Onboarding a new DSP route: the first 30 days, in order

The first 30 days of a new Amazon DSP route are the days that decide whether you spend the next 12 months profitable or panicking. Most of the failure modes that show up at month six were already visible at week two — owners just didn't have the dashboard to see them.

This is the playbook we walk new DSPs through — built around the failure modes that show up again and again in first-month fleet operations. It's not the Amazon-issued onboarding checklist (you've already got that). It's the operational layer that sits underneath it — what to track, week by week, so you know whether you're on a healthy trajectory or sleepwalking into a margin trap.

Week one: instrument before you optimize

Week one is the worst week to make process changes. Drivers are still memorizing the depot. Vans are being assigned and reassigned. The Amazon scorecard is still showing baseline. The temptation is to micromanage. Don't. Spend week one putting instrumentation in place.

What "instrumentation" means in practice:

• A vehicle register that knows which van is on which route. Not a spreadsheet — a system where reassignment is one tap and the audit trail is automatic. If you can't trace a Tuesday parcel back to a specific van and driver in 30 seconds, you're not ready for week four.
• A consistent photo handoff before and after each shift. Establish the four exterior views, dashboard and cargo-area photos from the start. Use the shift handoff checklist to assign the capture and review steps.
• Prepare the mobile workflow before the first shift. Install Fleet by Elevera, check access and walk through the photo capture process during the briefing.
• A baseline incident-rate counter started. You don't have data yet, but starting the clock now means you have a real comparison point in week three. (See incidents per 100 shifts.)

The deliverable at end of week one isn't an optimization. It's a clean dataset starting Monday morning. Everything else this month is impossible without it.

Week one daily checklist for the dispatcher

07:00  – Confirm all drivers have Fleet by Elevera installed and logged in
07:15  – Verify vehicle assignments match the day's route plan
07:20  – Spot-check three handoff photo sets from yesterday
07:30  – Review any incidents flagged overnight
17:30  – Pull the day's completion rate, mark any open shifts
18:00  – Review post-shift photos for new damage

Five touchpoints. None of them takes more than 5 minutes. By Friday, the rhythm is automatic.

Week two: read the early signal

By week two you have ~10 days of clean data. This is the week where the patterns become visible, even though you don't yet have enough volume to act on them with confidence.

What to look at on Monday morning of week two:

1. Per-driver completion times. Not "did they finish" — that's the Amazon scorecard's job. Look at the spread. If your fastest driver finishes at 14:20 and your slowest at 17:50 on the same route mix, you have a coaching opportunity, a routing opportunity, or both. Don't act yet. Note it.

2. Pre-shift photo compliance rate. If you're below 90% by day 10, your workflow is broken. Either the photos are taking too long (>60 seconds) or the briefing was too soft. Fix it this week — every missed photo in week two becomes an unrecoverable claim loss in month two.

3. Incident clustering. Look at where incidents are happening — driver, vehicle, route, time-of-day. Two weeks isn't enough to draw conclusions, but it is enough to see if there's a single van or driver that's an obvious outlier. Mark them for week-three review.

4. Vehicle uptime. Any van that's been in the workshop more than once in 10 days is signalling something. Check the maintenance log — is it the same root cause? Different ones?

5. Driver feedback (informal). Walk the depot at 06:30. Ask three drivers what's slowing them down. Write it down. The answer is rarely what you expected. (A classic of the genre: the answer to "what slows you down" is "I can't find the keys" three weeks running. Solving that one costs €8 in key hooks.)

The goal of week two: you've identified your top 3 operational risks. You haven't fixed any of them yet. Resist the urge.

Week three: first interventions

Now you act — but on three specific things, not twelve.

The rule we give new DSP owners: pick the smallest intervention that addresses the largest risk. Not the most exciting one. The smallest.

Three categories of typical week-three interventions:

Category A: a single driver. If one driver's incident rate is 3× the team average, the intervention is a 20-minute one-on-one — show them their own data, ask what they're seeing, agree on one thing to change. Don't fire anyone in week three. You don't have enough data and you'll set the wrong tone with the team.

Category B: a single vehicle. If one van keeps coming back damaged, the intervention is to swap which driver is on it for a week. You're isolating the variable. If damage continues, it's the van. If it stops, it's the driver–van combination (often a sightlines issue).

Category C: a single time block. If most of your incidents are clustering between 16:00 and 18:00, that's not a fleet problem — it's a fatigue problem. Intervention: shift one of your routes 30 minutes earlier and re-measure.

You're running three small experiments in week three. By week four you'll know which ones worked.

Week four: lock in the operating rhythm

The first 21 days were diagnostic. Week four is where the operating rhythm gets locked in. This is the rhythm that should still be running unchanged at month six.

What "rhythm" means:

Monday 07:00 — weekly metric review. Review the incident rate with its underlying counts, vehicle availability and claim status. Use the cost analytics overview alongside the relevant incident and maintenance records.

Tuesday 09:00 — coaching block. 20 minutes per week with one driver. Not the worst, not the best. The middle. The middle is where the gains come from.

Wednesday 14:00 — vehicle review. Walk the lot with the maintenance log open. Anything overdue gets scheduled today, not next week.

Thursday end of shift — claims review. Check open incidents, missing evidence and the next action for each claim. Follow the reporting requirements and deadlines that apply to the specific case; do not wait for the weekly review when an earlier action is due.

Friday 17:00 — driver feedback loop. Three names from the depot, one informal question each. Write the answers down. Once a month, you'll find a €4,000-a-year process improvement hiding in those answers.

Five touchpoints, total time under two hours per week. That's the operating rhythm of a DSP that holds margin.

The dashboard you should have by day 30

If you've followed this playbook, by day 30 you should be able to answer all of these in under 60 seconds, on one screen:

• What's our incident rate per 100 shifts this week vs the rolling 4-week average?
• Which three drivers are above the team baseline, and by how much?
• Which van has had the most workshop time this month, and why?
• How many claims do we have open, what stage are they in, and what's the expected resolution date?
• What's our cost per shift trending toward this month vs last month?

Use those questions to check whether your records support the review. The cost analytics workflow covers the financial view; vehicle availability and open claims need their own underlying records.

What month two should look like

If month one ran clean, month two is when the gains start compounding. Per-100-shift rates start drifting down. Claim disputes start tipping in your favour. Vehicle uptime stabilises. The dispatcher stops fighting fires and starts running plays.

If month two doesn't look like that, almost always one of two things happened: instrumentation slipped (the photos stopped getting taken, the dashboard stopped getting reviewed) or the operating rhythm wasn't actually locked in. The fix in both cases is the same — go back to the week-four protocol and run it for two weeks straight.

If you are standing up a new route, review the Fleet by Elevera workflows for Amazon DSPs against your operating plan. Pricing uses fleet-size bands; check the current plans and included modules. Start a 14-day trial for one van — card required, nothing charged for 14 days, cancel any time.

---

Related reading on the Fleet journal:

• Incidents per 100 shifts: calculation and limits
• Photo evidence at handoff: the claims cost model
• Preventive vs reactive maintenance: a 12-month cost comparison