When the power blinks, your site is read on a phone at four percent
The visit you should design for is one hand, cellular data and a homeowner who is already annoyed.
Rain events and Entergy outages produce a very specific kind of traffic. No desktop, no patience, sometimes no home internet, and a battery the reader is rationing.
Design the first screen for that person. What you do, which parishes you cover, and a phone number that dials on tap. Not a full width video, not a slideshow, not a cookie banner covering the only button.
Put the coverage answer in words, not on a map graphic. Orleans, Jefferson and St. Tammany, with the neighborhoods listed underneath. A map image is heavier, slower and harder to read on a small screen than a plain sentence.
Give the emergency path its own route. A person with active water coming in should reach a call button in one tap, without passing through a services menu that assumes they are shopping.
One question about piers changes which crew and which ladder shows up
Every field on the form should change what your dispatcher does, and the fields that matter here are not the ones on a national template.
Ask whether the house is raised or on a slab. Much of the older city sits on piers rather than slab, which affects access, ladder setup and how long the crew is on site.
Ask the covering, the number of stories and whether water is coming in right now. Those four answers, plus the address, decide who is sent and how quickly. A budget field decides nothing and costs you submissions.
Keep it short and specific. A form that reads like the first sixty seconds of a phone call with a good estimator gets finished, because every question obviously has a reason.
Add a photo upload. Homeowners here will send a picture of the ceiling stain or the shingles in the yard, and one photo saves an appointment that did not need to happen.
Confirm what happens next in plain words on the confirmation screen. Who calls, roughly when, and what they will ask. Silence after a submission is where inquiries go cold.
Photograph the staging problem, not the finished ridge
The pictures that persuade are the ones showing you have solved the access and mess problem on a lot like theirs.
Finished roofs all look alike from a drone. What a homeowner in the Marigny or Uptown is actually worried about is a narrow lot, no driveway, a fence between the truck and the ladder, and a neighbor twelve feet away.
Show the setup. Ground protection, where the dumpster went, how you reached a second story from a side yard barely wide enough to walk, how a gallery or ornamental detail got covered before tear-off began.
Use your own crews and your own trucks. Stock photography reads as stock photography, and in a market where storm chasers appear regularly, proof that you own equipment and employ people is a real differentiator.
Caption the photographs with the neighborhood and the roof type, not with adjectives. A reader recognizing their own street does more work than any slogan you could put beside the image.
The chat widget loads before your phone number, and in a rainstorm that is the whole site
Third party scripts are usually the reason a roofing site is slow, and speed here is a business problem rather than a technical score.
A typical roofing site carries a chat bubble, a review carousel, a tracking pixel or three, a font service and a gallery of full resolution photographs. Together they can outweigh the entire page of text the visitor came for.
Set a budget in kilobytes and hold to it. Compress and correctly size every photograph. Load the gallery only when someone scrolls to it. Defer everything that is not needed for the first screen to render.
Test on a mid-range Android phone over cellular data, not on the office fiber connection with a new laptop. The gap between those two experiences is where the abandoned visits live.
Keep the phone number in markup that renders immediately rather than inside a script that loads a widget. A number that appears two seconds late is a number a wet ceiling did not wait for.
Every shotgun with a rear addition has two roofs, and the site should show both
The site architecture should match the housing stock instead of a generic service menu.
A pitched front and a low-slope rear addition are extremely common in the older neighborhoods, and the transition between them is where problems tend to show up. Most local roofing sites have no page for that at all.
Build pages that follow the roof: pitched replacement, low-slope and flat sections, slate and specialty work, and repairs. A homeowner navigating by roof type finds the right page faster than one reading a list of service names.
Separate repair from replacement clearly. A person with a ten year old roof and one leak will leave a site that only sells full replacement, and they are the same person who will need a full replacement later.
Give commercial its own section if you do that work. A manager comparing contractors for a converted warehouse in Bywater needs different evidence than a homeowner in Gentilly, and a shared page serves neither.
Your warranty page should be written for Louisiana, not lifted from a national kit
Louisiana operates under civil law rather than the common law used in every other state, which makes borrowed contract and warranty copy a liability rather than filler.
Anything on the site describing agreements, warranties, deposits or claims should be written or reviewed by your own Louisiana counsel. Template language bought with a website theme was not written for this state.
Never promise an insurance outcome anywhere on the site. Describe what your crew documents and photographs, keep it general, and tell readers to confirm coverage questions with their insurer and their own attorney.
Where you mention approvals for work on a visible exterior in a historic district, hedge and point people to the relevant commission or the city. Requirements vary and change, and a confident sentence that turns out to be wrong is the one a reader remembers.
Put the license number, proof of insurance and financing options on a single page an office manager can text to a customer. Buried credentials are the same as no credentials during a five minute comparison between three companies.
Questions we actually get
- How many fields should our estimate form have?
- As few as possible, with every one earning its place by changing dispatch. Address, whether water is coming in now, raised or slab, covering type, stories, and a photo upload will carry most of the load. Cut anything that only feeds a report, and never ask for a budget on the first form.
- Do we need a separate page for each neighborhood we serve?
- Only where the page has real content: roof types common in that area, work you have done nearby, and the questions you actually get. Pages spun from one template with the neighborhood swapped in look automated to readers and to search engines, and they make the site feel less trustworthy rather than more local.
- Is a live chat widget worth adding?
- It depends entirely on whether a person answers it. An unattended chat that replies during business hours only, with a form behind it, adds weight and delay for nothing. If real staff or a monitored answering service handle it quickly, load it after the main content and keep the phone number visible regardless.
- Should we show prices on the site?
- Ranges tied to clearly stated assumptions can filter out mismatched inquiries, but hard numbers on a roof nobody has inspected create arguments later. A more useful approach is explaining what drives the price on the kinds of roofs common here: access, pitch, tear-off layers, decking condition and the covering itself.
- How fast should the site be?
- Fast enough that the phone number and first screen appear immediately on a mid-range phone using cellular data. Test in that condition rather than in the office. Speed here is not a vanity score, it is whether a homeowner standing under a leak waits for you or backs out and taps the next result.
