How the weekly scorecard is organized
The scorecard has two levels. The company level gives your DSP an overall standing for the week and the metrics behind it. The driver level lists each delivery associate by name and transporter ID, with their own standing and their own numbers for the same metrics. Some metrics only make sense at one level, and the legend on your scorecard says which.
Many DSPs download the driver-level table as a spreadsheet each week. That file is what you work from, because you can sort it, filter it and keep last week's beside it.
The standing tiers from Fantastic Plus down
Standings run from Fantastic Plus at the top down to the lowest tier. The tier names are not stated the same way everywhere: our sample file and the filters in SXT Ops use Fantastic Plus, Fantastic, Great, Fair and Poor, while a scorecard guide published in April 2026 lists six levels, with Good and At Risk among them. Use the names and cut-offs printed on your own current scorecard, and note the date you read them.
Standings change with Amazon's thresholds and weighting, so a driver who was Fantastic last quarter at the same numbers may not be now. That is why this guide gives no numbers for the tiers.
Safety metrics and where Mentor and Netradyne fit
The safety columns come from what happens in the van. Two systems are common at DSPs. eDriving's Mentor app uses the driver's phone sensors to measure harsh acceleration, hard braking, harsh cornering, phone distraction and speeding, and gives each driver a FICO Safe Driving Score. Netradyne's Driveri cameras capture drive time and raise alerts on driving events; Netradyne rates drivers with its GreenZone score.
On the driver table these usually appear as event rates. In our sample file the safety columns are seatbelt-off rate, speeding event rate, distractions rate, following distance rate and sign or signal violations rate. Your file may differ; the column headers and legend on your scorecard are the definition that counts.
Because safety events come from devices, a wrong event is possible. Coaching from Netradyne and Mentor events covers how to handle one.
Delivery quality metrics
These are the columns most owners know by their initials. The meanings below are the common ones; confirm each against your scorecard's legend.
| Column | Usual meaning |
|---|---|
| DCR | Delivery completion rate: the share of assigned packages delivered. |
| DNR DPMO | Delivered, not received: packages marked delivered that the customer reported missing, counted per million opportunities. |
| POD | Photo on delivery: whether drivers took the required delivery photo. |
| CC | Contact compliance, where your scorecard includes it. Read the legend for what counts as a contact. |
| CE | Customer escalations: complaints about a driver's behavior, safety or package handling that reached Amazon. |
| CDF | Customer delivery feedback, where your scorecard includes it. |
| DSB | A delivery behavior measure on some scorecards. Read the legend for its definition. |
Two columns tell you the most about a driver's week. DCR shows whether the route got done, and DNR shows whether what got done reached the customer. A driver with a high DCR and a high DNR rate may be worth a closer look: packages marked delivered that customers report missing.
What changed in 2026
DEFT surveyed DSP owners after a 2026 scorecard update and published the results on July 23, 2026. Before the update, 74.2 percent of respondents said they reached Fantastic Plus between 80 and 100 percent of the time. After it, 22.7 percent did, and 90.6 percent said the new thresholds had hurt them financially. DEFT notes these are answers from respondents, not a count of every DSP.
The practical point: if your standing dropped this year while your drivers' numbers held, you may be looking at a change in the scale rather than in your station. Compare your drivers against their own past weeks, not against an old tier chart.
Matching scorecard rows to associates
Before you coach anyone, make sure each scorecard row is the right person. Match by transporter ID first, because names repeat and change. Then fall back to your own employee ID, and only then to an exact name.
Rows that match nobody are worth a look every week. In our sample files, Aaliyah Martinez appears on the scorecard but not on the roster export, and Brandon Ramirez appears twice on the roster. Either one would throw off a weekly coaching list if nobody caught it.
| Associate | Standing | DCR | DNR DPMO | POD | Seatbelt off rate |
|---|---|---|---|---|---|
| Brandon Ramirez | Fantastic Plus | 99.3% | 72 | 98.3% | 0.21 |
| Mei Murphy | Fantastic | 99.2% | 611 | 99.3% | 0.12 |
| Priya Davis | Fair | 99.0% | 98 | 95.4% | 0.09 |
| Tyrone Ortiz | Poor | 99.6% | 1040 | 99.6% | 0.18 |
The sample is made up, but it shows something real scorecards do too: a high DCR alone doesn't explain a standing; here DNR is the outlier. Tyrone Ortiz has the highest DCR of these four, the worst DNR and the lowest standing. Read the whole row before you decide what to talk about.
Turning a bad week into coaching
Write your coaching rules down before you look at names. For example: anyone below your DCR line, anyone with a seatbelt-off event. Then apply them to everyone the same way, and if you skip someone, write down why. Rules set in advance are fairer to drivers and easier to defend if a decision is ever questioned.
Keep coaching separate from discipline. A coaching note says what the numbers showed, what you talked about and what help you offered, with a date to check in. For example: "Talked about DCR last week. Two missed deliveries were apartment lockers. Showed the locker code screen and set a check-in for next Friday." A write-up is a different step with its own record.
Check in on the date you set. A note with no follow-up reads, months later, as if nothing was done.
How SXT Ops handles this
On the Station plan ($399 per station per month), SXT Ops imports the weekly scorecard as a CSV or Excel file and matches each row by transporter ID, then employee ID, then exact name; rows that match nobody show as Not linked. Each associate's record holds their scorecard history, and drivers see their own scorecard in the driver app unless owners hide it. After an import, Coaching lists people who cross the rules you set; the default rules are placeholders, not Amazon's tiers. Nothing is created until a person chooses to, and drivers don't see coaching items in the app.
More on scorecards and coaching, the Netradyne and Mentor coaching guide, or bring one station to the pilot.
Changes to this guide
- : First published. Tier names compared with a public guide dated April 12, 2026; survey figures from DEFT, July 23, 2026.