Most last-mile scorecards lead with on-time percentage and cost per stop. First attempt delivery rate belongs above both of them, because a shipment that fails on visit one contaminates every other number on the page — it inflates cost per delivery, breaks the promise date, and generates a support ticket before anyone notices. This guide covers how to define the metric precisely, what a defensible benchmark looks like, what a second attempt actually costs, and which fixes reliably move it.
What the metric actually measures
First attempt delivery rate (FADR) is the share of shipments completed on the first physical visit to the delivery address — no redelivery, no hold-at-terminal, no reschedule.
The formula is trivial:
FADR = (Shipments completed on first attempt ÷ Total shipments attempted) x 100
The difficulty is entirely in the definitions, and this is where most shipper-carrier disputes start.
Three places the definition breaks
Denominator drift. The denominator must be unique shipments, not delivery events. If a package is attempted three times, that is one shipment in the denominator, not three. Carriers that count attempts inflate the rate; a 200-shipment day with 40 failures and 60 total reattempts reads as 80% under a shipment-based denominator and 87% under an attempt-based one. Fix the denominator in the contract, not in the QBR.
What counts as an attempt. A driver who arrives at the geofence, dwells 90 seconds, and departs has made an attempt. A driver who runs out of hours and never leaves the terminal has not — that is a service failure, but it belongs in a separate bucket. Without geofence entry plus a timestamped scan or photo, “attempted” is a self-reported claim. Require GPS-stamped proof before an exception code is accepted.
Exclusions. Decide up front whether refusals, customer-requested reschedules, weather embargoes, and closed-business stops sit inside or outside the denominator. There is no universally correct answer, but there is a wrong one: changing the rule mid-year. Publish a gross FADR (everything included) and a controllable FADR (carrier-attributable failures only) and report both.
What good looks like
Public benchmarks are thinner than vendors imply, so treat any single number cautiously. The most useful anchor comes from Loqate’s Fixing Failed Deliveries research, which found that roughly 8% of U.S. first-time deliveries fail — implying a national reference point somewhere near 92% for standard parcel-style residential volume.
That is an average across a very wide range. Segment before you judge yourself:
- Commercial B2B, weekday, dock or receiving desk — the highest band, and it should be. Failures here are almost always appointment or hours-of-operation errors.
- Residential, no signature, release authority granted — high, because the driver has a fallback.
- Residential, signature or ID required — drops sharply. This is a policy choice with a measurable cost.
- Appointment-based big-and-bulky — strong when the appointment is confirmed inside 24 hours; poor when it is confirmed at booking and never re-confirmed.
- Multi-family, gated communities, and HOA-controlled properties — consistently the weakest segment, and in dense South Florida markets it can be a large share of the route.
A blended 90% that hides 97% commercial and 78% multi-family is not a benchmark. It is an average of two different businesses.
The real cost of a second attempt
The Loqate research put the average cost of a failed U.S. first-time delivery at $17.20 per order. That figure covers the direct redelivery and handling cost — it does not price the downstream damage.
Work the arithmetic on your own network. A second attempt consumes a second stop’s worth of driver time, vehicle cost, and route capacity while producing zero incremental revenue, and it displaces a paying stop that could have occupied the slot. Add the storage day at the terminal, the support contact, and the elevated probability that the order converts to a return.
Then there is retention. Research from Convey, cited by Loqate, found that 84% of shoppers are unlikely to return after a poor delivery experience. And the problem is not shrinking: in the fourth annual ecommerce home delivery study from Descartes and SAPIO Research, released in May 2025 and based on 8,000 consumers across North America and Europe, 79% of respondents reported experiencing a delivery problem — up from 67% the prior year, with dissatisfaction concentrated among the under-35 shoppers driving online growth.
Instrument it so it is actionable
A single percentage tells you nothing about what to fix. Every failed first attempt needs a mandatory, mutually exclusive reason code captured by the driver at the point of failure:
- Consignee not present / no safe release location
- Address incorrect, incomplete, or unlocatable
- Access denied (gate code, front desk, elevator not reserved, no COI on file)
- Refused at the door
- Recipient rescheduled
- Damage or shortage discovered on the truck
- Route capacity — stop not attempted
Free-text notes are where root-cause analysis goes to die. Codes are what let you say “31% of our failures are access, not addressing” and staff accordingly.
The fixes that actually move it
Validate the address at capture, not at the dock. In the Loqate survey, 71% of businesses identified inaccurate delivery addresses as a primary cause of failure. Address autocomplete and verification at checkout, plus a unit/apartment field that cannot be skipped, removes the single largest failure category before the order is ever picked.
Capture and verify a mobile number. An email address is not a delivery tool. A same-day SMS with a narrowing ETA window converts “not home” failures into completed stops.
Re-confirm appointments inside 24 hours. Confirmation at booking is stale by delivery day. A confirm-or-reschedule prompt the day before shifts failures from the truck to the phone, where they cost almost nothing.
Build an access intelligence layer. Gate codes, elevator reservation rules, certificate-of-insurance requirements, loading dock hours, and building contacts should live on the address record and travel with every future shipment to that location.
Enforce proof of delivery discipline. Geotagged photo, timestamp, and code selection on every stop — success or failure. Without it, your FADR is an estimate.
Set a reporting cadence
Review controllable FADR weekly by segment and reason code, gross FADR monthly at the network level, and hold the definition constant for at least four quarters so trend lines mean something. Tie a portion of carrier scorecard weighting to controllable FADR rather than raw on-time, and the incentives line up.
The Bottom Line
First attempt delivery rate is a measurement problem before it is an operations problem. Define the denominator, force reason codes, segment by property type, and the fixes tend to identify themselves — usually starting with address data and appointment confirmation rather than routing software.
Go LTL runs appointment-based residential and commercial delivery across South Florida with geotagged proof of delivery and reason-coded exceptions on every stop. See our last-mile delivery services or request a quote from Go LTL to benchmark your current network.
Recent Comments