What is the absolute minimum for an MVP?

Auth (one method), the core user flow end to end, payments if you monetize, basic push for transactional events, baseline analytics (~12 events), Sentry, and accessibility basics. Everything else is a defer.

Should I include chat in my MVP?

Almost never. Chat is a multi-month feature surface. Unless chat IS your product, use email or Intercom for support and ship the rest of the MVP on time. You can add chat as a focused follow-up release.

Do I need an admin panel for launch?

No. Use SQL access, Retool/Forest Admin, Postman collections, and scripts for the first 500 users. Build a focused admin only when a specific ops workflow becomes a weekly bottleneck. Building admin upfront usually costs 2–4 weeks you do not have.

How do I handle analytics in an MVP?

Mixpanel, Amplitude, or PostHog plus Sentry. Instrument ~12 well-chosen events covering signup, activation, key action, and retention markers. Skip custom data warehouses, BI tools, and Segment until your data volume genuinely justifies the complexity.

What about accessibility — can I add that later?

Do the basics now. VoiceOver/TalkBack labels, proper contrast, 44pt tap targets, and dynamic type are quick wins that are far cheaper to do during build than to retrofit. Save the formal WCAG AA audit for post-MVP if you need it.

Mobile App MVP Features: What to Include and What to Skip

Introduction

Scope discipline is the single biggest determinant of MVP success. Every additional feature increases cost, delays launch, multiplies QA surface, and delays the one thing that matters most — user feedback on your core hypothesis. This guide walks through each common mobile feature category and gives you a clear include/skip recommendation.

Authentication — what to include

Include: email/password OR social login (Google, Apple Sign-In). Not both. Password reset flow. Session management. Apple requires Sign-In with Apple if you offer any social login on iOS.

Skip: multi-factor auth (unless regulated), SSO/SAML, team/organization management, email verification gating (use soft verification), magic links on top of other methods.

Include: one auth method + reset + session management
iOS: Sign-In with Apple required if you have any social login
Skip: MFA, SSO, team management, verification gating

Payments — what to include

Include (if your product monetizes in MVP): Apple Pay / Google Pay for one-time purchases, Stripe for subscriptions if you have web parity, or StoreKit/Play Billing for subscriptions sold through the App Stores. Never store card data yourself.

Skip: complex tiered pricing, coupon codes (unless central to pricing strategy), invoicing, multi-currency (launch in your primary market's currency), tax calculation beyond what Stripe/Apple handle.

Include: Apple Pay / Google Pay / StoreKit / Stripe; never store cards
Skip: tiers, coupons, invoicing, multi-currency, complex tax

Push notifications — what to include

Include: push permission request after the user has completed onboarding (never before), basic transactional notifications (new messages, order updates), OneSignal or Firebase Cloud Messaging as the provider.

Skip: sophisticated segmentation (wait for 1k+ users), rich media notifications, in-app messaging UI, geo-triggered notifications, A/B tested push copy.

Include: basic transactional push via OneSignal or FCM
Ask for permission after onboarding, not during
Skip: segmentation, rich media, geo-triggers, push A/B testing

Chat and messaging — usually skip

Building chat in an MVP is almost always wrong. Real-time sync, typing indicators, delivery receipts, media uploads, notification integration, and moderation are each multi-week projects. Unless chat IS your product, skip it.

Credible alternatives: email-based threads for async communication, Intercom or Crisp for customer support chat, Stream or Sendbird if you genuinely need embedded chat.

Chat is a multi-month project — skip unless it IS the product
Use Intercom/Crisp for support, Stream/Sendbird if embedded chat is needed

Analytics — include the basics only

Include: Mixpanel, Amplitude, or PostHog for product analytics; Sentry for errors; a dozen well-chosen event names covering the core funnel (signup, activation, key action, retention markers). That's it.

Skip: custom data warehouse, BI dashboards, attribution tooling beyond a basic UTM setup, sophisticated cohort analysis, Segment (until volume justifies it — start with direct SDK integration).

Include: Mixpanel/Amplitude/PostHog + Sentry + ~12 well-chosen events
Skip: custom warehouse, BI dashboards, heavy attribution, Segment until volume

Admin panel — SKIP

Admin panels are entire separate products that founders routinely smuggle into MVPs. Do not build one. Use: direct SQL access for the founder, Retool or Forest Admin for ops tools, Postman collections for one-off actions, scripts for repeated actions.

If you hit 500+ active users and operations start to slow down, then build a focused admin for the specific actions your team does weekly. Never start there.

Skip admin panel in MVP
Use: SQL, Retool/Forest, scripts, Postman
Build focused admin at 500+ users for specific slow actions

Offline mode — usually skip

Offline mode is expensive. Conflict resolution, sync queues, offline UI state, and edge cases add weeks of engineering. Skip unless your users are genuinely offline for extended periods (field workers, travel apps). Most consumer and B2B apps do not need it.

Offline adds weeks for conflict resolution and sync
Skip unless users are genuinely offline (field, travel)

Accessibility — INCLUDE the basics

Include: VoiceOver/TalkBack labels on interactive elements, 4.5:1 color contrast, minimum 44x44pt tap targets, dynamic type support for iOS. These are quick wins that prevent later rework and broaden your user base.

Skip (until later): full WCAG AA audit, screen reader flow optimization for complex UI, reduced motion support (pleasant to have but rarely launch-blocking).

Include: VoiceOver/TalkBack labels, 4.5:1 contrast, 44pt tap targets, dynamic type
Skip (until later): full WCAG AA audit, advanced screen-reader flows

Conclusion

A disciplined MVP ships fewer features better. The categories above cover roughly 80% of the 'should I include this?' conversations we have with founders. When in doubt, cut. You can always add the feature in the second sprint with real user signal; you cannot get back the weeks you spent building something users did not need.