DSP Operations Guide
The Amazon DSP Scorecard, Explained
Your scorecard is the single most important number in your DSP — it drives your bonus, your standing with Amazon, and whether your business grows or gets squeezed. Here's what Amazon actually measures, how the tiers and FICO score work, why scorecards slip, and how to fight incorrect penalties — from operators who live in the scorecard every week.
The short version
Every week — usually Wednesday — Amazon drops a scorecard PDF plus supporting files (Netradyne safety data, EOC, PPS, CX feedback, escalations) into your DSP portal. The scorecard scores four categories — Safety and Compliance, Reliability, Team, and Quality — and averages them into an Overall Standing that maps to a tier from At Risk up to Fantastic Plus. The data lags several days, so weekly review alone is too slow — the operators who hold Fantastic Plus monitor daily and coach before a metric crosses a threshold. (Nizod)
The four categories
Your Overall Standing is an equally weighted average of four categories — so a single weak category drags your whole tier down, even if the other three are strong.
Safety & Compliance
Driven by your FICO score from Netradyne plus Engine Off Compliance (EOC) and Proper Parking Sequence (PPS). This is where speeding, seatbelt, harsh braking, and parking-safety behavior land.
Reliability
Delivery Completion Rate (DCR) and delivery quality — did your team deliver what it was assigned, without packages coming back to station? Highly sensitive to dispatch and rescue quality.
Team
Reflects how you staff, retain, and onboard drivers. Turnover and staffing gaps show up here and quietly cap the other three categories.
Quality
Customer-facing outcomes: Delivered Not Received (DNR), Photo on Delivery (POD) compliance, and customer escalations. The metrics your end customers actually feel.
Tiers and the Fantastic Plus bonus
Amazon rates every DSP on a tiered ladder: Fantastic Plus, Fantastic, Great, Good, Fair, and At Risk. Most stations sit between Good and Fantastic; reaching and holding the top is what separates the best operators. Because Amazon evaluates rolling windows, Fantastic Plus is a consistency game — one bad week can undo a strong month. (Nizod)
Amazon rewards Fantastic Plus with a bonus — roughly $5,000 each time it's achieved per Nizod — which for a well-run station can offset a meaningful slice of operating costs. Exact bonus mechanics and thresholds vary by contract and shift over time, so confirm yours with your business coach.
The metrics that move your score
Not all metrics are equal — some recover in days, others linger for weeks. These are the ones that decide your tier:
DNR
Delivered Not Received
Customer reports a package was marked delivered but never arrived. Heavily weighted, slow to recover, and hard to dispute without a timestamped POD photo.
DCR
Delivery Completion Rate
Share of assigned packages actually delivered. Sensitive to dispatch quality and call-out coverage — but it responds to operational fixes within days.
POD
Photo on Delivery
Whether drivers take required delivery photos. Almost entirely a training and habit issue — and your primary evidence for winning DNR disputes.
FICO
Netradyne Safety Score
Aggregate fleet driving safety: speeding, harsh braking, distracted driving, seatbelt-off, sign/signal violations, following distance. A declining FICO is a coaching problem.
EOC
Engine Off Compliance
Whether drivers shut off the engine at each stop. A quiet metric many owners never coach — but it compounds across hundreds of stops per route.
CX
Customer Feedback (PRR)
Positive Response Rate from customer sentiment. Amazon's Fantastic threshold is roughly a 98% PRR. Escalations for behavior or mishandling drag it down fast.
Metric definitions compiled from operator resources including Bridgetown Delivery and Nizod. Exact metrics and thresholds are set by Amazon and change over time.
Why scorecards slip
It's rarely a data problem — Amazon gives you more performance data than almost any contractor relationship in logistics. It's an action problem:
- Nobody's watching between Monday and Friday — a metric that moved Tuesday has days to compound before anyone looks.
- Coaching happens after the threshold, not before it, so the scorecard hit is already baked in.
- Dispatch problems create metric problems — bad routes drive late deliveries, DCR misses, and escalations downstream.
- Documentation gaps lose disputes — no photos, no incident logs, no coaching trail means you absorb penalties you could have fought.
Operator note
The stations that hold Fantastic Plus don't have better drivers — they have a daily rhythm. Someone reviews yesterday's metrics every morning, watches pace during the day to catch rescues before they happen, logs incidents every night, and coaches from data once a week. That's the whole game, and it's exactly the work that falls apart when one person is also running dispatch, HR, and everything else.
Disputing incorrect penalties
A meaningful share of scorecard penalties are wrong or outside your control — GPS misfires, unverified customer DNR reports, station sorting errors, miscategorized Netradyne events. Amazon's dispute window is typically about ten days from when a penalty appears, which isn't much time once you notice it, gather evidence, and submit. (Nizod)
Winning disputes comes down to evidence captured in real time: timestamped delivery photos (your best defense against DNR), driver incident logs, device-failure records, and documented coaching. The operators who win don't start when a penalty lands — the documentation already exists as a daily habit, so a dispute takes hours, not days.
Protect your scorecard
Turn scorecard data into daily action
The scorecard rewards the operators who catch problems on Tuesday, not the ones who read the PDF on Saturday. Sortd's performance module pulls your metrics into one view, flags drivers before they cross a threshold, and keeps the coaching and documentation trail that wins disputes — alongside scheduling and live dispatch that fix the upstream causes. Built by DSP operators, read-only, and TOS-safe — it stores zero Amazon credentials.
Frequently asked questions
What is the Amazon DSP scorecard?
The Amazon DSP scorecard is a weekly performance report Amazon delivers to every Delivery Service Partner. It aggregates dozens of delivery, safety, customer-experience, and compliance metrics into four categories — Safety and Compliance, Reliability, Team, and Quality — and rolls them into an Overall Standing that places your station in a performance tier.
What are the Amazon DSP scorecard tiers?
From top to bottom: Fantastic Plus, Fantastic, Great, Good, Fair, and At Risk. Most stations sit between Good and Fantastic. Your tier is based on rolling performance windows, so one bad week can pull down a month of strong numbers, and one good week cannot rescue a bad month.
What is a FICO score on the Amazon DSP scorecard?
The FICO score is a driving-safety score generated by Netradyne's in-vehicle cameras. It aggregates speeding, harsh braking, distracted driving, seatbelt-off events, sign and signal violations, and following distance across your fleet, then feeds into the Safety and Compliance category. It also affects your insurance costs and standing with Amazon.
How does the Fantastic Plus bonus work?
Fantastic Plus is the top scorecard tier. Amazon rewards it with a bonus — Nizod reports roughly $5,000 each time it is achieved — though Amazon's exact bonus structure and thresholds vary by contract and change over time. Holding Fantastic Plus is about consistency across every metric, not a single strong week.
Why does my DSP scorecard keep slipping?
The most common causes are nobody monitoring metrics between Monday and Friday, coaching that happens after a threshold is already crossed, poor dispatch that creates downstream DCR and escalation problems, and documentation gaps that lose disputes. Scorecards reflect whatever your operation naturally produces without systematic daily intervention.
Can you dispute Amazon DSP scorecard penalties?
Yes. Amazon's dispute window is typically about ten days from when a penalty appears. Customer-reported metrics like DNR, customer escalations, and DCR exceptions caused by station errors are the most disputable — but only with timestamped delivery photos, incident logs, and coaching records captured at the time, not reconstructed later.
Stop managing your DSP the hard way.
Join operators who’ve automated the boring so they can focus on growth.
Not ready to book? Drop your email and we’ll send the operator’s rundown — no call required.