Abstract
Amazon Flex has loud, recent driver pain around offers, routing, support, standings, battery, and app reliability, but the most valuable fixes sit inside Amazonβs controlled logistics system. A solo builder should not try to build a true replacement; the realistic opportunity is a modest paid companion for mileage, profitability, route evidence, and support-dispute documentation.
Background
Amazon Flex is Amazonβs free driver app for signing up, scheduling blocks, accepting same-day offers, navigating deliveries, and getting on-road support. It still shows a 4.6β lifetime average and 8,097,720 installs, so the issue is not lack of demand; it is that recent reviewers describe the app as unreliable at the exact moments that determine whether they get work or get penalized.
What it does
Lets drivers onboard, schedule or request blocks, scan packages, navigate stops, contact support, and track delivery standing.
What changed
Recent reviews mention security-check failures, update loops, instant-offer bugs, request-only scheduling complaints, and navigation regressions.
Why it is soft
Drivers are angry, but Amazon controls the jobs, pay, standings, and routing inputs. A new app can assist drivers, not replace the marketplace.
Strengths
The incumbent still has massive structural advantages: Amazon owns the delivery demand, driver eligibility, warehouse workflows, payments, support channels, and block inventory. Even several critical reviewers admit the flexibility, guaranteed pay, or side-hustle value can be useful, which means a solo competitor would be fighting a habit and supply network it cannot recreate.
Supply lock-in
Drivers open Flex because that is where Amazon blocks exist; a third-party app cannot manufacture those shifts.
Retained trust signal
A 4.6β lifetime average and 8,097,720 installs create store credibility a new app will not have on day one.
Still useful when working
Some reviewers still describe it as a good source of money or appreciate guaranteed payouts when the workflow behaves.
Market gap
The gap is real, but it is fragmented. The most common-looking pain is not a missing consumer feature; it is drivers feeling blind, underpaid, unfairly ranked, and unable to get human help when the app or warehouse workflow fails.
Blocks, bots, and access
Many reviewers say offers disappear instantly, bots or multiple phones win blocks, and request-style scheduling removes control from drivers.
Blind route economics
Drivers repeatedly ask to see mileage, distance to first stop, package count, route area, gas burden, and EV range fit before accepting.
Navigation and app reliability
Reviews cite wrong turns, lag, road-closure ignorance, private-property routing, battery drain, connectivity errors, and failed updates.
Support and standing penalties
Drivers say support gives generic answers while app, warehouse, access, safety, and customer-claim problems still damage their standing.
It's not good, and not much has improved in my 7yrs being an Amazon Flex driver. It's always buggy, and nowadays it's oversaturated with bots snagging all the blocks, so most blocks that DO show up are already reserved. It's pretty frustrating. Amazon Flex needs a better way to reserve blocks. All these years later and they've yet to work out a better system.
Something wrong with routing. First 10(of 40) stops scattered over entire route. First stop will be in a group of 20's, next stop will be 10+ minutes away in a group of teens. Until around stop 12-15 where the grouping makes more sense. This is K*LLING me in gas! Or the time it takes to reroute. Please fix!
Amazon Flex should show the total mileage before drivers accept a block so drivers can know if it is worth it or not. Many times it says 4 hours, but it takes more than 5 hours to finish. Also, the mileage is so high that the pay does not even cover gas expenses. This feels unfair to delivery drivers and wastes their time and money.
- Build a companion, not a block-sniping app. Offer automation is the tempting gap, but it is also the riskiest and least defensible surface for a solo builder.
- Start with mileage, route, and profit evidence. A manual or screenshot-assisted log can help drivers decide whether a block paid enough after gas, wear, distance home, and time overrun.
- Add a standing-dispute locker. Timestamped screenshots, check-in proof, package counts, return notes, and support-message exports address the support-and-penalty cluster better than another map clone.
Build complexity
Estimated, not derived: one developer, part-time, no funding, and no paid acquisition. This assumes a compliant companion app with manual entry, screenshot/photo capture, local logs, GPS trip tracking, and exportable evidence β not an app that logs into Flex, reserves blocks, or changes Amazon routing.
| Workstream | Weeks |
|---|---|
| Android shell, local storage, onboarding | 1.0 |
| Block, route, and package logging | 2.0 |
| Mileage, GPS trip tracking, and profit calculator | 2.0 |
| Evidence locker, exports, and support templates | 1.0 |
| Billing, Play listing, QA, and policy cleanup | 2.0 |
| Total | 8.0 |
Expected revenue
Estimated, not derived: assume a free basic log and a $4.99/month Pro plan for unlimited route history, exports, evidence bundles, and advanced profitability reports. With no paid acquisition, month-12 revenue is likely small because this app cannot promise more blocks or higher pay.
| Month 12, monthly | ||
|---|---|---|
| Installs | 1,500 | |
| Activated | 600 | 40% |
| Retained | 240 | 16% |
| Paid | 120 | 8% |
| Revenue | $600 |
Biggest challenges
The hard part is not CRUD screens; it is staying useful without touching Amazon-controlled surfaces. Drivers want more offers, better routes, fairer standings, and real support, but a third-party app can only document and advise around those problems.
Cannot control supply
No companion app can create Amazon blocks, raise base pay, reverse deactivations, or force support to respond.
Policy and trust risk
Anything that looks like botting, screen-scraping, or automated offer grabbing would be risky for users and for Play review.
Background location burden
Trip tracking means battery, privacy, permission, and accuracy work β exactly the area reviewers already punish harshly.
AI angle
There is a limited but genuine AI angle: AI can turn messy driver evidence into structured logs and appeal drafts. It cannot fix Amazonβs scheduling, routing, pay, or support decisions, so it should be positioned as documentation help, not a magic way to get more blocks.
Useful AI
Extract block time, package count, station notes, screenshots, mileage notes, and support-chat snippets into a clean dispute packet.
Not useful AI
Predicting which hidden route is profitable or defeating bots would require data and control the solo app does not have.
Conclusion
Skip it. Do not build a true Amazon Flex competitor; Amazon owns the marketplace, routing inputs, support outcomes, and payment system. If you still want to serve this audience, build the smaller companion app and expect side-project revenue, not a venture-scale wedge.
8 weeks
Time to v1
$4.99/mo
Pro companion plan
$600
Month-12 revenue
Amazon controls blocks
Deciding constraint