A diagnostic is simply a structured way to find out where work leaks out of your business and what each leak costs. You can do a useful version yourself with exports from your phone system, job software and accounting, a spreadsheet and two or three quiet weekends. This checklist walks through seven areas, gives you a formula for each and ends with a way to rank what you find.
Before you start: gather the numbers
Pull twelve months of data if you can, so both heating and cooling seasons are included. You'll need:
- inbound calls by month, answered and missed, from your phone system
- jobs booked and invoiced by type (service, maintenance, replacement, new install) from your job software
- quotes issued and won for replacements and installs
- marketing spend by channel, plus agency and software fees
- active maintenance agreements and renewals
- technician hours paid and hours billed
- monthly revenue, gross margin and cash balance from your accounts
Keep a list of anything you can't find. Missing data is a finding in its own right.
The five core numbers
Five numbers explain most of an HVAC company's results. Write them on one page before you go further.
| Number | Formula | What it tells you |
|---|---|---|
| Answer rate | answered calls ÷ total inbound calls | Whether demand is reaching you |
| Booking rate | jobs booked ÷ answered calls | Whether your office converts interest into appointments |
| Average ticket | service revenue ÷ service jobs | Whether pricing and diagnosis are working |
| Close rate | replacement quotes won ÷ quotes issued | Whether your sales process works |
| Utilisation | hours billed ÷ hours paid | Whether you have room to grow or need to hire |
Area 1: demand
Ask where calls come from and what each source costs per booked job, not per lead. Divide each channel's monthly cost by the jobs it produced. If you can't connect calls to sources, that is your first fix: call tracking or even a "how did you hear about us" field filled in on every booking.
Check also how demand moves through the year. Many companies are full in January and July and nearly empty in April and October. Those shoulder months are where maintenance campaigns and replacement follow-ups pay for themselves.
Area 2: call response
Count missed calls by hour of the day. Then put a cost on them:
missed calls per month × booking rate × average ticket = monthly revenue at risk
A worked example with round numbers: 50 missed calls a month × 55% booking rate × $400 average ticket ≈ $11,000 a month. Not every missed caller is lost, since some call back, so treat this as an upper bound and say so.
Common causes are a single office person at lunch, morning rushes before 9 a.m. and after-hours calls going to a full voicemail box.
Area 3: quoting and follow-up
For each replacement quote, record the date issued, the date of first follow-up and the outcome. Then check:
- how many quotes had no follow-up at all
- the median days between quote and first follow-up
- close rate for quotes followed up within 48 hours compared with later
If quotes followed up quickly close noticeably better, the fix is a follow-up rule, not more leads.
Area 4: pricing and mix
Compare average ticket by technician and by call type. Wide differences between technicians doing similar work usually point to inconsistent diagnosis or pricing, which is a training or price-book question. Look at your revenue mix too: what share comes from replacements, repairs and maintenance agreements? A company that depends heavily on emergency calls has a less predictable year than one with a strong agreement base.
Area 5: maintenance agreements
Agreements smooth out the year and feed replacement work. Check:
- active agreements now compared with twelve months ago
- renewal rate, and whether renewals are asked for or simply hoped for
- whether agreement visits are scheduled into shoulder months
If 310 agreements renew at 60% instead of 80%, that is 62 fewer customers a year who receive a scheduled visit, and a smaller pool for replacement conversations.
Area 6: capacity
Utilisation tells you whether to sell more or hire. Below about two-thirds billed, you probably have capacity and a demand or booking problem. Near full utilisation in peak season with long waits for customers, you have a hiring or scheduling problem. Check also how long the office takes to schedule a booked job, and whether dispatch decisions are written down or held in one person's head.
Area 7: cash
Calculate cash on hand in weeks of payroll at the slowest point of last year. Note how long invoices take to be paid and how many jobs are invoiced more than a few days after completion. Growth plans that ignore cash usually stall in the first slow month.
Ranking what you find
Put every finding in one table with four columns: what you found, the estimated yearly cost, how hard it is to fix (easy, medium, hard) and who would own the fix. Then sort by cost. Start with the most expensive problem that is easy or medium to fix, not the one that annoys you most.
| Finding | Yearly cost estimate | Difficulty | Owner |
|---|---|---|---|
| [Finding] | [amount] | [easy/medium/hard] | [name] |
Mistakes to avoid
- Mixing seasons. Comparing July with March tells you about the weather, not your business.
- Counting leads instead of jobs. Always anchor on booked and invoiced work.
- Skipping the uncomfortable numbers. Cash and utilisation are the ones owners avoid most, and often matter most.
- Fixing everything at once. Take the top three findings into a plan, using the method in our 90-day plan guide.
When a self-review is enough
A self-review is often enough for a company with clean data and a team that agrees on the problem. If the numbers conflict, if staff accounts contradict each other or if a major decision rests on the answer, our Business Diagnostic does the same work from outside, with staff interviews and reconciliation across systems. See the packages overview for how it compares with the other options. This checklist covers the business only; technical and code questions belong to your licensed team and your provincial regulator.
Frequently asked questions
How much data do I need to start?
Twelve months of job history is ideal, because it covers both heating and cooling seasons. Six months is workable if you note which season it covers. Anything less, and seasonal swings will mislead you.
What if my numbers don't agree with each other?
That is normal and it is itself a finding. Agency reports, phone systems and job software often count differently. Decide which source is closest to real money, usually your invoices, and use that as the anchor.
How often should I repeat the checklist?
Once a year in full, ideally in a shoulder season, plus a quick look at the five core numbers every month.
When should I pay for an outside diagnostic instead?
When the numbers contradict each other and you can't reconcile them, when staff explanations conflict, or when a large decision depends on the answer. An outside reviewer has no stake in who was right.