Ad Network Guide
Playable Ads for Moloco
Moloco is a programmatic DSP built around machine learning, and it optimises delivery differently from traditional ad networks. Rather than using server-side models trained on aggregated historical data, Moloco deploys on-device ML inference that learns in real time from each publisher app's own user signals. This allows Moloco to predict the lifetime value of an individual user with significantly higher precision than conventional DSPs, which suits ROAS-optimised UA campaigns, where the goal is users who spend money rather than installs alone. Moloco serves mobile gaming advertisers as well as e-commerce and fintech app developers seeking high-value, high-retention installs.
Playable ads on Moloco benefit from the platform's ML signal loop in a specific way: Moloco's model incorporates creative engagement data (including how long users interact with a playable, whether they reach the CTA, and post-install behaviour) into its bidding decisions. A playable that brings in engaged users is shown to more users like them over time, so the campaign gets more efficient without manual bid changes. Technically, Moloco documents a single data-URI-compressed file with no external references, and it is explicit on two points that trip up builds made for other networks: the file must not be zipped for upload, and it must not include mraid.js. Moloco scans every upload for mraid.js and XMLHttpRequest, and the same file can also run as an Interactive End Card after a video.
The upload workflow for Moloco varies depending on your account type. Larger advertisers with managed-service relationships upload creatives through their Moloco account manager, who handles campaign configuration and creative approval. Advertisers with self-serve access use the Moloco Cloud dashboard to upload HTML files under Creatives > Interactive. In both cases the creative is a single HTML file, but not the same file you would upload to Unity Ads or AppLovin: a Moloco build carries no mraid.js and exits with FbPlayableAd.onCTAClick(), the same call Meta uses.
Export specifications
| Specification | Value |
|---|---|
| Accepted format | Single .html: must not be zipped |
| Playable Ads Maker exports | Single HTML file |
| Published size limit | 5 MB |
| Playable Ads Maker builds to | 5 MB |
| MRAID | Must not include mraid.js |
| Store-exit call | FbPlayableAd.onCTAClick() |
| External requests | None allowed at runtime |
| Optimisation targets | CPI, ROAS, in-app events (purchase, subscription, etc.) |
| Inventory source | Open RTB exchange + direct publisher deals |
| Review time estimate | 1–2 business days (managed via account manager) |
Size, format, MRAID and store-exit rows are taken from Moloco’s own documentation; “Playable Ads Maker builds to” is our exporter’s own target. Every network side by side: playable ad size limits by network.
If you already have a build, drop the HTML file into the free Playable Ad Spec Checker and it will grade it against these Moloco requirements (size, packaging, MRAID and the store-exit call) in your browser, with nothing uploaded.
Check my Moloco playable adHow to upload to Moloco
- Build your playable in the Playable Ads Maker, focusing on an interaction that clearly conveys the value proposition of your app: Moloco's ML optimises for users who install and engage, so a playable that accurately represents your app's experience drives better post-install signal.
- Select the Moloco preset in the Export modal and download the single HTML file. It carries no mraid.js tag and wires your CTA to FbPlayableAd.onCTAClick(). Do not zip it. Moloco documents that the file must not be zipped for upload.
- Check the size: Moloco publishes 5 MB, and Playable Ads Maker builds Moloco exports to 5 MB. If using managed-service, share the file with your Moloco account manager via the agreed transfer method (usually email or a shared drive).
- If you have self-serve access to Moloco Cloud, log in and navigate to Creatives > Interactive / Playable.
- Upload the HTML file and fix anything Moloco flags before the creative goes live.
- Work with your account manager (or use the self-serve UI) to configure the campaign: set your optimisation goal (CPI or ROAS), define the target in-app events to optimise for, and specify geo and device targeting.
- Launch the campaign and monitor the Moloco dashboard for creative performance metrics including engagement rate, click-through rate, and post-install conversion. Moloco's ML requires approximately 1–2 weeks of data to fully calibrate delivery.
File size tip
Moloco publishes a limit of 5 MB; Playable Ads Maker builds Moloco exports to 5 MB. If your export is too large, reduce image resolution, switch to WebP at 75–80% quality, or trim animation frame counts in the Playable Ads Maker Export modal. The live size estimate updates as you adjust each setting.
Tips for Moloco playable ads
Set up post-install event tracking so ROAS optimisation can run
Moloco's main strength is that it can optimise for post-install revenue events rather than just installs. To activate ROAS optimisation, you must have your mobile measurement partner (MMP) correctly configured to pass purchase or subscription events back to Moloco's attribution pipeline. Without this signal, Moloco defaults to install optimisation, which is significantly less efficient for revenue-focused campaigns. Before launching your playable campaign, verify that your MMP integration (AppsFlyer, Adjust, or Singular) is live and sending event postbacks.
Moloco's ML needs creative signal data, so start with broader targeting
Unlike rule-based DSPs where you manually specify narrow audience segments, Moloco's ML model discovers the best audiences automatically from broad initial targeting. When launching a new playable creative, avoid tight geo, age or interest filters. Start with the broadest reasonable audience for your app and let Moloco's model identify which user segments produce the best post-install events. Manual narrow targeting can starve the model of signal and slow the optimisation feedback loop.
Provide multiple creative variants to help Moloco's model differentiate audiences
Moloco's system allocates budget between creative variants based on which performs best for each user segment. Uploading two or three playable variants with clearly different opening hooks (for example, a mechanic-demonstration variant versus a character-story variant) allows the ML to learn that different audience segments respond to different creative approaches. Running several variants consistently improves overall campaign ROAS compared to running a single playable against the full audience.
Playable Ads Maker does the Moloco packaging (Single HTML file), wires the FbPlayableAd.onCTAClick() store exit and inlines every asset when you export.
Build your playable ad