Measuring HVAC missed-call recovery should begin with a defined baseline, explicit event states, and a fixed observation window. A delivered text, a reply, a qualified handoff, a booking, and a verified completion are different outcomes.

The measurement plan should be written before launch. Otherwise teams tend to count the easiest available signal and rename it "recovered revenue," even when booking and job records cannot support that conclusion.

Illustrative first-party measurement chain
  1. Eligible missed call
  2. Acknowledgement attempted
  3. Message delivered or failed
  4. Customer replied or opted out
  5. Human contact completed
  6. Booking and job outcome verified separately

Fix the denominator and observation window

Eligible missed-call denominator: all missed inbound calls recorded during one fixed observation window, minus documented exclusions. Exclusions should be decided before review and can include known spam, duplicate records for the same call, internal tests, known opt-outs, unsupported number types, and records without evidence that the missed-call trigger occurred.

Record the window’s start and end timestamps, the exclusion reason on every removed record, and changes in campaigns, hours, or seasonality. Do not silently change the denominator after seeing the result.

Define five separate rates

  • Acknowledged rate = acknowledged eligible records divided by the eligible missed-call denominator. “Acknowledged” means the approved response attempt is recorded; delivery remains a separate event.
  • Replied rate = eligible records with an inbound reply divided by the eligible missed-call denominator. A reply is not qualification or booking.
  • Qualified-handoff rate = eligible records meeting the approved routing rules and assigned to a human divided by the eligible missed-call denominator.
  • Booked rate = verified bookings in the authoritative scheduler divided by the eligible missed-call denominator.
  • Completed rate = records with the agreed completion state in the business job system divided by the eligible missed-call denominator.

The attribution boundary is strict: count a booking or completion only when the authoritative business system links it to the same inquiry identifier inside the fixed observation window. Unlinked records remain unverified; message delivery, replies, and bookings are never called revenue.

Fully synthetic cohort calculation

This illustrative test cohort contains 40 raw missed calls; 8 excluded under the prewritten rules; 32 eligible in the eligible missed-call denominator.

  • 28 acknowledged / 32 eligible = 87.5% acknowledged rate.
  • 12 replied / 32 eligible = 37.5% replied rate.
  • 8 qualified handoffs / 32 eligible = 25% qualified-handoff rate.
  • 5 verified bookings / 32 eligible = 15.6% booked rate, rounded to one decimal place.
  • 4 verified completions / 32 eligible = 12.5% completed rate.

These rates are not revenue. They are synthetic operational examples and do not predict a contractor’s results.

Keep value claims outside unsupported data

If collected-value records are incomplete or cannot be joined through the same inquiry identifier, report operational states only. Do not multiply replies or bookings by an assumed ticket value and present the result as revenue.

The measurement framework defines the site's current evidence boundaries. The synthetic evidence package demonstrates event records without claiming a live outcome.

Review the HVAC missed-call text-back service for the commercial scope this measurement plan supports.