Attribution & Measurement

Emulated Devices

Also known asEmulatorsDevice EmulationApp EmulatorsVirtual Devices

Software that mimics a real mobile device on a computer or server. Legitimate for development and QA, but also a primary tool for mobile ad fraud — faking installs and engagement at scale.

Key takeaways

  1. 01An emulated device is software that simulates a real phone/tablet — running the OS and apps virtually rather than on physical hardware.
  2. 02Two faces: legitimate (developers and QA teams test on emulators) and fraudulent (fraudsters spin up thousands of emulated devices to fake installs, clicks, and post-install events).
  3. 03Emulator-driven install farms are a core mobile-ad-fraud vector — cheap to scale, hard to distinguish from real devices without detection tooling.
  4. 04MMPs detect emulators via device fingerprint anomalies: missing sensors, datacenter IPs, impossible hardware/OS combinations, and behavioral signatures (no real touch entropy).

An emulated device is software that mimics a real mobile device — running a mobile operating system (iOS or Android) and its apps virtually, on a computer or server, instead of on physical phone hardware. The same technology has two very different uses: a legitimate one (development and quality assurance) and a fraudulent one (faking app installs and engagement at industrial scale).

The legitimate use: development and QA

Developers and QA teams routinely run apps on emulators (Android Emulator in Android Studio) and simulators (the iOS Simulator in Xcode) to test across device models, OS versions, and screen sizes without owning every physical device. This is standard, sanctioned practice — emulators are a core part of the mobile build-and-test loop. The catch for measurement: these QA sessions can pollute analytics and attribution if they aren't filtered out, inflating install and event counts with internal, non-user activity.

The fraudulent use: install farms and fake engagement

The darker side is mobile ad fraud. Fraudsters spin up large fleets of emulated devices — hundreds or thousands running in parallel on cheap cloud servers — to fake the entire user journey: a click on an ad, an "install", an app open, even scripted post-install events to mimic an engaged user. Because each emulated device can be reset to look brand new (fresh advertising ID, wiped storage), a small server farm can manufacture a huge volume of fraudulent installs that bill against a UA budget. Emulator farms are one of the cheapest, most scalable fraud techniques, which is why they remain a persistent problem — see [[ad-fraud]] for the broader fraud taxonomy.

How MMPs and fraud tools detect emulators

The major MMPs (AppsFlyer Protect360, Adjust Fraud Prevention, Singular Anti-Fraud) and the big self-attributing networks bundle emulator detection into their anti-fraud layers, flagging or rejecting emulator-sourced installs before they're attributed and billed. For advertisers, the practical defense is to run UA through an MMP with active fraud filtering, watch for traffic sources with abnormal emulator rates, and exclude QA/emulator traffic from production analytics with a test-device allowlist.

Quick answers

What is an emulated device?

An emulated device is software that simulates a real mobile phone or tablet — running a mobile OS and apps virtually on a computer or server rather than on physical hardware. Developers use emulators (Android Emulator) and simulators (iOS Simulator) for legitimate testing; fraudsters use them at scale to fake app installs and engagement for ad fraud.

Why are emulators a problem for mobile ad fraud?

Because they're cheap and scalable. A fraudster can run hundreds of emulated devices on cloud servers, each resettable to look brand new, and manufacture huge volumes of fake clicks, installs, and post-install events that bill against an advertiser's UA budget. Emulator-based install farms are among the most common and cost-effective mobile-ad-fraud techniques.

How do you detect emulator traffic?

Detection combines device-fingerprint anomalies (impossible hardware/OS combinations, missing sensors), network signals (datacenter rather than residential/carrier IPs), and behavioral signatures (no real touch entropy, identical timing across devices, impossibly fast funnels). MMPs like AppsFlyer Protect360, Adjust Fraud Prevention, and Singular Anti-Fraud automate this and reject emulator-sourced installs before attribution.

Should I exclude emulators from my analytics?

Yes — for production analytics and attribution, filter out both fraudulent emulator traffic (handled by your MMP's fraud layer) and your own QA/test sessions (handled with a test-device allowlist). Internal testing on emulators is legitimate but pollutes install and event counts if it leaks into production dashboards.

Back to glossary