Quotes & revenue · 9:09 video
How can AI help speed up HVAC quotes?
AI can help speed up HVAC quotes by summarizing engineers’ notes, organizing site information and flagging missing details before the office prepares a quote. Your team keeps control of pricing, scope and approval. Start with the handover that causes the most waiting, then test whether quotes move faster without creating more corrections.
English captions available in the player. Watch on YouTube ↗ · Read the transcript ↓
Find the delay between the repair and the approved quote
An engineer spots a repair, captures the details and sends them to the office. The office prepares a quote, someone follows it up, and the customer decides. In the video, Shaz maps the gaps between these steps: missing details, emails going back and forth, work waiting in a queue and follow-up pushed back. Improving that handover is a more specific starting point than giving everyone another AI tool.
Give AI the preparation; keep decisions with your team
The proposed split is practical: AI summarizes notes, organizes photos, flags missing information and drafts the next step. Coordinators and other responsible team members decide the scope, price, approval and customer response. A useful pilot starts by agreeing what a complete quote-ready handover contains and who checks it. AI should surface a missing detail rather than invent it.
Measure the whole route to approval
The video proposes four measures: time to quote, quotes that stall, quote approval rate and coordinator administration. Record a baseline before the pilot, agree a review period and compare like-for-like work. Faster drafting alone does not establish that more work was won. Include rework and errors in the review so a quicker first draft is not hiding extra work later.
Turn the example into a workshop for your business
Bring an anonymised engineer’s note, the resulting quote and the steps between them. In a practical workshop, your team can map the handover, practice preparing information with AI and agree the checks. The output is a workflow, a playbook and a pilot plan your business can own. Integration and ongoing system operation need a separately agreed scope.
Before you decide
Questions from business owners and operations teams
Can AI price HVAC repairs automatically?
This guide recommends keeping pricing and scope with your team. AI can help prepare the information, but prices, exclusions and commitments need an authorised person and reliable business data.
Will faster quotes mean more revenue?
That is a hypothesis to test. Track approval rate and the value of work won alongside turnaround time. The review described here did not establish a measured revenue increase.
Do we need a new field service system?
Start by mapping the tools and information you already use. Whether a proposed workflow can run inside them depends on their capabilities, permissions and integrations; the video does not prove compatibility with every system.
Video transcript
English caption transcript, reviewed for content and grouped by topic. Spoken wording is retained. Examples, estimates and earlier service references reflect the recording; the explanation above describes the current workshop offer.
Read the full transcript 9:09
0:00 Find the commercial moment: repair to approved quote
So, I recently worked with a HVAC company, and I want to take you through how we optimize their workflow around AI. It was between repair to quote. That was where the biggest commercial moment was. Um and let's draw that out. So, the first thing is the engineer basically identifies um a need for a repair on a specific um site. So, it's a site issue. Site issue, not sites, but there we are. Um that information then obviously has to be captured, and it has to be clearly captured so that it can move back into the office. So, that will be captured typically through notes. The office then has to turn that into a quote. The customer then obviously approves this, so there's an approval stage. And the commercial moment, or where I kind of saw the biggest value, is the gap between finding the repair and securing uh approval, which is basically this stage. This is where the value is, yeah? Um boom, just put a dollar sign. Um
1:40 Map one workflow and its handovers
so, that meant for me working with this business was to start mapping the route itself. Um so, it wasn't just thinking about AI obviously in the abstract, and that looked like this. So, obviously the way I like to do things is think about applying to one pilot, one workflow first. Um So, it's one workflow that we focused on to begin with. And that looked like this. So, you'd have the engineer. Um they're the one that finds the issue. Then you've got the office uh who are involved. They're receiving the information. Uh the quote's created. There's a follow-up. So, someone obviously has to chase, someone has to follow up on this quote. And then obviously the customer makes the decision on whether they accept. And then obviously in between each of these steps, there's basically friction.
3:01 Missing details and delayed follow-up
Okay, so between office and quote specifically, right? Lots of detail is missing. You've got things like, you know, um emails going back and forth. Maybe detail missing within the emails. Lots of waiting. Um a lot of work queue building up. And then you've also got a major friction point with the follow-up. These follow-ups are getting delayed downstream because of these two friction points here around uh information within the office and the back and forth with the quoting process. So, that overall means obviously, slower um output of cash, yeah, for this business. Slow right now.
4:03 AI prepares; people price, scope and approve
So, that kind of pushed pushed me and the team into really asking which part of this work would benefit most from AI, and what was actually kind of consuming the most amount of office capacity. So, let's talk about that a bit more. And what we figured out is the right model here really isn't just pure automation without control. It's It's actually the preparation pieces, doing that quicker, the mechanical stuff, with obviously human judgment in place. So, the way that we ended up uh proposing AI was AI preparing, and then the team pricing and approval approving. So, how that looks like is we would have AI basically on this side, and AI would be doing the things like note summaries, so summariz- summarizing, um it will be um sorting out assets, like photos, for example, that are being um sent back and forth of the site, um flagging maybe missing details. Uh and then maybe also drafting what the next step will be. Then that piece here is basically the AI preparing, and then we would have the team, the coordinators, deciding. So, people on this side, yeah? And they would do the stuff that's important. Pricing, scoping, approving, customer relationship, all of that bit, which is important. So, this was the operating model that we suggested as it was much clearer on the delineation between what AI would do and what people would do.
6:26 Choose the pilot measures
Um and the way that we wanted to measure this when we rolled rolled this out is through the following kind of measurements that are typical for, say, a COO or an MD uh within uh within a trades business. So, where do I do that? So, let's put it here. So, we're going to put measurement here, number four. And the measures we looked at were is time to quote time to quote reducing? Um is the number of quotes being sent um that are stalling reducing? So, quote stalls reducing, as well. Um is the approval rate So, I don't know where I'm going to put this. Let's put that here. Is the approval rate of quotes up? And then, is the admin for the coordinators down, All So, these were the pilot measures within this workflow that we proposed. And obviously, these metrics are commercially driven and operationally driven. So, ultimately, this is this is kind of the redesign of their workflow. And this is why working in this way, looking at a specific route and a specific workflow, will help in terms of AI adoption because it's much clearer where the commercial value is being driven. Whereas, generic AI adoption, so generic tool access, for example, may give you some marginal improvements, uh but that won't necessarily drive or improve this bit. The value.
8:44 Start with one route and a proper pilot
So, my practical rule always is kind of start with one route and then run a proper pilot. Now, I appreciate that was a lot to think about. If you would like to talk about this further, please just send me a message and I can give you more information or visit customerjourneys.ai and you can fill in your information. I'll get back to you on the same day. Thank you.
