White-glove deliveryJuly 7, 2026· Updated October 14, 20264 min read

The Final-Mile KPIs Every White-Glove Operator Should Track

On-time rate alone hides the problems that matter most in white-glove. Here are the metrics that actually predict retention and margin.

Two-person white-glove crew carrying a wrapped sofa into a luxury home entryway at sunset

White-glove is a high-margin business with a lot of ways to lose the margin, and parcel or LTL numbers don't carry over. Take on-time percentage. The parcel world worships it. In white-glove it tells you almost nothing, because a truck that arrives in the window and scratches a $12,000 cabinet did not have a good day.

If you run a final-mile white-glove operation, these are the numbers we'd put on the wall, because they line up with retention, referrals, and gross margin. Track them weekly. Show the crews.

1. Window adherence, not on-time rate

Nobody paying for white-glove cares that you arrived "by end of day." They care that you showed up in the two-hour window you promised, because they took the morning off to be there. Window adherence is the share of jobs that started inside the committed window. Measure it at the gate or the front door. Not at "dispatched."

Below 90% and customers have stopped trusting your windows, whether they've said so or not. Above 95% and you have something competitors can't easily copy. The gap between the two is nearly always scheduling discipline, not driving speed.

2. First-attempt success

The share of deliveries completed on the first visit, no return trip. A failed first attempt is brutal. You burn the truck and the two-person crew, you spend down the customer's patience, and you still owe them the delivery.

Most failures trace back to access or product. Access means nobody reserved the elevator, or the building wanted a certificate of insurance and never got one. Product means the wrong item is on the truck, or the right one got damaged on the ride over. Both get fixed with a better pre-delivery checklist and a call to the property manager the day before.

3. Damage-claim rate per 1,000 items

Per item, not per delivery. Jobs in white-glove run anywhere from one armoire to a whole house, so a per-delivery rate lies to you. A per-item rate tells you whether handling and packing standards are holding.

If the number is creeping up, look at your intake photos before you look at the crews. Damage "discovered" at delivery is often damage that happened before pickup and never got written down. The fix is tighter intake, with mandatory six-side condition reports on every piece before it leaves the warehouse.

4. Customer-signed proof-of-delivery completion

How many jobs end with a digital signature, a condition acknowledgement, and at least one photo of the item in place, all uploaded before the truck leaves the curb? It should be 100%. Every job short of that is a damage dispute you'll lose later, with nothing to show.

This one measures process, not opinion. If it's sitting at 87%, you have a crew-training problem. Not a software problem.

5. Hours per job, by job type

Track average crew-hours separately for in-home placement, simple drop-off, assembly, and packing. What matters is the trend. When hours drift up, costs are rising and nobody has said anything yet. When they drift down and complaints don't spike, you've found real productivity. Go find out which crew found it.

6. Revenue per stop

Total revenue divided by completed stops, week over week. This is the cleanest single number for whether the business is growing or just getting busier. Those aren't the same thing. Two stops at $400 beat four stops at $150 most weeks, because the cost lives in the stop, not in the revenue.

7. Customer NPS or post-delivery rating

One question, sent within four hours of completion. Don't wait until the next morning. The response rate triples when the survey lands while the customer is still looking at the sofa in its new spot, and the comments are far more specific. Use the answers to work out which crews and routes produce five-star days.

How to make the metrics move

A dashboard changes nothing on its own. The numbers move when they show up in the morning huddle, the weekly ops review, and the monthly conversation about pay. Good white-glove delivery software works these out in the background and puts them where the team already is: the dispatch screen, the crew app, the manager portal. Nobody should need a separate analytics tool to find out how Tuesday went.

If on-time rate is still your headline KPI in 2026, you're measuring the wrong thing. Your customers don't remember that you arrived. They remember the chipped cabinet, the missed window, and the second trip. Build the dashboard around the numbers that predict those three, and the business gets better in the places people actually notice.

Frequently asked questions

Why is on-time rate a poor KPI for white-glove delivery?

Because it measures the wrong promise. Customers were told a two-hour window, not “by end of day,” and they judge the job on placement, care, and condition as much as timing. A delivery that lands inside the window and damages a piece scores perfectly on on-time rate and terribly with the customer.

What is window adherence and what is a good target?

It is the share of jobs that started inside the committed window, measured at the gate or the front door rather than at dispatch. Below 90% is a trust problem; above 95% is a competitive advantage. The gap is almost always scheduling discipline rather than driving speed.

Why measure damage claims per 1,000 items instead of per delivery?

White-glove jobs vary enormously in item count, so per-delivery rates flatter big jobs and punish small ones. A per-item rate tells you whether handling and packing standards are holding. When it drifts up, check intake photos first. Damage “discovered” at delivery often happened before pickup and was never documented.

What proof-of-delivery completion rate should we expect?

One hundred percent. Every job should end with a digital signature, a condition acknowledgement, and at least one delivered-state photo uploaded before the truck leaves. A rate of 87% is a crew-training gap, not a software gap, and it is open exposure on any future dispute.

Is revenue per stop more useful than total revenue?

For week-over-week health, yes. Variable cost sits in the stop (the truck, the two-person crew, the drive), so two $400 stops usually beat four $150 stops. Watching revenue per stop tells you whether the business is growing or just getting busier.

When should we send the post-delivery survey?

Within about four hours of completion, while the experience is fresh. Response rates rise sharply compared with next-day sends and the comments get specific enough to tell you which crews and routes produce five-star results.

How do we get these KPIs to change crew behavior?

Put them where the work already happens (the dispatch screen, the crew app, the weekly ops review) instead of in a separate analytics tool nobody opens. Metrics move when they show up in the daily huddle and the compensation conversation.