Start typing to search this publication.
techcirkle
Open menu

Subscribe to techcirkle

Get new posts delivered straight to your inbox.

Pods, Squads or Platform Teams? Topologies for Native and Cross-Platform Mobile

Compare pods, outcome squads and platform groups for a mobile app development team, and see how native or cross-platform choices change staffing.

techcirkle
5 min read
Cover image for Pods, Squads or Platform Teams? Topologies for Native and Cross-Platform Mobile
UI/UX designers and developers reviewing an app structure chart and screen wireframes at a meeting
UI/UX designers and developers reviewing an app structure chart and screen wireframes at a meeting

Conway's law applies to mobile with unusual force, since the org chart ends up inside your module graph, your release train and your API contracts. Deciding how to arrange a mobile app development team is therefore partly an architecture decision. Get it wrong and you will see the result as features that drift across platforms, duplicated networking layers, or a backlog stuck behind another team's priorities.

This post focuses on topology and stack-driven staffing. For the wider context, including roles, costs and sourcing, read our complete mobile app development team guide.

Settle the stack before the org chart

Job descriptions come after the technical approach, because each option has its own staffing signature.

Fully native: Swift and Kotlin

Expect at least one specialist per platform, and two per platform for coverage on anything business-critical. In return you get the deepest OS integration and the best performance for camera, Bluetooth, background processing and on-device machine learning. Duplicated work is the price, along with features that gradually diverge unless design and product are strong.

Cross-platform: Flutter or React Native

A smaller group can cover both stores, and engineers can rotate between features. The popular claim that this halves your cost is optimistic. Native modules, permissions, background behaviour and store rules still demand platform knowledge, so staff at least one engineer with real iOS or Android experience.

Kotlin Multiplatform: shared logic, native UI

This is a reasonable middle path for groups that already have native strength and want less duplicated business logic. Build complexity goes up, so it suits teams where a strong senior engineer will explicitly own that complexity.

How AI shifts the comparison

Generated code makes translating a feature between Swift and Kotlin cheaper than it was, which softens the pain of native duplication. Review, test and release effort are still doubled, however. AI narrows the gap between approaches without removing the trade-off.

Topology 1: the single cross-functional pod

Pod
 ├── Product manager
 ├── Designer
 ├── Mobile engineers
 ├── Backend engineer
 └── QA + release owner

One roadmap, one standup, direct communication, minimal process. This is the right default up to about eight people, and most teams split too early. A pod's limit shows up when two distinct user outcomes begin competing for one roadmap, and every planning session becomes a negotiation over which one wins.

Topology 2: outcome squads with a platform group

Squad A (acquisition)     Squad B (core product)
 mobile, backend,          mobile, backend,
 design, QA, PM            design, QA, PM
        \                       /
         Mobile platform group
 design system | networking | build tooling | release train

Each squad owns a user outcome end to end, while a small platform group maintains shared components. Teams commonly reach for this around eight to ten people, and it becomes a formal, dedicated group beyond about 15.

The failure mode is cultural. A platform group that treats squads as customers multiplies everyone's speed. One that behaves as a gatekeeper becomes a new queue with a roadmap of its own. Some practical guardrails:

  • Measure the platform group by squad lead time, not by the number of components shipped.

  • Let squads contribute to shared modules through pull requests rather than waiting for tickets.

  • Version the design system and networking layer so that an upgrade does not stop a squad's work.

  • Keep the release train owned in one place, with squads deciding what boards it.

Topology 3: iOS team and Android team

Software engineers at side-by-side desks collaborating with a teammate holding a laptop in a modern office
Software engineers at side-by-side desks collaborating with a teammate holding a laptop in a modern office
iOS team  <--> Backend team <--> Android team
       \_______ shared backlog _______/

You will find this in older organizations, and it tends to be the least effective shape. Every feature requires coordination between two client teams and a backend team, which means three queues and frequent drift between platforms.

If you have inherited this structure, you can migrate gradually. Pick one user outcome, embed an iOS and an Android specialist alongside backend and design, and let that squad run a full cycle. Keep platform chapters, or regular guild-style meetings, for craft standards. Moving toward squads with embedded platform specialists usually speeds up delivery noticeably.

The rule that applies to all three

Keep the app and the API it depends on close together, ideally on the same team. Many slow mobile organizations are not short of mobile engineers; they are waiting on a separate backend team with different priorities. If you cannot merge the teams, at least give the mobile group a clear contract and a way to get endpoint changes prioritized. Without that, engineers tend to build fragile logic into the client to work around the wait.

Size bands to plan with

These assume a typical app with a backend, so adjust for regulation and integrations.

  • 3 to 5 (MVP): one pod, one or two senior mobile engineers, shared QA, partial design.

  • 5 to 8 (post-launch): still one pod, plus dedicated QA and a full-time PM. Native teams add a second platform specialist.

  • 8 to 15 (scale-up): usually two squads with shared platform capability, an analytics owner and a formal release manager.

  • 15 and up (multi-product): a dedicated platform team, with feature squads building on top.

Verifying your topology with data

Assess the shape by looking at where work waits. A long review queue suggests too few senior reviewers per squad, and this is worth watching closely with AI assistants raising PR volume. Long lead time paired with short queues suggests handoffs between groups. Duplicate components across squads suggest a platform group that is too small, or one that squads avoid.

If you are reshaping a team or staffing a new build, our mobile app development engineers can help you pressure-test the plan.

Frequently Asked Questions

What is the difference between a pod and a squad?

Usage varies. Here, a pod is a single cross-functional team working on one roadmap, while squads are multiple cross-functional teams that each own a user outcome and share a platform group.

When does a mobile team need a platform group?

Once several squads start rebuilding the same components and release tooling. A small shared capability can begin in the scale-up band, with a dedicated team usually arriving beyond about 15 people.

Why are separate iOS and Android teams usually slower?

Each feature needs coordination across two client teams, often plus a backend team. That adds handoffs and makes it easier for platforms to drift apart.

Does cross-platform remove the need for native engineers?

No. Native modules, permissions, background behaviour and store rules still require platform knowledge, so keep at least one engineer with real native experience.

Is Kotlin Multiplatform a good choice for a new team?

It fits best where native strength already exists and a senior engineer will own the added build complexity. A new team without that person may find it a heavy start.

Should backend engineers sit inside the mobile squad?

Ideally, yes. When the app and its API are owned by people working closely together, mobile engineers spend less time blocked by another team's priorities.

Subscribe to techcirkle