Industry Research

Last updated June 2026 Every figure sourced & cited

App development statistics for 2026, and what they mean if you're shipping one.

The state of building apps in 2026 — by the numbers. The app economy, what a build actually costs and how long it takes, why most apps fail, where AI fits, and what it takes to stay live. Every figure below is dated and cited; nothing is estimated.

Compiled by AppMakers USA from primary sources (Sensor Tower, Apple, Google, Stack Overflow, Statista and others). Full citations at the bottom.

$167B

consumer spend in apps, 2025

Sensor Tower

150B

app downloads in 2025 — a record

Sensor Tower

3.6h

a day per user, spent in apps

Sensor Tower

94%

of users churn within 30 days

AppsFlyer benchmark

How to read this page. Figures marked typical ranges are agency-consensus estimates — no authoritative primary survey exists for app cost or build time, so they're shown as ranges, not precise numbers. A few load-bearing benchmarks are older than 2023 and are dated-flagged where they appear. Everything else is a current primary or named-primary figure. Sources are listed in full at the bottom — built to be cited.

01The App Economy

The app economy is still growing — and it runs on attention.

$167B

Consumer spend across iOS and Google Play in 2025, up 10.6% year over year.

Sensor Tower · 2026 State of Mobile (2025 data)

~150B

App downloads in 2025 — a new record, surpassing the prior peak.

Sensor Tower · 2026 State of Mobile

5.3T

Hours spent in apps globally in 2025, up 3.8% year over year.

Sensor Tower · 2026 State of Mobile

~3.6h

Spent in apps per mobile user, every day, in 2025.

Sensor Tower · 2026 State of Mobile

~70%

Of all app consumer spending happens on iOS — despite Google Play taking far more downloads.

Business of Apps · App Revenues (2025)

$83.6B > $83.2B

For the first time ever, non-game apps out-earned games in consumer spend in 2025.

Business of Apps (2025)

3.8B

Generative-AI app downloads in 2025 — double 2024; their in-app revenue tripled past $5B.

Sensor Tower · 2026 State of Mobile

The app economy has a split personality: Android distributes, iOS earns

Google Play takes nearly three-quarters of downloads, but the App Store collects roughly 70% of the money — iOS skews toward higher-income markets where users spend far more per install.

Share of consumer spending · 2025

App Store (iOS)

70%

Google Play

30%

Share of downloads · 2025

Google Play

73.5%

App Store (iOS)

26.5%

Source: Business of Apps — App Revenues & App Statistics (2025 data)

What shipping right looks like

This is the economy we build in. The products we make are designed to earn a real place in it — Number Hive reaching 700K+ students, ShiftPass live and rated on both stores. See what we've built.

Browse case studies →

02What It Costs To Build

There's no single "app cost" — only ranges, and what drives them.

$5K–$150K+

Simple apps ~$5K–$60K; medium-complexity ~$50K–$150K; complex or enterprise builds run $400K and up.

Typical industry ranges · agency consensus

$120–$250/hr

Typical billing rate for senior US mobile developers.

Typical industry ranges · agency consensus

$25–$80/hr

Offshore rates — ~$25–$55/hr (India), ~$40–$80/hr (Eastern Europe).

Typical industry ranges · agency consensus

40–70%

Cost reduction offshore teams can offer versus US-based — though quality and rework vary widely by process.

Typical industry ranges · agency consensus

What shipping right looks like

The cheapest quote isn't cheapest if the app ships twice. A surprising share of our work is rebuilding apps that came in low and cracked under real use — we'd rather scope it honestly once.

See app rescue →

03How Long It Takes

Most real apps take months, not weeks — and the plan is the tell.

4–8 months

Where most app builds land; the full range runs ~3–12 months by complexity.

Typical industry ranges · agency consensus

2–4 mo

A simple app; professional-grade ~4–9 months; highly complex 12 months and up.

Typical industry ranges · agency consensus

1–2 mo

Planning, before a line of product code — then 3–6 months building, 1–2 months pre-launch.

Typical industry ranges · agency consensus

What shipping right looks like

A realistic timeline is the first sign of a serious team. We scope the planning phase as real work, not a formality — it's why builds like the iPermit contractor portal hold up in production.

Read the iPermit build →

04Failure & Retention

Building the app is the easy part. Getting people to keep it is the war.

~24%

Average Day-1 retention across apps — roughly three in four users don't return the next day.

AppsFlyer benchmark · via Business of Apps

~5–6%

Average Day-30 retention — meaning ~94% of users have churned within a month.

AppsFlyer benchmark · via Business of Apps

~30% / ~11%

Fintech retains best — ~30% Day-1 and ~11–12% Day-30; shopping apps sit far lower (~24% / ~6%).

Industry retention benchmarks (aggregated)

The retention cliff: where the average app's users go in 30 days

Most loss is immediate — roughly three in four users never return after day one. By day 30 only about one in twenty remains. The drop is driven by weak onboarding, slow or buggy first sessions, and unclear value, not by price.

Day 1

~24%

Day 7

~10%

Day 30

~5–6%

Bars scaled to the highest value for readability; figures are the cited cross-app averages.

Source: AppsFlyer retention benchmarks, via Business of Apps (2024–2025)

How we approach it

Retention is engineered before launch, not chased after. When the product is fast and just works, people stay — which is why we sweat the hard parts, like the live-map flicker we hunted down and fixed in ShiftPass.

See the engineering story →

05Native vs Cross-Platform

Flutter and React Native lead cross-platform — but "lead" is smaller than the blogs claim.

A common myth puts cross-platform "market share" near 50%. The real, sourced figure is developer usage — and it's far lower.

9.4%

Of all surveyed developers did extensive work with Flutter in the past year — narrowly the most-used cross-platform framework.

Stack Overflow 2024 Developer Survey

8.4%

Used React Native — second among cross-platform frameworks, just behind Flutter.

Stack Overflow 2024 Developer Survey

Top 2

Flutter and React Native are the two most-used cross-platform frameworks among developers, ahead of all others.

Stack Overflow 2024 Developer Survey

~47.2M

Developers worldwide as of early 2025 (including part-time) — ~50% more than three years earlier.

SlashData · global developer population (2025)

~$35B

Projected low-code/no-code platform market by 2030, from ~$6.8B in 2022 (~23% CAGR) — it speeds simple apps, not complex ones.

Grand View Research

Cross-platform "dominance" is smaller than the blogs claim

This is developer usage — the share of all surveyed developers who did extensive work with each framework in the past year — not "market share." Flutter narrowly leads React Native; everything else trails.

Flutter

9.4%

React Native

8.4%

Bars scaled for readability; values are % of all surveyed developers (usage, not market share).

Source: Stack Overflow 2024 Developer Survey

How we approach it

We pick the stack for the product, not the trend. ShiftPass ships and performs on both iOS and Android — the right choice is the one that holds up live, not the one that's loudest this year.

See our builds →

From our build log

We've paid for the wrong framework call — so now we choose deliberately.

On media-heavy builds we spent months stabilizing React Native's video stack — init races, mid-playback crashes, state desyncs — and now steer streaming-first apps away from it. On another build we went Flutter against the React Native default for its render performance and shorter build times. The stack follows the product.

Anonymized from AppMakers' internal engineering log · client builds

06AI in App Development

Almost every developer now uses AI — and trusts it less than they did.

84%

Of developers are using or planning to use AI tools in their workflow — up from 76% a year earlier.

Stack Overflow 2025 Developer Survey

51%

Of professional developers now use AI tools every day.

Stack Overflow 2025 Developer Survey

~33%

Trust the accuracy of AI output — while ~46% actively distrust it. Adoption is up; confidence is down.

Stack Overflow 2025 Developer Survey

55%

Faster task completion for developers using an AI coding assistant in a controlled experiment.

GitHub / Microsoft Research2022 study

AI adoption is near-universal — but trust is sliding the other way

84% of developers now use or plan to use AI tools, yet only about a third trust their accuracy and favorability has fallen from 70%+ to ~60%. Adoption is up; confidence is down — which is exactly why human review still ships the product.

84%use AI

Using or planning to use AI tools84%

Not using / not planning16%

…yet only ~33% trust the output, vs ~46% who distrust it.

Source: Stack Overflow 2025 Developer Survey

What shipping right looks like

We build AI into client products — and into our own. Echo Journal is our voice-based AI journaling app. AI accelerates the work; judgment and testing are still what ship a product worth keeping.

Talk about an AI build →

From our build log

We engineer AI for production, not demos — and pick the model for the job.

On an EdTech build we run a model-per-job pipeline: one model generates and grades reading-level definitions, another scores whether images match — behind a model-agnostic layer so we can swap or upgrade without re-plumbing. On a document-parsing build we deliberately chose a cheaper, schema-constrained model for a bounded task and hard-gated cost per request. We benchmark models against the work; we don't chase the loudest one.

Anonymized from AppMakers' internal engineering log · client builds

07App-Store Reality

Roughly 4.8 million apps are live — and the stores reject millions more.

~2.39M / ~2.37M

Apps live on the Apple App Store and Google Play respectively (live counts).

42matters · live counts, June 2026

7.7M

App Store submissions Apple reviewed in 2024 — and rejected more than 1.9 million of them.

Apple Newsroom · transparency report (2024)

146K+

Developer accounts Apple terminated for fraud in 2024; 400K+ submissions rejected for privacy violations.

Apple Newsroom · transparency report (2024)

2.36M

Policy-violating apps Google blocked from Google Play in 2024, plus 158K+ developer accounts banned.

Google · Android Security Blog (2024)

1.75M+

Policy-violating apps Google prevented in 2025; over 92% of harmful-app reviews are now AI-assisted.

Google · Android Security Blog (2025)

What shipping right looks like

Getting in is the easy part — staying compliant and live is the work. ShiftPass cleared review and is live and rated on both stores; we build to the guidelines, not around them.

See a live, dual-store app →

08Security & Privacy

Most apps are far leakier than their owners realize.

Across 500,000+ assessments, the picture is consistent — and it's usually not malice. Most teams simply don't know what their own third-party dependencies collect or how weakly data is protected.

75%

Of iOS apps tested carried both sensitive data and tracking domains (since Aug 2025).

NowSecure privacy research (2025)

~⅔

Of apps used broken or weak encryption; ~1 in 5 had hardcoded encryption keys.

NowSecure · 525K+ assessments (2025)

97%

Of iOS apps were missing required Privacy Manifests for their third-party SDKs.

NowSecure (2025)

400K+

App submissions Apple rejected for privacy violations in 2024; $9B+ in fraud blocked over five years.

Apple Newsroom (2024)

What's actually inside the average mobile app

Share of assessed apps carrying each risk. The gap between "it works" and "it's safe" is exactly the maintenance-and-audit work that separates a professional build from a cheap one.

iOS missing privacy manifests

97%

iOS: sensitive data + tracking

75%

Weak / broken encryption

~66%

Source: NowSecure mobile app risk research (2025)

From our build log

On a regulated medical build, encryption couldn't be an afterthought.

For an FDA-class patient app, the obvious offline database had no at-rest encryption — unacceptable for medical data. Rather than swap libraries and lose its sync engine, we forked the database and patched encryption in at the native layer, keeping the offline sync intact. The fix was open-sourced. When the data is sensitive, security is an architecture decision, not a checkbox.

Anonymized from AppMakers' internal engineering log · client build (in development)

How we approach it

If your app handles sensitive data, the encryption and privacy story has to be engineered in from day one — not bolted on before a store review. That's the kind of build we take seriously.

Build it right →

09How People Use Apps

People live in apps — but their attention concentrates on a handful.

3h36m vs 18m

US adults spend 3h36m a day in apps versus 18 minutes in mobile browsers — an ~88% / 12% split.

eMarketer (2024)

~18

Different apps a typical US smartphone user opens in a day.

RealityMine · panel data

~½ on 6

Nearly half of all app time is concentrated on just six apps — the bar for staying installed is brutal.

RealityMine · panel data

~59%

Of global retail e-commerce now happens on mobile (~$2.5T in 2025); the US trails at ~44%.

eMarketer / Statista (2025)

81% / 51%

Median push-notification opt-in on Android vs iOS — consent rules structurally cap how many users you can even reach.

Business of Apps / Airship

Mobile time lives in apps, not the browser

Of US adults' mobile-internet time, roughly 88% is spent inside apps versus ~12% in mobile browsers — 3h36m a day in apps against just 18 minutes on the web.

88%in apps

Time in apps~88%

Time in mobile browser~12%

Source: eMarketer (Insider Intelligence), 2024 — US adult mobile-internet time

How we approach it

Earning a place in someone's daily six is the real goal. We build products people come back to — like Number Hive, used by classrooms of students at national scale.

Read the Number Hive story →

10Outsourcing & Failed Projects

A $634-billion outsourcing market — and most software projects still don't fully succeed.

$634B

Projected size of the worldwide IT-outsourcing market in 2026, growing ~6.2% a year toward $807B by 2030.

Statista Market Forecast (2026)

31%

Of IT projects succeed outright; ~50% are "challenged" (late, over budget, or under-scope) and ~19% fail.

Standish Group CHAOS2020, latest edition

~$53B → $65B

Custom software development market, 2025 to 2026 — compounding ~22% a year as businesses outgrow off-the-shelf tools.

Precedence Research

Most software projects don't fully succeed

Barely three in ten projects land on time, on budget, and on scope. Half come in "challenged" and roughly one in five fail outright — usually on scope and communication, not raw engineering.

31%

50%

19%

Succeeded — 31%

Challenged — 50%

Failed — 19%

Source: Standish Group CHAOS (2020, latest published edition)

What shipping right looks like

Most projects don't fail at the code — they fail at the scope and the handoff. We take over stalled and offshore builds, and rebuild operations that off-the-shelf tools couldn't hold, like the iPermit portal.

See how we rescue apps →

From our build log

We've inherited the codebases that landed in the "challenged" column.

On one rescue we took over a buggy, unstable codebase from a previous team and stabilized it under a tight budget so it could even run reliably. The honest lesson we now give every client: decide fix-vs-rebuild upfront — trying to do both on a thin budget is how a solid idea stalls.

Anonymized from AppMakers' internal engineering log · takeover/rescue build

11After Launch: Maintenance

The build is the cheap part. Keeping it alive is most of the cost.

60–80%

Of a software product's total lifecycle cost goes to maintenance, not the initial build.

IEEE / lifecycle-cost literature

15–20%

Of the original build cost is a typical annual maintenance budget, every year after launch.

Gartner / Forrester-cited benchmark

55–80%

Of IT budgets go to maintaining existing systems rather than building new ones.

Gartner (via secondary)

Where the money really goes over a product's life

The initial build is the small slice. Most of what a software product costs over its lifetime is everything after launch — updates, fixes, dependency upkeep, and keeping it compatible with new OS releases.

60–80%maintenance

Maintenance / post-launch~60–80%

Initial build~20–40%

Source: IEEE / software-lifecycle-cost literature (ring shows the ~70% midpoint)

What shipping right looks like

This is the number most agencies don't plan for. We don't build and walk away — we stay to run what we ship, the way the iPermit portal still runs a national permitting operation today.

See a product we still run →

From our build log

On our own app, we treat upkeep as routine engineering — not "future cleanup."

On Echo Journal — our own voice-based AI journaling app — we moved the framework across nine major versions, deliberately one step at a time rather than one big jump. Each increment surfaced small, isolatable conflicts; a multi-year catch-up migration would have compounded them. Continuous maintenance is why the apps we run keep running.

From AppMakers' internal engineering log · Echo Journal (AppMakers-owned)

★AppMakers, by the numbers

We've spent years building the apps behind these statistics.

400+

apps shipped across iOS, Android & web

verified client reviews on Clutch

10+

globally recognized brands we've built for

app we own end-to-end — Echo Journal, our voice-based AI journaling app

Who we build for

WalmartFREENOWTeradataBayutitalkiTrillerEurosportbeIN SPORTSRightmoveNaked Insurance

Public AppMakers figures (appmakersla.com · Clutch). "Who we build for" reflects brands from AppMakers' published client portfolio — work delivered for these companies, not ownership of their products.

—The Takeaway

The numbers say most apps fail. Ours ship, and stay running.

The data on this page is the industry baseline — high churn, blown timelines, projects that don't fully land. Here's what beating it looks like in production.

Number Hive

700K+ students reached

EdTech engineered to scale through its own growth instead of buckling under it.

Read it →

ShiftPass

5.0★, live on iOS & Android

A two-sided shift marketplace on a live map — and the hard live-map bug we hunted down.

Read it →

iPermit · ARCXIS

Live across hundreds of US cities

A B2B contractor portal that bends to permit work — built, and still run by us.

Read it →

Want to be on the right side of these numbers?

Whether you're building from scratch or rescuing a stalled one, we scope it honestly, build it to last, and stay to run it with you.

Get a Free Project Estimate →

Sources & methodology

Every figure was read on its source — primary reports where they exist, named-primary aggregators otherwise. Cost and timeline figures are agency-consensus ranges (no authoritative primary survey exists). Last updated June 2026.

2026

Sensor Tower — 2026 State of Mobile (2025 data). sensortower.com

2026

42matters — live App Store & Google Play app counts. 42matters.com/stats

2025

Apple Newsroom — App Store transparency report (2024 data). apple.com/newsroom

2026

Google — Android Security Blog, keeping Google Play safe (2025). blog.google/security

2025

Google — Android Security Blog (2024). blog.google/security

2025

Stack Overflow — 2025 Developer Survey (AI). survey.stackoverflow.co/2025

2024

Stack Overflow — 2024 Developer Survey (technology). survey.stackoverflow.co/2024

2025

AppsFlyer retention benchmarks, via Business of Apps. businessofapps.com

2026

Statista — IT Outsourcing market forecast, worldwide. statista.com

2024

eMarketer — mobile time spent, app vs browser. emarketer.com

2022

GitHub / Microsoft Research — Copilot productivity experiment (dated). github.blog

2020

Standish Group — CHAOS report, project outcomes (latest published edition). opencommons.org

—

Cost, rate & timeline ranges — agency-consensus estimates (Topflight, LitsLink, SPDLoad and similar build guides); presented as ranges, no primary survey exists.

2024

IEEE lifecycle-cost literature & Gartner-cited maintenance benchmarks (via Vention, Savi). ventionteams.com

2025

Business of Apps — App Revenues & App Statistics data hubs (iOS/Android split, categories). businessofapps.com

2025

NowSecure — mobile app security & privacy research (500K+ assessments). nowsecure.com

2025

SlashData — global developer population trends. slashdata.co

2025

Precedence Research — custom software & enterprise mobility market forecasts. precedenceresearch.com

2025

Grand View Research — low-code development platform market. grandviewresearch.com

2025

eMarketer / Statista — mobile-commerce share of retail e-commerce. statista.com

—

Airship push-notification opt-in benchmarks, via Business of Apps. businessofapps.com

—

AppMakers USA — portfolio & engineering log (the "build log" receipts; anonymized client builds, public figures via appmakersla.com & Clutch).