What national posts should steal from Amazon and Flipkart — and what they shouldn't

What national posts should steal from Amazon and Flipkart — and what they shouldn’t

By Soham Chokshi, CEO

National posts have been told for a decade they need to “become more like Amazon.” Most of that advice is wrong, but the small part that is right has been the difference between posts that modernized and posts that are being restructured by their treasuries. Here is the honest version.

What most CXOs believe

The dominant posture in national posts for the last ten years has been defensive modernization. Digitize the record layer, add track-and-trace, launch a parcel product line to capture e-commerce, partner with consumer marketplaces. The underlying belief has been that national posts can defend the letter-mail business while slowly ramping parcel, and that e-commerce volumes will lift all boats.

This has produced a recognizable class of mid-stage modernization: a decent mobile app, a functioning track-and-trace page, a parcel product that looks competitive on the surface, and an internal ops model that is essentially unchanged from 2010. The letter-mail subsidy has been quietly funding the parcel build-out, and for most posts the subsidy has run out or is about to.

The deeper belief underneath this — and this is where I think the industry has been strategically wrong — is that the posts can’t move as fast as Amazon or Flipkart because of political, labor, and regulatory constraints. That is partially true and mostly an excuse. The posts that have broken out — Post Luxembourg, Posti, Bpost, Qatar Post — did it by picking very specific operational playbooks from e-commerce logistics and refusing to adopt others. The winners understood that “become like Amazon” is the wrong goal. The right goal is to become the default parcel infrastructure for a national market, which is a different operating model.

What’s actually happening

The posts that have modernized successfully did three things, and refused to do two.

What they stole from e-commerce — AI-native routing and first-attempt-success obsession. Amazon and Flipkart treat every undelivered parcel as a data signal: address quality, time-window mismatch, customer preference, building access. They feed this into dynamic routing models that improve first-attempt-success meaningfully every quarter. Post Luxembourg, Posti, and Qatar Post have adopted versions of this discipline — Qatar Post runs at 90% FADR, and Post Luxembourg moved delivery-window adherence from ~30% to 90%+ after deploying AI micro-cluster routing. The posts that did not — who still route by static zone with fixed driver beats — run with first-attempt-success rates materially below the modernizing cohort, and that gap compounds into cost and churn.

What they stole — hybrid operating models. Amazon runs a mix of in-house fleet, gig drivers (Flex), contracted fleets, and locker/access-point networks. Flipkart runs a similar hybrid. The modernizing posts have embraced the same: Bpost’s hybrid model for urban Brussels, Post Luxembourg’s parcel-locker network, Qatar Post’s franchise-branch model for last-mile. Static employed-fleet-only operating models cannot compete on cost at e-commerce parcel volumes. The hybrid operating model is the single biggest strategic decision most posts have to make and most have not.

What they stole — the multi-carrier allocation layer. A modernized post is not a single-carrier operator. It is the orchestrator of a national network that includes its own fleet, partner carriers for specific lanes, international carriers for cross-border, and access-point networks for pickup. An agent-driven multi-carrier allocation layer — choosing the right carrier for each parcel in real time — is the unlock that makes a hybrid model operable at scale.

What they wisely refused to steal — the retail relationship. Amazon and Flipkart logistics serve an internal retail business. They optimize for the retailer’s economics. A national post that tries to be its own retailer usually fails — the merchant relationships, the working-capital model, and the product pricing disciplines are not native to a post. The winning posts stayed firmly in infrastructure mode: be the parcel rail for every retailer in the country, including ones that compete with each other.

What they refused to steal — the speed-at-all-costs posture. Amazon optimizes for ever-faster delivery because the retail P&L pays for it. A post that prices same-day at cost ends up subsidizing e-commerce margins out of the letter-mail base. The modernizing posts offered speed tiers — next-day default, same-day premium — but kept their pricing discipline. Posti and Bpost are good examples of this.

What to do in the next 90 days

Publish your FADR number internally and treat it as the KPI that drives everything. If your first-attempt-delivery rate is below 85%, you have a routing and access problem that no amount of volume growth will fix. The margin on a re-delivered parcel is negative. Set an 18-month target to move FADR into the 92%+ range. IKEA runs a 95% FADR in big-bulky retail. The posts who get there in parcel are the ones who survive.

Commission an honest cost-per-parcel audit against at least one modernized post. Benchmark your cost-per-parcel against a modernized peer such as Post Luxembourg or Posti. If the gap is material, you are likely running a pre-e-commerce ops model. The conversation after that audit — whether to hybridize the fleet, introduce parcel lockers, or reallocate route planning to an agent layer — is the conversation that matters.

Stand up an AI-native routing pilot on a single urban corridor. Not nationwide. One high-volume urban lane. Measure FADR and stops-per-hour against your existing routing. This gives you the internal proof point you need to make the national investment case. The modernizing posts all started with a single-corridor pilot and scaled.

Rewrite your partnerships charter. A post that is not actively signing partner-carrier and access-point network deals in 2026 is giving up ground to private CEP operators. The modernization playbook requires a partnership-first posture: you are the infrastructure orchestrator, and partners expand your capacity and coverage faster than in-house hiring ever will.

Pick a single agent use case — FADR rescheduling is the highest-leverage one — and deploy it end-to-end. An agent that rebooks a failed delivery based on customer preference, access-point availability, and driver capacity resolves exceptions that today queue into re-delivery cost. This is where AgentFleet (specifically Astra for operational orchestration and Clara for customer-facing communication) lands cleanest in a postal context.

Why this matters now

The regulatory clock is ticking on letter-mail cross-subsidies in most markets. Postal modernization was always going to be forced eventually — it is being forced now. The posts that have already crossed the modernization threshold are going to consolidate regional parcel share over the next five years. The posts that are still debating hybrid-fleet models will not exist in their current form by 2030.