The first question is which floor, and the second is whether it rained
Two questions asked in the first thirty seconds sort a roofing inquiry more accurately than five minutes of general conversation.
Water on the top floor after rain is usually a roof call. Water on the third floor of a six story building, in dry weather, frequently is not, and finding that out on the phone saves an estimator half a day.
Then get the building: type, number of floors, whether it is attached on both sides, and who owns it. A caller who turns out to be a shareholder rather than the owner changes the whole path, because the decision belongs to the board and the managing agent.
Write the script and hold to it. Everyone answering the phone should ask the same things in the same order, so the notes are comparable and nothing depends on who happened to pick up.
Recording calls will show you where the conversation falls apart, usually in the same place every time. Recording laws vary and change, so confirm the current rules with your own counsel before you turn anything on.
A roof survey with no key to the bulkhead door is a wasted morning
In this city the site visit fails on access far more often than it fails on price.
Before a survey is booked, somebody has to answer how the crew gets onto the roof, who holds the key or the code, and whether a resident has to be home. Locked bulkheads, roofs reached through a top floor apartment and fire escape access are ordinary here, and none of them sort themselves out on the morning.
Parking is part of the same problem. An estimator crossing into Manhattan for a ten o'clock appointment loses more time to the street than to the roof, so cluster appointments by borough and by neighborhood rather than by the order calls came in.
Confirm the day before, in writing, with the access details in the message. A survey that runs is worth far more than a survey that is booked, and those are two different numbers.
Track the failures. If a fifth of your appointments cannot get onto the roof, that is not bad luck, it is a missing question in the intake.
Between your price and the work sits a board packet nobody owns
Across much of the housing stock the proposal is not the last document, it is the first one in a submission that somebody has to assemble.
A co-op or condo job generally needs an alteration agreement, certificates of insurance naming the entities the building specifies, and a freight or access schedule before work can start. Requirements vary building by building and are worth confirming with the managing agent, but the fact that a packet exists does not vary.
Decide internally that assembling it is your job, not the owner's. The contractor who sends complete paperwork the first time gets scheduled ahead of the one who sends it three times with corrections.
Build the templates now, while nothing is pending. A standard cover note, a sample certificate request, an access and staging plan, and a schedule that fits inside a building's working hours.
Say so on the call. Telling an anxious shareholder that you handle the board submission is often the sentence that wins the job, and it costs you nothing except the preparation you should have done anyway.
Silence for six weeks usually means the packet is stuck, not that you lost
Proposals in this market go quiet for reasons that have nothing to do with your price, and treating quiet as a loss throws away work you already paid to create.
Boards meet on their own schedule, and a proposal can sit until the next meeting no matter how good it is. Your follow up has to survive that gap without becoming noise.
Make each contact useful. Send the insurance certificate the building will need. Send the access plan. Send a short note explaining what the board will likely ask about the coating warranty. Useful beats persistent every time.
Ask the managing agent one direct question: what is outstanding, and when does the board next meet. Most people will answer it, and the answer tells you whether to schedule crew time or move on.
Keep a real cadence rather than an impulse. Something at day three, day ten, day thirty and then monthly, until you get an actual no. The jobs that come back six weeks later are the cheapest work you will ever book.
One leak can produce five inquiries, and your CRM will call them five leads
In multi unit buildings the same problem gets reported by several people, and counting contacts instead of buildings makes your numbers fiction.
Three shareholders in the same line, plus the super, plus the managing agent, can all reach out about one roof. Any report built on contact records will show five leads and one job and conclude your close rate is terrible.
Key the pipeline on the building address. Every call, form and email attaches to the building, and the deal lives there. Everything downstream gets more honest immediately.
Tag by borough and by building type as well. A survey in Manhattan costs more to deliver than one eight blocks from your yard, and the reporting should be able to show it.
Then measure the small number of things that matter: inquiries by building, surveys completed, proposals submitted, jobs signed, and the time between each step. Where the time stretches is where the money is leaking.
Change one big thing per quarter, because the sample will never settle an argument
A single metro does not produce enough roofing inquiries to prove a small difference, so stop pretending a split test can rule on anything here.
Testing a button color or a headline against this volume produces a result you could get by flipping a coin. Waiting for significance means waiting past the point where the answer would have been useful.
Change things large enough to see. Rewrite the entire intake script. Replace the estimate document. Put access questions on the form. Move the phone number. Those produce differences big enough to read in a quarterly number.
Use qualitative evidence where the sample fails. Twenty recorded calls will show you the question that ends conversations far more reliably than any experiment your traffic can support.
Then hold the measurement steady. Compare quarter to quarter with the same definitions, and accept that seasonality and weather are part of the picture. One clean quarterly comparison beats four noisy weekly ones.
Questions we actually get
- How fast do we need to call back?
- For a leak, minutes, not hours. Somebody with water coming into a top floor apartment is calling several companies in a row and will book with whoever answers and can say a real time. For planned replacement the clock is slower, but the pattern still holds: the first competent response frames the whole comparison. If nobody can answer after hours, use a service that takes real details rather than a voicemail box.
- What should we actually measure?
- Five numbers, tied to buildings rather than contacts: inquiries received, surveys completed, proposals submitted, jobs signed, and the time between each step. Split them by borough and by building type. Anything more elaborate than that tends to be ignored, and anything less cannot tell you where inquiries are being lost.
- Is it worth chasing proposals that have gone silent?
- Yes, and it is the cheapest work available to you. Board driven jobs go quiet for structural reasons: meetings, budgets, an agent on vacation. A follow up that sends something useful, an insurance certificate or an access plan, gets answered far more often than one asking whether they have had a chance to review. Keep going until somebody gives you an actual no.
- Should we offer free inspections?
- It is the most common offer in the trade, which is exactly why it persuades very little on its own. What differentiates here is specificity: naming what the survey will document, how long it takes, how you get onto the roof, and what the owner receives afterward. Being clear about what happens is worth more than being loudest about the price of it.
- Can you split test our website?
- Honestly, not usefully at the volume one metro produces. Small effects need traffic you do not have, and waiting for a statistically sound answer would take longer than the answer is worth. We would rather rewrite the intake, rebuild the inquiry form around access and building type, and compare quarters with consistent definitions. Bigger changes, read over longer periods, is the method that fits this market.
