A photo of the sill triages better than any question you can ask
One image sorts the whole house job from the single stuck door before anyone spends a morning driving.
Make photo upload the first thing the form asks for and make it work from a phone camera. Homeowners send them willingly, because a picture is easier than describing a rot pattern to a stranger.
A photograph tells your estimator the material, roughly the age, whether the unit is original wood or a later aluminum retrofit, whether the sill is gone, and often whether the opening has racked. None of that survives a phone conversation with someone who does not know the vocabulary.
Use it to route. Repair intent goes to a different conversation than a whole house quote, and a small job in a far parish can be handled with a different path rather than a full in home visit.
Keep the images attached to the record in your system. When the same inquiry resurfaces four months later, the photo is what lets whoever answers pick the thread back up without starting over.
Send the inquiry to a phone in the truck, not to an inbox checked at lunch
Almost every dealer says they respond quickly, and almost every dealer has an inbox that proves otherwise.
Route form submissions to a phone as a text message, with the name, the parish and the photo attached, going to a named person with a named backup. An address that goes to a shared inbox is not a routing plan.
Set an internal standard for the first response and let the owner see the misses. The point is not a slogan about speed, it is that the second dealer to call is usually the second dealer considered.
Have a scheduling path that works after hours, because most of this traffic arrives in the evening when someone is finally looking at the window that bothers them. A callback window offered plainly beats a chat widget that collects a message nobody reads until morning.
Track missed and abandoned calls separately from voicemails. Missed calls are the leak nobody sees, because they leave no record in the CRM at all.
The wait between the measure and the emailed quote is where jobs go cold
Enthusiasm peaks in the room and decays every day the paperwork sits.
Time the gap. Pull twenty recent jobs, note the measure date and the date the written quote actually reached the customer, and look at the spread. The number usually surprises the owner, and it correlates with what closed.
Prepare the quote to leave the same day where the job allows it, even in a simple form with the detail to follow. A document that arrives while the customer still remembers the conversation is doing different work than one that arrives next week.
Write it opening by opening rather than as one figure. A homeowner who can see which two windows cost the most can make a partial decision instead of postponing the whole thing, and partial decisions turn into second phases.
If the job runs through a historic district and the paperwork adds time, tell the customer the sequence up front and give them a date you will follow up. Silence during a wait gets read as being forgotten.
Custom sizes mean lead times, and lead times lose the job at the end
The most avoidable cancellations in this trade happen after the customer has already said yes.
Nothing in a hundred year old shotgun house is a stock size, and openings that have shifted with the ground do not come off a shelf. Ordered units take time to arrive.
Say the lead time before the deposit, in writing, in a range you can live with. A customer who learns the timeline after signing spends the wait deciding whether they were misled, and some of them cancel over it.
Build two or three touches into the waiting period. A short note when the order is placed, another when it ships, another to confirm the install day. The contact costs a few minutes and it prevents the quiet period that competitors call into.
Measure cancellations after signature as their own number. Dealers usually track close rate and stop there, which hides a category of loss that is entirely fixable with communication.
Revenue per completed measure survives a job list this mixed
A close rate averaged across a two window repair and a full house restoration tells you nothing you can act on.
Track a small set of numbers all the way through: inquiries, contacted, booked measures, completed measures, quotes delivered, signed jobs, and revenue. The drop between booked and completed is where most shops are losing the most, and almost nobody watches it.
Split the report by parish and by job type. A measure in Algiers or Slidell costs more of a day than one in Lakeview, and an average across the metro hides which work is actually worth chasing.
Use revenue per completed measure as the headline number. It respects the fact that one Uptown restoration can outweigh a month of small jobs, and it stops the team from optimizing toward easy quotes.
Set the definitions once and write them down. Half the reporting arguments in a small dealership come from two people counting a booked measure differently.
Decide what counts as a win before the test, because the sample will tempt you
One metro will not produce the volume to settle small questions, so the discipline has to come from the method.
Pick changes big enough to matter. Requiring a photo, moving the response to a text, sending quotes opening by opening, adding a follow up sequence during the order wait. Changes at that size can show up in signed jobs. A button color cannot.
Write down beforehand what you expect to move, over what period, and what result would make you keep it. Without the rule set in advance, a good month gets credited to whatever was changed most recently.
Run one change at a time and give it a full season where seasonality is involved. Heat drives demand here, and a change made in the middle of a busy stretch will look better than it is.
Where numbers are thin, use call recordings and the estimator's account of the visit as evidence. Twenty recordings will not give you significance, and they will tell you exactly which question is being asked badly.
Questions we actually get
- Where do most window and door inquiries actually leak?
- In most shops, three places. Calls that are never answered and leave no record, booked measures that never get completed, and the weeks between a quote and a decision when nobody follows up. Instrument those three before changing anything on the website, because they are usually larger than any on page issue.
- How fast should we respond to a web inquiry?
- Faster than the next dealer, which in practice means minutes rather than hours during working time and a clear callback window after hours. There is no magic threshold worth quoting, but the pattern is consistent: the first useful conversation carries a large advantage in a market where homeowners collect several quotes.
- We do not have enough traffic for split testing. What should we do instead?
- Test sequentially and test big things. Change one part of the process, define in advance what would count as success in signed jobs, and give it a full season. Supplement thin numbers with call recordings and estimator debriefs, which surface intake problems that no test would have isolated anyway.
- Should we charge for measures in far parishes?
- Some dealers do, particularly for long crossings to the Northshore, and some qualify harder instead by requiring photographs and an opening count first. Either approach works. Charging without qualifying tends to reduce good appointments along with the bad ones, so it is worth trying qualification first.
- What is the single number an owner should watch weekly?
- Completed measures, alongside revenue per completed measure. Lead counts move for reasons outside your control and signed jobs lag by weeks. Completed measures respond quickly to how the phone is handled and how appointments are held, which makes it the number that tells you whether this week went well.
