The overflow answering plan you buy in May is the one that saves August
Peak demand and peak understaffing arrive in the same week, and unanswered rings are the single most expensive leak in this trade.
Pull the call records for the worst stretch of last summer and count the calls nobody answered, plus the ones answered after a long hold. Most shops have never looked, and the number is usually worse than the owner expects.
Answer rate is a staffing decision made months earlier. An overflow service, a second trained person on the phones during peak hours, or a rotation that moves someone off dispatch into intake all cost real money, and all of it is less than the work walking away.
Test what the after hours service actually says. Call it at eleven at night from a phone nobody recognizes and listen to the whole thing. The script is often years old and promising something the company no longer does.
Web forms deserve the same scrutiny. A form submitted at four in the afternoon that gets a callback the next morning is competing against whoever picked up the phone in the meantime.
A changeout survey in West Lake Hills fails when nobody asked about the side yard
Intake exists to make a productive visit possible, and around here access is the detail most often missing.
Take the address first. It settles county, drive time, and which utility serves the property, and it decides whether the job is worth the slot at all.
Then capture what the visit needs. Equipment age and location, whether the outdoor unit sits at grade or on a slope, gate width, driveway length, and anything overhanging the equipment. Homes west of the city in West Lake Hills, Lakeway and Bee Cave regularly have all four working against a technician at once.
Ask for photos during the call and text the customer a link. A picture of the data plate and the side yard removes a second trip more reliably than any script change.
Where a replacement is likely, note whether the pad may need to move and whether ground conditions are known. Shallow rock near the surface can change what relocating equipment involves, and nobody should be assessing that from an office. Say plainly that a site evaluation settles it.
Newcomers do not haggle, they compare documents, so the proposal has to read alone
A large share of replacement buyers here moved recently, have no reference point, and will decide by reading rather than by negotiating.
A quote that only makes sense with the salesperson standing there is a quote that loses to a clearer one. The document goes to a spouse, sometimes to a parent in another state, and it has to argue on its own.
Include what was measured and why, what is being installed, what the install involves, what is excluded, and the financing options. Explain the method rather than just listing a total. Buyers with nothing to compare against reward that heavily.
Then follow up on a schedule somebody owns. A replacement decision often runs a week or more while the household collects other quotes. The contractor who checks back with something useful, rather than a discount, is frequently the one who signs.
Log the reason for every loss in one field with a short list of options. Price, timing, went with a warranty, never responded. Reading that list at the end of a season tends to be more instructive than any report an agency will send you.
One casual promise about a utility program becomes a write off at collection
Eligibility generally depends on which utility serves the address, and a phone estimate of what a customer will get back is a discount you have already given away.
Austin Energy is owned by the city, which makes this market different from most of Texas, where households pick a retail provider. Anyone on your phones who learned the trade elsewhere in the state may be carrying assumptions that do not hold here.
Boundaries also do not follow city limits or county lines neatly. An address in Pflugerville or out past the Hill Country edge may not be served the same way as one in the middle of town, and the person answering the phone is not the right person to be ruling on it.
Set one rule: nobody quotes an amount or confirms eligibility. The office confirms current terms with the utility serving the address, in writing, before it appears in a proposal.
Then listen to a handful of recorded calls specifically for that word. Where a customer heard a number that was never confirmed, you generally find out at collection, and the difference comes out of the job.
Separate the jobs you never booked from the ones you rolled a truck to and lost
Two very different failures get averaged into one conversion rate, and the fix for each is nothing like the fix for the other.
The first failure is intake. Someone called, or filled in a form, and never got scheduled. The causes are answer rate, hold times, a dispatcher declining work that was actually profitable, or an arrival window the customer would not accept.
The second failure is in the field. A truck went out, a diagnosis or a quote happened, and no work was released. The causes are pricing, presentation, the proposal document, or a technician who is excellent mechanically and uncomfortable asking for the job.
Measure them apart, and break both out by county. Booking rate near the shop in Travis County and booking rate on a Hays County run to Kyle or Buda usually differ, and the gap is often drive time rather than demand.
None of it works until the call tracking and the dispatch software agree on a single job number. Reconciling those two systems is the least interesting project on this list and reliably the one that pays for the rest.
Pre-commit to one number per change, or a heat wave will let you claim any result
One metro does not produce enough volume for small split tests, so the discipline has to come from method rather than from sample size.
Testing a button color needs traffic this business will never have. Worse, demand here swings hard with the weather, so any before and after comparison will happily confirm whatever you already believed.
Change big things instead, one at a time, and write down beforehand which single number should move and by roughly how much before you would call it a success. Guessing afterward is how bad changes survive.
The candidates worth the effort are all operational. The after hours script. Whether the arrival window is promised as a time or as a range. Whether the proposal is emailed or walked through. Whether a maintenance agreement is offered by the technician at the truck or by the office a week later.
Compare like periods. This July against last July, one shoulder month against another. And hold the change long enough that a single bad week of weather cannot explain the whole result.
Questions we actually get
- What should we fix first if we can only fix one thing?
- Answer rate during your busiest hours, almost always. Pull the call records from the hottest stretch of last summer, count what went unanswered or waited on hold, and multiply by your average job value. The number usually makes the staffing decision for you, and no improvement anywhere else in the funnel matters if the phone is not picked up.
- How do we know whether we are losing jobs at the phone or in the field?
- Track them as two separate rates. Inquiries that never became a scheduled visit is one number, visits that never produced released work is another. They have completely different causes, so a single conversion rate averages them into something you cannot act on. Break both out by county too, since drive time distorts the picture across a service area this wide.
- Is call recording worth setting up?
- Yes, and it is usually the highest value data a service business already has and never reviews. Listen for how the arrival window is promised, what gets said about utility programs, and where a dispatcher declined profitable work. Check your obligations on consent and notification, and post the notice properly; it is worth confirming the current requirements with your own counsel.
- Can we A/B test our way to a better booking rate?
- Not with the volume a single metro produces, at least not for small changes. Statistical testing on button copy needs traffic this business will not have. Change substantial things one at a time, decide in advance which number should move, compare like periods across seasons, and use call recordings as your qualitative sample.
- Should the technician sell the maintenance agreement or should the office?
- Both are defensible and it is a genuinely good thing to test. The technician has trust and is standing in front of the equipment. The office has a script and does not have another call waiting. Pick one approach, run it for a full season, count agreements per completed job, and then try the other rather than doing both halfway.
