Case study›Solar · Bay Area · CallTools
88 appointments, and a file ranked on every one of them
A California solar call center dialed two 25,000-record Bay Area files and booked 88 appointments in 15 days. We read every dial back into the next file: ranked on what booked, the top 40% of records held 23 of 25 bookings.
- Who
- A California solar call center
- Sells
- Solar appointments for an installer, by phone
- Market
- San Francisco Bay Area
- Team
- 11 agents, in-house
- Dialer
- CallTools predictive
- Volume
- ~31,000 dials per weekday
The result
Two files, fifteen days, 88 appointments
The floor took the first 25,000-record file on August 31 and the second on September 8. Every call on both went through their dialer and back to us through its API, so this is their log, not our count: 88 appointments, 49 on the first file and 39 on the second, with 14 of them coming in as callbacks from homeowners in the file.
A booked call ran 6½ minutes at the median, against 19 seconds for the typical connect. The last dialing hour, 6 to 7 pm, booked more appointments than any other.
What their agents tagged
Every call an agent finished got a disposition. These are the three that grade a file, on the first two days of the first file.
The third dial books more than the first
Pickup falls with every pass, as it does on any file. Appointments do not. Per live connect, the third dial booked 2.4 times what the first one did, and the fifth and sixth kept booking. A record is not spent after one dial, so the right move is to work a file three times before asking for the next one.
What booked
With 88 appointments joined back to the record behind each one, five traits separated the homes that booked from the ones that did not, each adjusted for the others. Two of them, phone activity and time in the home, we had never ranked a file on before.
Pool homes were the surprise. They pick up at a normal rate and almost never book: over the first three passes, one appointment in roughly 18,700 dials. They are also one of the strongest predictors of who installs solar, which is what the first file was built on. Who installs and who books a cold call are different lists.
The file, ranked
To check the traits were real and not a story told after the fact, we scored the second file with weights learned on the first file only. Its top fifth booked 2.8 per 1,000 records. The bottom three fifths, together, booked 2 of 25. The third file shipped on September 15 ranked this way.
23 of 25 bookings in the top 40% of a ranked file, on 24,968 records, 25 bookings. Weights from the first file only; the five traits were chosen looking at both.
Why the second file felt better
The floor said the second file was much better. At the same dialing depth it was level: 20 appointments against the first file’s 21 after three passes over the same number of records. What changed was the conversations. Scheduled callbacks nearly doubled, from 1.2 to 2.2 per 100 tagged calls, and “doesn’t qualify” fell from 4.0 to 2.7. But only 3.4% of the numbers that asked for a callback ever booked. Warmer calls feel like a better list. They are not the same thing, so we grade on appointments set.
What to take to your own floor
Work every record three times. Connects fall with each pass; bookings per connect rise. Replace the file after the third dial, not the first.
Rank on phone activity and time in the home. A mobile with carrier activity in 5 of the last 12 months and an owner in year zero to five each roughly triple bookings per record. Skip pool homes.
Grade the file on appointments set. Callbacks and friendlier conversations are sentiment. Fewer than 1 in 25 callbacks became a booking.
Send the export. Every finding above came from the dialer’s own log joined to the delivered file. That is how the next file gets ranked.