LiveOps Platforms Compared: Metaplay, PlayFab, Satori, Beamable, and Unity

This comparison looks at how five game backends (Metaplay, PlayFab, Satori, Beamable, and Unity Gaming Services) handle the four jobs that make up day-to-day live operations: A/B testing, remote configuration, live events, and player segmentation. PlayFab and Satori carry the most built-in experimentation analysis and ML: PlayFab runs on Microsoft's internal stats engine, and Satori runs a Databricks pipeline. Unity layers a LiveOps surface onto Remote Config for Unity-native teams, and Beamable covers the same jobs for Unity and Unreal studios.

Metaplay ships the LiveOps dashboard, server, and game logic as source code you own, so experiments and events run server-side inside your own code, with no built-in ML and analytics streamed to your own warehouse. Where Satori and PlayFab provide churn models and segmentation ready-made, Metaplay puts the logic in source you extend.

Quick comparison of LiveOps features by platform

Category	Metaplay	PlayFab	Satori	Beamable	Unity
A/B Testing	Config-variant experiments; runs in your source	Experiments; up to 9 variants; 95% confidence	Multi-phase experiments; variant allocation	Trials; percentage or criteria bucketing	Game Overrides; statistical significance
Remote Config	Game Configs; any C# structure; source access	Title Data; key-value	Feature Flags; boolean and multivariate	Content; real-time updates	Remote Config; key-value and JSON
Live Events	Calendar + cron scheduling; config or dashboard	Scheduled Tasks; CloudScript triggers	Time-based events; presets + CRON	Schedule builder; events + announcements	Via Remote Config; no dedicated scheduler
Segmentation	Custom rules in source; any player attribute	Segments API; properties and behaviors	Audiences; properties and computed values	Rules-based; private stats	Audiences; behavior and attributes
Dashboard	Included; full source for customization	Game Manager; polished, not customizable	Satori Dashboard; separate from Nakama	Beamable Portal; functional	Unity Dashboard; unified
AI & ML	Rules/config experiments; no ML; export to your warehouse	Churn-risk segmentation (Azure ML); AI chat moderation	Databricks ML (churn/LTV) → segments/A-B; GenAI	None documented	Safe Text/Voice ML moderation
Pricing	Included in all tiers	Included; ADX extra for analytics	Separate product; from $600/month	Included in tiers	Usage-based, free tier

Platform overview: owner, deployment, and primary use case

Platform	Owner	Deployment Model	Primary Use Case
Metaplay	Independent (founded in 2019 by people with backgrounds in the game and software industries)	Metaplay Cloud (managed) or self-host	Mobile F2P, server-authoritative games
PlayFab	Microsoft (acquired 2018)	Azure SaaS	Cross-platform games, Xbox/PlayStation titles
Satori	Heroic Labs	Heroic Cloud or self-host	LiveOps for Nakama users or standalone
Beamable	Independent	Beamable Cloud (AWS)	Unity and Unreal games
Unity Gaming Services	Unity Technologies	Unity Cloud	Unity-native projects

A/B testing and experimentation

Feature	Metaplay	PlayFab	Satori	Beamable	Unity
Variant model	Config variants per audience	Up to 9 variants	Multi-phase variants	Cohorts via stat/percentage	Rollout % per group
Statistical analysis	Via analytics export	95% confidence, SRM detection	Goal metrics, CSV export	Basic reporting	Significance tracking (no targeting)
Multi-phase experiments	No	No	Yes	No	No
Audience targeting	Segments, percentage, capacity	Segments, percentage	Audiences, probabilistic	Private stats, percentage	Rules-based
New player targeting	Yes	Yes	Yes	Via stats	Yes
Tester/preview mode	Yes	Yes	Yes	No	No
Source code access	Yes	Client SDKs only	No	No	No

PlayFab and Satori differ in how their experimentation works. PlayFab Experiments runs on the same internal platform behind Minecraft and Office, supports up to nine variants, and ships Sample Ratio Mismatch detection at 95% confidence to catch a treatment that crashes clients or drops logging. Concurrent override experiments need mutually exclusive audiences to avoid client-side conflicts.

Satori is the only platform here with multi-phase experiments: a player who lands in a variant during a phase stays locked to it regardless of later audience changes, which keeps variant assignment stable across the phase. Unity's Game Overrides randomly bucket players and track significance, but targeting and significance tracking can't both run on the same test. Metaplay's experiments are config-variant overrides that execute server-side inside game logic you own and can extend, and results are analyzed in your own warehouse rather than a built-in stats panel.

Remote configuration and over-the-air updates

Feature	Metaplay	PlayFab	Satori	Beamable	Unity
Config format	Any C# class (IGameConfigData<>)	Key-value	Feature flags, configs	JSON/YAML content	Key-value, JSON
Hot reload / OTA	Yes	Yes	Yes	Yes	Yes
Segment targeting	Yes	Yes	Yes	Yes	Rules-based
Version history	Yes	Yes	Yes	Yes	Yes
Dashboard editing	Yes	Yes	Yes	Yes	Yes
Source code access	Yes	Client SDKs only	No	No	No

This area divides into key-value stores and typed config. PlayFab Title Data, Unity Remote Config, and Satori feature flags all push key-value or JSON to clients at runtime, which covers tuning numerical values.

Metaplay is the typed end of that spectrum. Configs are built from Google Sheets or CSV into binary archives published over-the-air, and any C# class works as long as it implements IGameConfigData<>, so a config change is a strongly typed object the client and server both compile against rather than a JSON blob. Beamable sits between the two: you can edit content values remotely through the Portal, but adding or removing content types still needs a game update.

Live events and scheduling

Feature	Metaplay	PlayFab	Satori	Beamable	Unity
Scheduling	Calendar, cron	Scheduled Tasks (cron)	Start/end, CRON, presets	Schedule builder	Via Remote Config
Recurrence	Yes	Yes	Yes (presets or CRON)	Yes	Manual
Audience targeting	Segments	Segments	Audiences, opt-in	Segments	Rules
Webhook notifications	No	Via PlayStream	Yes	No	No
Dashboard management	Yes	Yes	Calendar + list views	Yes	Limited
Event templates	Via config	CloudScript	Duplicate existing	Via prefabs	No

Satori ships a dedicated events system covering battle passes, tournaments, and limited-time shops with opt-in audiences, sticky membership, feature-flag overrides for participants, and webhooks that fire on event start and end.

Metaplay has two models. Config-driven events resolve identically on client and server, so there's no sync traffic, though every future event has to be declared up front. LiveOps Events are created and edited from the dashboard at runtime with no code push. Unity has no dedicated scheduler, so live events are assembled out of Remote Config and Game Overrides with rules-based targeting.

Player segmentation and targeting

Feature	Metaplay	PlayFab	Satori	Beamable	Unity
Condition types	Custom properties, ranges, booleans	Stats, tags, location, purchases	Properties, computed, geographic	Private stats	App attributes, behavior
Compound conditions	AND, OR, nested segments	AND/OR	AND, OR, regex	AND	AND
Real-time evaluation	Yes	Yes	Yes	Yes	Yes
External audience import	No	No	Yes	No	No
Actions on segment entry	Via game logic	CloudScript, grants	Feature flags, events	Via code	Config delivery

Satori's segmentation covers default and computed properties, regex matching, manual include/exclude lists that override rules, and importing audiences from an external BI system, the only platform here that does that last one. PlayFab ties segment entry to automatic actions, firing CloudScript or granting items and currency the moment a player crosses a threshold. Metaplay evaluates Player Conditions against typed PlayerModel state with AND/OR logic and nested segments. Because segments are code, a custom property type is a PlayerPropertyId subclass you write rather than a fixed field set, which keeps segmentation in your codebase rather than a no-code panel.

AI and ML in LiveOps

LiveOps is increasingly where AI appears, both as in-product ML (churn scoring, ML-driven segmentation, GenAI analytics) and as AI moderation for live communities. It also splits from developer AI tooling, which is about how fast your team builds the LiveOps surface itself.

Platform	In-product AI / ML for LiveOps	Developer AI tooling
Metaplay	None built in; experiments, segmentation, and config are rules-based; stream analytics to your own warehouse for ML	Metaplay Agent SDK skills (public preview) + Docs/Portal MCP servers over source you own
PlayFab	Churn Prediction (Azure ML) feeds segmentation and churn-risk offers; AI chat moderation via Azure AI Content Safety	PlayFab Copilot in Game Manager (Private Preview, GDC 2024)
Satori	Databricks integration runs ML (churn, LTV, engagement) that writes scores back for segmentation and A/B targeting; GenAI features query datasets and summarize feedback	None documented (Heroic Labs publishes LLM-optimized docs)
Beamable	None documented	Markets "AI automation" (no specifics)
Unity (UGS)	Safe Text and Safe Voice ML moderation; Remote Config and Game Overrides are themselves rule-based	Unity AI in-editor assistant + Unity MCP server (engine/editor)

Satori and PlayFab put ML directly into LiveOps (churn, segmentation, moderation), Unity adds ML moderation, and Metaplay keeps LiveOps rules- and config-based while streaming analytics out for whatever ML pipeline you choose, with its AI investment in build-time developer tooling instead. Satori and PlayFab provide turnkey churn and segmentation ML; Metaplay leaves the modelling to you, running on analytics you export and source you own.

Dashboard and operations tooling

Feature	Metaplay	PlayFab	Satori	Beamable	Unity
Dashboard included	Yes	Yes (Game Manager)	Yes	Yes (Portal)	Yes
Dashboard customization	Full source access	No	No	No	No
API access	Yes	Yes	Yes	Yes	Yes
Audit logging	Yes	Yes	Yes	Yes	Yes
Role-based access	Yes	Yes	Yes	Yes	Yes

Game Manager, the Satori Dashboard, Beamable Portal, and the Unity Dashboard all cover player management, config editing, and event control through a web UI you configure but can't change. Metaplay ships the LiveOps Dashboard as Vue.js/TypeScript source, so a studio can add its own pages, rewire existing tools, or build a workflow that doesn't exist out of the box. Source-level customization suits a long-running live game with bespoke operations; the configurable dashboards cover teams that work from the standard panels.

LiveOps pricing across the five platforms

Platform	LiveOps Included	Pricing Model	Headline Numbers
Metaplay	All tiers	Flat monthly + per-day DAU	Production €1,985/mo + €2.95/1k DAU per day (first 5k DAU/day free)
PlayFab	All tiers	Usage-based	Free dev tier; Standard $99/mo; Premium $1,999/mo
Satori	Separate product	Flat, separate from Nakama	From $600/month
Beamable	All tiers	Tiered + usage	Trial $0, then Developer $125/mo and up
Unity	All tiers (free tier)	Usage-based, per service	Free tier; billed per service above limits

Metaplay includes every LiveOps feature in all paid tiers with no per-experiment or per-event charge. The licence is a flat monthly fee plus a DAU fee that is billed each day and summed across the month: Production is €1,985/month plus €2.95 per 1,000 DAU per day, with the first 5,000 DAU/day free. A game holding 100,000 DAU across a 30-day month bills (100,000 − 5,000) ÷ 1,000 × €2.95 = €280.25/day, so about €8,408 in DAU fees on top of the €1,985 base. See Metaplay pricing for the worked example.

It is worth putting that roughly €10,393/month at 100,000 DAU in context. That is about the fully loaded cost of a single experienced backend engineer. The alternative to a managed platform is building and operating these systems in-house, which means staffing a full backend and DevOps team for the life of the game. Among the options here, Metaplay is the only one that delivers a source-available backend as a managed service, so a studio gets those systems without building that team and without lock-in. Metaplay's own analysis puts an in-house backend build at around $21M over a typical multi-year build: The true cost of building your own backend.

PlayFab is usage-based with a free development tier and meter credits at higher tiers (Standard $99/month includes $400 in meters, Premium $1,999/month includes $8,000), and advanced analytics push you onto Azure Data Explorer at extra cost. Satori is priced as its own product from $600/month, so a studio already paying for Nakama adds that on top. Beamable is tiered: a 90-day $0 trial, then Developer at $125/month and up, with MAU caps as the constraint at scale. Unity bills per service with a free tier; LiveOps via Remote Config sits inside that usage-based model.

Decision guide: which LiveOps platform to choose

If you need...	Consider
Full source code access for LiveOps tools	Metaplay
Microsoft and Xbox ecosystem integration	PlayFab
LiveOps separate from your game backend	Satori
Unity-native with minimal setup	Unity Gaming Services
A backend covering Unity and Unreal	Beamable
Multi-phase experiments	Satori
SRM detection and statistical rigor	PlayFab
A dashboard you can extend with custom tools	Metaplay
Turnkey churn/LTV ML feeding segments	Satori or PlayFab
Server-authoritative experiment execution	Metaplay

Talk to us about LiveOps on Metaplay

If you want LiveOps tooling you can read, extend, and run inside your own code (experiments, events, segmentation, and a dashboard shipped as source), book a demo and we'll walk through how it maps to your game. The LiveOps overview covers the day-to-day workflow in more depth.

For head-to-head detail on individual platforms, see Metaplay vs PlayFab and Metaplay vs Nakama. If you're choosing a backend for a mobile game, the best backends for F2P mobile compares the same platforms across economy and IAP.

FAQ

Which LiveOps platform is cheapest?

It depends on scale and how you count. Unity Gaming Services and PlayFab both have free tiers that cost nothing for a small or pre-launch game, so they are cheapest at the bottom. Satori starts at $600/month flat. Metaplay's Production tier is €1,985/month plus a per-day DAU fee. As DAU climbs, usage-based meters (PlayFab) and per-DAU fees (Metaplay) weigh more heavily than the base price, so the comparison turns on actual DAU; there's no single cheapest platform across all sizes.

Do I need a separate analytics tool to run experiments?

On Metaplay, yes: experiments emit analytics events that you stream to your own warehouse (BigQuery or S3) and analyze there, so you bring your own analytics layer. PlayFab and Satori include statistical analysis in the platform: PlayFab reports significance with SRM detection inside Experiments, and Satori tracks goal metrics with CSV export. Unity tracks significance through its own Analytics. Beamable provides limited built-in reporting, so deeper reads typically go to an external tool too.

Which platforms put ML directly into LiveOps?

Satori and PlayFab. Satori's Databricks integration runs churn, LTV, and engagement models and writes the scores back so you can target segments and A/B tests on them, plus GenAI features for querying datasets. PlayFab uses Azure ML for churn prediction that feeds segmentation and churn-risk offers, with Azure AI Content Safety for chat moderation. Unity adds ML for Safe Text and Safe Voice moderation but not for experiments or segmentation. Metaplay and Beamable keep LiveOps rules- and config-based; Metaplay's path is to export analytics and run your own models.

Can I use Satori without Nakama?

Yes. Satori is Heroic Labs' standalone LiveOps and experimentation product. It works with any backend, including ones that aren't Nakama, and ships client SDKs for Unity, JavaScript, Godot, Java/Android, and Swift, with server-side support for TypeScript, Go, and Lua.

Does PlayFab require Azure?

PlayFab's backend services are Azure-hosted, though game clients can run on any platform. Advanced analytics features depend on Azure Data Explorer, and CloudScript via Azure Functions needs a separate Azure subscription.

Which platform has the best A/B testing?

They emphasize different things. PlayFab focuses on statistical analysis (up to 9 variants, SRM detection, 95% confidence on Microsoft's internal engine). Satori focuses on experiment design (multi-phase tests with locked variant assignment). Metaplay runs experiments server-side in source you own and extend, with analysis in your own warehouse. The fit depends on whether statistical tooling, experiment flexibility, or owning the logic matters most for your team.

Can I customize the LiveOps dashboard?

Only Metaplay ships dashboard source code: the LiveOps Dashboard is Vue.js/TypeScript you can extend with custom pages and actions. PlayFab, Satori, Beamable, and Unity offer configuration but not code modification of the dashboard itself.

How do these compare for mobile F2P games?

Metaplay and PlayFab are both used by major mobile F2P titles. Metaplay's server-authoritative architecture executes economy and LiveOps logic on the server, making client-side cheating of those systems structurally hard; PlayFab offers broader engine and console support. For a fuller breakdown, see the best backends for F2P mobile.

What about real-time multiplayer?

This comparison covers LiveOps. For multiplayer, PlayFab offers Multiplayer Servers and the Party SDK, Satori pairs with Nakama's authoritative match handlers, Unity has Relay and Netcode, and Metaplay provides async multiplayer entities and an MMR matchmaker (no real-time physics netcode). See each platform's multiplayer docs for detail.