Editorial Review by TopAdNetworks
Moloco’s defining trait is that it was built on deep learning from day one, not retrofitted onto legacy infrastructure.
Founder Ikkjin Ahn built YouTube’s video profitability systems and Android’s data infrastructure at Google before starting Moloco in 2013. Every bid runs through 8 to 10 specialized models covering conversion probability, fraud detection, frequency capping, and LTV forecasting, executing up to 8 million predictions per second.
Performance credentials are independently validated: in the 2025 AppsFlyer Performance Index, Moloco ranked second globally for midcore gaming on iOS and in the top 5 across 19 Android categories, alongside AppLovin and Liftoff. Moloco’s August 2025 research with Sensor Tower and Singular found Google and Meta still capture 88% of mobile ad spend, and that diversifying beyond them can lift returns to 214%.
Moloco’s expansion into Retail Media and streaming monetization bets that the same ML engine built for mobile UA can be repointed at retailer first-party data. The 2025 Google Cloud partnership, migrating Retail Media to Vertex AI Vector Search, delivered roughly 10x retrieval capacity and up to 25% lower p95 latency.
The platform’s real limitation, noted across independent reviews, is accessibility: Moloco does not support open self-registration, requiring an approved account or agency relationship, with documentation skewing toward advanced buyers.
Company Overview
Moloco was founded in 2013 in Redwood City by Ikkjin Ahn (CEO) with co-founders including Dong-hwan Park and David Sehyuk Park (CIO), all formerly of Google. Ahn was an ML engineer at YouTube (2008-2010) and Android’s data team (2010-2013).
The company has raised approximately $200-227 million total, including a $150 million Series C led by Tiger Global in August 2021, valuing it at $1.5 billion.
As of 2026, Moloco employs 700-plus people across 12 global offices, including the US, UK, South Korea, China, Japan, and Singapore, with over 60% in engineering and data science. The company reports five consecutive years of profitability and an estimated $400 million revenue run-rate as of 2025, up from $100 million-plus at its 2021 Series C, citing 4,455% three-year growth on the 2022 Inc. 5000 list.
Leadership includes DJ Park as CTO, who joined in 2015 from Google’s Cloud Storage team, and Pat Copeland, GM of Moloco Commerce Media, who previously scaled Amazon’s Sponsored Brands.
Confirmed clients include King Digital, Playrix, and Netmarble, reflecting Moloco’s roots in mobile gaming UA before its retail media expansion.
How The Platform Works
Moloco Cloud DSP is a fully ML-driven bidding system. It ingests first-party advertiser data, including in-app events and purchase signals, through a modular pipeline: data ingestion, a prediction engine forecasting conversion and LTV, real-time RTB bidding, and a continuous optimization loop retraining models as data accumulates, refining accuracy without manual rule-setting.
Inventory access comes via integrations with major in-app supply partners including AppLovin, ironSource, Unity, and Vungle, giving reach across roughly 3 million apps and nearly 10 billion devices. Campaigns center on a single core objective, typically CPI or in-app event optimization, with ML handling bid pacing and creative rotation.
Advertiser Onboarding
- Apply for a verified Moloco account; self-registration is not available
- Connect a mobile measurement partner (MMP) for postback tracking required by the ML engine
- Define a primary objective: CPI, in-app event optimization, or ROAS-based goals
- Upload video and banner creative assets; video generally outperforms static banners
- Fund the account with a recommended starting budget, commonly $300 to $500 per campaign
- Launch the campaign; Moloco’s AI handles bidding, pacing, and optimization automatically
Ad Formats And Placement
- In-app video: full-screen and rewarded video, generally outperforming static banners
- Display and banner: standard in-app placements across partner inventory
- Native: content-matched in-app native placements
- CTV: connected TV inventory via Moloco Performance CTV
- Playable and interactive ads: interactive formats for gaming promotion
Targeting And Optimization
- ML-driven bidding: 8 to 10 models per impression covering conversion, fraud detection, frequency capping, and LTV
- Geo, OS, device, and app version: foundational parameters set at campaign launch
- Behavioral optimization: core mechanism, based on first-party conversion data rather than declared interest
- Real-time budget reallocation: spend is redistributed by hour and day
- LTV-based bidding: optimize toward predicted lifetime value rather than initial install cost
Key Differentiators
Moloco’s competitive position rests on deep learning built from inception, not layered onto legacy infrastructure, massive bid-processing scale, and one ML engine applied across mobile UA, retail media, and streaming.
1. Deep Learning Built From Day One
Unlike DSPs that added ML as an optimization layer over time, Moloco’s bidding engine was architected around deep learning from its founding in 2013. Every impression passes through 8 to 10 specialized models executing up to 8 million predictions per second, enabling behavioral precision that rule-based systems cannot replicate.
2. Independently Validated Performance Rankings
In the 2025 AppsFlyer Performance Index, Moloco ranked second globally for midcore gaming on iOS and top-five across 19 Android categories, an independent third-party benchmark rather than a self-reported claim, placing Moloco alongside established DSPs like AppLovin and Liftoff.
3. Massive Scale: 600 Billion Daily Bid Requests
Moloco processes approximately 600 billion bid requests daily across 3 million-plus apps and nearly 10 billion connected devices in 190-plus countries, giving its ML models substantially more training data per cycle than smaller DSPs and directly improving prediction accuracy over time.
4. One Engine Across UA, Retail Media, and Streaming
Moloco applies one deep learning engine across three product lines: Cloud DSP for mobile UA, Retail Media Platform for first-party commerce monetization, and Moloco Monetization for streaming and OTT. The 2025 Google Cloud Vertex AI Vector Search integration added semantic ad matching, improving retail media retrieval capacity by roughly 10x.
Ideal Use Cases
Moloco is best suited for performance-focused mobile advertisers, particularly in gaming and e-commerce, who want reliable first-party conversion data and ML-driven scale beyond Google and Meta.
- Mobile gaming UA: studios optimizing install volume and in-app event conversion via behavioral ML bidding
- E-commerce and retail media: retailers monetizing first-party shopper data via the Retail Media Platform
- Diversification beyond walled gardens: advertisers seeking reach across the independent app ecosystem
- Streaming and OTT monetization: media companies building ad-supported revenue via Moloco Monetization
- LTV-focused performance marketing: fintech and e-commerce advertisers optimizing toward long-term value
Target Clients
Advertisers
Moloco’s advertiser base spans mobile gaming, e-commerce, fintech, and DTC brands, with confirmed clients King Digital, Playrix, and Netmarble. Clients range from developers with under 100,000 users to those exceeding a billion users, with some spending over $1 million monthly. Access requires a verified account or an agency relationship, which favors established performance marketing teams.
Publishers
Moloco accesses publisher inventory through integrations with AppLovin, ironSource, Unity, and Vungle, rather than a direct publisher program. Retail and streaming publishers access Moloco differently, through the Retail Media Platform and Moloco Monetization, building their own ad businesses on Moloco’s infrastructure.