: fetching the quote from the files, checking the status, writing, sending.

• Today: 12 x 6 minutes = 72 minutes.
• With the workflow: entering 30 x 1 minute = 30 minutes. Following up 20 quotes where nothing has arrived after the deadline, at 2 minutes each to check the draft and send: 20 x 2 = 40 minutes. Total 70 minutes.

todaywith workflow

• Entering quotes in the list30 x 1 minute, assumption

  today 0 min
  with workflow 30 min

• Following up12 x 6 against 20 x 2, assumption

  today 72 min
  with workflow 40 min

• Total20 instead of 12 follow-ups

  today 72 min
  with workflow 70 min

• Total with 20 follow-ups today20 x 6 against 30 + 20 x 2

  today 120 min
  with workflow 70 min

Source: Invented example, not measured

Text version of the graphic

• Entering quotes in the list: today 0 min, with workflow 30 min (30 x 1 minute, assumption)
• Following up: today 72 min, with workflow 40 min (12 x 6 against 20 x 2, assumption)
• Total: today 72 min, with workflow 70 min (20 instead of 12 follow-ups)
• Total with 20 follow-ups today: today 120 min, with workflow 70 min (20 x 6 against 30 + 20 x 2)
• Source: Invented example, not measured

The finding is sobering: in this example the workflow saves no time, it moves it from writing to entering. The difference lies in the count: 20 instead of 12 quotes are followed up. A time gain would arise only against an office that would follow up all 20 anyway: there it would be 20 x 6 = 120 minutes against 70.

The assumptions are on the favourable side. Entering a row with seven fields in one minute is tight. If entering takes two minutes instead of one and checking and sending three instead of two, it would be 30 x 2 + 20 x 3 = 120 minutes, more than today. Time saved is therefore not the argument. The argument is that the list and the draft make following up reliable. Anyone who keeps the list by hand and looks at it every morning achieves the same without a workflow.

Whether more follow-ups bring orders is open. Thought experiment: 8 additional follow-ups x 10 per cent that still become an order x EUR 3,500 order value gives EUR 2,800 of turnover a month. The 10 per cent and the EUR 3,500 are invented. It is turnover, not profit; set-up, maintenance and tool costs are missing. How to collect your own figures is shown in Measuring time savings instead of promising them.

What you can do now

1. For four weeks, keep a list by hand with quote number, date sent, follow-up date and outcome. Afterwards you know your conversion rate and your time needed.
2. Enter a validity period in every quote and decide who follows up and when, including during holidays and sickness.
3. Ask your lawyer whether and in what form you may follow up, by e-mail and by telephone, and which notices belong in the quote and in the enquiry form.
4. Build the workflow only once the list has worked for a few weeks, and check the timezone and the sending approval.