Automation is much easier to judge with a concrete example than in the abstract. So this post follows a single enquiry through a service business, from the first message to the quotation to the follow-up, and shows each step as it is often done by hand today and as it could be done with a sensible amount of automation.
Everything here is fictional: the business, the customer and the details. The shape of it will be familiar to most service businesses, whether they renovate kitchens, service air-conditioners or run a cleaning team.
The enquiry
A small renovation company receives a WhatsApp message on a Tuesday evening. The customer wants new kitchen cabinets for a four-room flat, mentions they would like work to start next month, and attaches three photos of the current kitchen. There is no measurement and no budget.
Step 1: capturing it
By hand: whoever sees the message first copies the details into a spreadsheet or a notebook, if they have time. The photos stay in the chat. If the message arrives late at night, it may not be seen until the next morning, by which point it is under twenty newer messages.
Automated: the message and its photos are logged against the customer as soon as they arrive, in one place the whole team can see. Nobody has to remember to copy anything.
Step 2: asking for what is missing
By hand: someone types out the usual questions: can you share the kitchen dimensions, do you have a budget in mind, which floor is the flat on. The same questions, typed slightly differently, many times a week.
Automated: the gaps in the enquiry are identified against what a quote needs, and a short reply asking for them is drafted. Depending on how the business prefers to work, it can be sent automatically or queued for someone to glance at and send.
Step 3: drafting the quote
By hand: once the details arrive, the owner opens last month's similar quote, changes the numbers, and hopes nothing from the old job is left in it.
Automated: a draft is prepared from the business's own price list and the details provided, with anything unusual flagged. This is where most of the saved time sits, and it is also where accuracy improves, because nothing is copied from an old document.
Step 4: reviewing and sending
This step does not change, and should not. A person reads the draft, adjusts it using judgement about this job and this customer, and approves sending it. Pricing is a business decision, and an automated system that sends quotes without review is a system that will one day send a wrong one.
We wrote about where to put this kind of check in our post on approval gates.
Step 5: following up
By hand: the intention is to chase in a week. In practice, it happens when someone remembers, which is sometimes never. Quotes that are not followed up are a common and invisible way service businesses lose work.
Automated: a follow-up is scheduled when the quote is sent, and cancelled automatically if the customer replies first. The message itself can be reviewed in advance, so it sounds like the business rather than a robot.
Step 6: when the customer replies
Again, this stays with a person. The reply might be a question, a negotiation or a yes. Each needs judgement and a human voice. The automation's job is simply to make sure the reply is seen promptly, with the whole history of the enquiry next to it.
What this example shows
Of the six steps, four are preparation and two are decisions. The preparation is where the time goes, and it is what software is good at. The decisions are where the business's judgement lives, and they stay with people.
That split is the right starting point for almost any workflow. Our Automation Sprint takes one repeating workflow like this one, connects it across the tools you already use, and keeps approval gates where judgement is needed. If your enquiries look anything like this, it is a good one to start with.
