The first three questions on a flooded basement call decide the next four hours
Intake is triage, and the order of the questions changes what the truck carries and when it leaves.
Water still coming in or stopped. Clean water or sewage. More than one fixture affected. Those three answers tell a dispatcher whether this is a sump pump, a main line or a supply failure, and whether it is a tonight job or a tomorrow job.
Put the safety question early. If water is near the electrical panel or the furnace, the caller should be told to stay clear and, where appropriate, to contact ComEd or an electrician. Say it as a caution, never as an assessment of their property.
Write the sequence down and put it in front of whoever answers. Most shops keep it in one experienced dispatcher's head, which works until that person is on another line during the first cold snap of the year.
Track where calls stall. If callers drop out mid intake, something in the sequence is asking for a last name before it has earned one.
An answering service that cannot ask about the cleanout is a voicemail with a bill
Most after-hours setups take a message, and a message is not a booking.
Plumbing calls go stale fast. Someone with water spreading across a basement floor calls the next company on the list within minutes, so the question is not whether you return the call. It is whether anyone answered.
If you use a service, hand them your triage script and test it. Call your own number at eleven at night and listen. Most owners who try this find the operator collects a name, a number and nothing a dispatcher could act on.
Set an escalation rule that fits the trade. Emergencies get a live handoff or a callback inside a short, defined window. Everything else can wait for morning, provided the caller is told which category they are in.
Measure it directly. Count after-hours calls, count how many were answered live, count how many became jobs. The gap between the first and third number is usually the largest single leak in the business.
A quote with a permit timeline beats a quote with a discount
Large plumbing jobs stall on uncertainty about time and process far more often than on price.
Sewer line replacement, overhead sewer conversion and flood control work involve permits, inspections and a hole in someone's yard or basement floor. A homeowner sitting on a quote is often not comparing numbers. They are trying to picture the week.
Put the sequence in the proposal. What happens on day one, when the permit is typically applied for, what the inspection step generally involves, how long the excavation stays open, what gets restored. Practice varies between the city and the surrounding municipalities, so say it varies and note that specifics are worth confirming with the local building department.
Include the camera footage and the photos. A homeowner cannot see the pipe, and one still frame of a collapsed section does more work than another paragraph of description.
Then chase properly. A large quote needs a date for the next conversation, agreed while the technician is still on site, and a record of whether it happened. Proposals nobody follows up on are the quietest loss in a plumbing business.
Ask how many buildings they own before you quote the one blocked stack
The caller with a backed up line in a three-flat may be a portfolio, and a single job quote treats them like a stranger.
Chicago holds a deep stock of two-flats, three-flats and courtyard buildings owned by small landlords and local managers. One of those relationships is worth many single visits, and it is usually won or lost on the first call.
Change the intake for that caller. How many units, how many buildings, who lets the crew in, who signs, and whether they prefer a call or an email. Answering a landlord like a homeowner is a mismatch they notice immediately.
Offer what they actually want, which is predictability. A named contact, a response commitment during business hours, photo documentation, invoices formatted per building. None of it costs money and all of it is a switching cost for a competitor.
Track them separately in reporting. Revenue per customer over a year, not revenue per job, or the numbers will keep insisting these calls are ordinary.
Count the jobs you booked and never billed, then find out what happened
The gap between a full dispatch board and a week of invoices is a conversion problem hiding inside operations.
Jobs fall out after booking for reasons that repeat. Nobody home, the tenant never got the message, the alley was blocked, the customer found someone sooner, the crew showed up without the right equipment for the space.
Log the reason every time. A short fixed list beats free text, and it should include cancellations and the calls you turned down for being out of area, since both were paid for in advertising.
Read the list monthly and act on the top two lines. If access problems dominate, the intake form needs to ask about the gangway and the rear alley. If no shows dominate, the reminder needs to be a text the morning of, not an email the day before.
Reconcile against money rather than the calendar. A board that looks full and a deposit report that looks thin are two different stories, and only one of them pays.
Seasonality will fake every before and after you run in this market
In a market with a hard winter peak, a change made in November will look brilliant no matter what it was.
Any comparison spanning a change in the weather is measuring the weather. A new intake script launched days before a cold snap will show a booking lift, and none of that lift belongs to the script.
Compare like with like. The same season against the same season a year earlier where the history exists, or a stable stretch of weeks with no weather event in either half. When in doubt, hold the change and wait for a quiet period.
One shop in one metro also produces a limited number of calls each month, so small differences never separate from noise however patiently you wait. Put the effort where the effect is large enough to see: answering after hours at all, rewriting the intake sequence, attaching footage to quotes, following up on proposals.
Order the work by the money behind it and how quickly it can be done. Formal split testing earns its place later, once both call volume and a calm stretch of weather allow it.
Questions we actually get
- We already answer the phone. What else is there to fix?
- Answering is the first step, not the whole job. The common gaps are what gets asked on that call, what happens between eleven at night and seven in the morning, and whether large quotes receive a scheduled follow-up. Most shops have never counted how many after-hours calls were answered live, and that count usually starts the conversation.
- Should we use an answering service or hire someone?
- It depends on volume and on what a missed emergency costs you. A service is cheaper and works if you give them a real script and test it regularly. A person on staff handles judgment better and costs more. Either way the test is the same: call your own number late at night and see whether the person answering could tell a dispatcher anything useful.
- How do we know which fixes to do first?
- Estimate the money behind each leak and the effort to close it. Unanswered after-hours calls usually top the list, followed by unchased quotes on large jobs, then booked jobs falling out for access or no show reasons. Working in that order beats running experiments on details you cannot measure at your volume.
- Can conversion work be measured honestly in a seasonal market?
- Yes, with care. Compare the same months across years rather than month over month, avoid drawing conclusions from any period containing a cold snap, and track outcomes through to booked and billed rather than to form fills. We will not attribute a winter lift to a change made in November, because that comparison cannot be trusted.
- What should we track that we probably do not?
- After-hours answer rate, reason codes on every job booked and never billed, follow-up completion on quotes above a threshold you set, and revenue per customer over twelve months rather than per job. Those four numbers say more about where work is lost than any traffic dashboard ever will.
