The proposal nobody chased is a bigger leak than the form nobody fixed
Gated work does not close because you sent a good number; it closes because you were still present when the board met.
Ask a plumbing company how many association and commercial proposals they sent last quarter, then how many were followed up more than twice. The second number is usually close to zero.
A board approval is a sequence, not an event. The engineer raises a question. Scope changes. A meeting slips a month. The contractor still in the conversation at that point wins a surprising share of the time, and it is rarely the cheapest one.
Set a cadence and hold it. Confirmation the day you send. A check at one week. A check at three weeks. A check around when that board usually meets. A quarterly touch on anything gone quiet. Nothing clever, just applied consistently.
One person owns it and it lives in a system rather than an inbox. Proposals tracked in a spreadsheet nobody opens are the same as proposals not tracked at all.
A building manager and a panicking unit owner need two different first questions
The opening thirty seconds should sort the caller into emergency dispatch or scoped proposal, and those conversations share almost nothing.
The owner with water on the floor needs three things: reassurance, a shutoff instruction, and a truck time. Every extra question makes the call worse.
The manager needs the opposite. They are testing capability. Have you worked a building of this age and height. Are you insured to their requirements. How do you handle a shutdown notice to residents. When could you walk it.
Write both scripts and put them where the phone is. Then train whoever answers to read the caller inside the first two questions rather than running one script at everyone who dials.
The common failure is a single script for both. A manager pushed through the emergency script hears a lead being processed, and managers always have three other numbers.
A truck turned away at the loading dock is a lost sale with a labor cost attached
In Brickell and Edgewater the job can be lost after it is booked, at the service entrance.
A crew that arrives without a certificate of insurance on file, without a freight elevator booking, or outside the building's permitted work hours does not do the job. It burns two hours and most of the customer's confidence.
Fix it at intake. Building name, floor, whether the owner or the association is authorizing, whether a certificate is required and to whom it should be issued, permitted hours, and how the crew gets in.
Keep a building file. Once you have worked a tower in Edgewater or Coconut Grove, the access rules, dock hours and management contact should be stored and attached automatically to every future job at that address.
As a conversion fix this is among the highest return available, because the demand was already won and the cost of losing it is a paid crew standing in a garage.
Bilingual intake usually fails at the handoff, not at hello
Answering in Spanish is the easy part; keeping the conversation in Spanish through dispatch, the technician and the invoice is where companies drop callers.
A caller from Little Havana or Hialeah greeted in Spanish and then transferred to an English-only dispatcher has had a worse experience than if you had never offered.
Map the whole path. Greeting, intake, quote, scheduling confirmation, the technician at the door, the invoice, the follow-up. Any English-only step in that chain is a place people leave.
Where you cannot staff every step, say so plainly. Publishing the hours that have Spanish-speaking dispatch beats implying it is always available and being caught out at ten at night.
Record and review calls in both languages. Most owners have never listened to a Spanish-language call their own company took, and that is generally where the surprises are hiding.
Add board deferred and access denied to your outcome codes
You cannot fix leaks you never label, and standard plumbing outcome codes are too coarse for how work gets authorized in this county.
Booked, not booked and canceled tells you nothing useful. The codes that earn their keep are reasons: price, timing, no answer, tenant with no authority, outside service area, no certificate on file, access denied, board deferred, awaiting association approval.
Two of those are specific to Miami-Dade and both are recoverable. Board deferred is a follow-up problem. Access denied is a process problem. Neither is a marketing problem, which is exactly why marketing spend does not fix them.
Make the code mandatory before a job can be closed. Optional fields are blank on precisely the calls you most need to understand.
Review the codes monthly with whoever answers the phone in the room. The pattern is usually visible in one sitting, and the fix is usually a line in a script or an item on a checklist rather than a redesign.
With Miami-Dade volume, test the script and the follow-up, not the button
One metro will rarely hand a plumbing company enough sessions to prove a small design change, so put the effort where the sample is large enough to teach you something.
Small-effect split tests need traffic most regional service companies simply do not have. Running them anyway produces confident-looking results that do not repeat, and decisions get made on noise.
What is genuinely testable: large structural changes with large effects, offer and script changes measured over quarters, follow-up cadence tested by holding a group back, and the hours you cover the phone.
Qualitative work is always available and badly underused. Thirty recorded calls reviewed properly will tell you more about your intake than a year of button tests, and the sample costs nothing.
Stay honest about attribution while you do it. When a manager searches, asks two colleagues and calls six weeks later, no analytics package draws a clean line, and pretending otherwise is how companies cut the thing that was working.
Questions we actually get
- We answer every call during business hours. Is intake still our problem?
- Very possibly. Answering is the floor, not the finish. The questions asked, whether the caller is sorted into emergency or proposal, whether authority and building access are established, and whether anyone codes the outcome all sit above that floor, and all of them leak work quietly.
- How do we handle after-hours calls without paying for a full desk?
- Decide which calls you actually want at night and route accordingly. A trained answering service with your two scripts, a clear rule about what gets escalated to a technician, and a same-morning callback promise you genuinely keep will usually outperform a voicemail box and a good intention.
- What is a reasonable follow-up cadence on an association proposal?
- Confirmation on the day, a check at one week, a check at three weeks, a check timed around when that board meets, then a quarterly touch. Note the meeting cycle for each building you bid, because it varies and it is the single most useful thing to know about a slow proposal.
- Should we publish pricing to filter out shoppers?
- Residential flat rates for common jobs can screen usefully and reduce wasted calls. Association and commercial pricing is a different matter, since publishing it mostly informs whoever bids against you next. Filter with scope and process detail instead.
- How do we measure conversion when the board approves months later?
- Measure two funnels. The same-day funnel closes at the booked job. The gated funnel closes at the approved proposal, with the source carried through from the first call. Report them separately, expect the gated one to look worse in any single month, and judge it over quarters.
