Personalized article, guide and video recommendations delivered inside your native iOS or Android app — the same engine behind content recommendations on your website. Because the profile is shared, a reader who has been following a topic on your site sees related content in the app, and what they read in the app shapes what they see back on the web.
In the panel: New campaign › Content Recs for Apps.
This uses the same SDK as product recommendations for apps. If that is already integrated, there is no further engineering work — you create a content recommendation instead of a product one.
Before you start
The SDK has to be installed and reporting first. Installation for both platforms, the initialisation code, and how to confirm data is arriving are in Targeting & Personalization for Mobile Apps. If your device does not appear in Live visits, stop and fix that first — recommendations cannot be produced for a user Personyze cannot see.
You also need a content catalog: your articles, guides or videos, each with at least an internal ID, a title, a URL and an image. It lives in Settings → Recommendation setup → Article Catalog, or you can push it over the API with the articles object.
The campaign steps
Seven steps: Set up, Catalog, Interactions, Content, Target, Automation and Performance. As with all app campaigns there is no QA step, and Content comes before Target.
1. Set up
Connects the SDK, with a tab each for Android, iOS and Server (REST). It has three steps: add the library, initialise it with your API key, and report each screen. If product recommendations for apps already run, this is done.
2. Catalog
Your content catalog. Categories and tags on each item are worth the effort here — they are what lets the engine recommend along a topic rather than just by popularity.
3. Interactions
Report what people read. In an app this must be explicit: your app calls the SDK when an article is opened, and optionally when it is liked, commented on, or completed.
Content interactions are the raw material for interest inference — this is what builds the topic profile that both app and website recommendations draw on. An app that reports nothing will still receive recommendations, but they will be generic.
4. Content
Pick the algorithm — related to what they are reading, popular in a topic, trending, or personalized to their inferred interests (see Recommendation Algorithms) — and choose the rendering shape:
- HTML (WebView) — a rendered widget for a WebView.
- JSON — the Content Recommendations via JSON Feed returns structured data your app lays out natively. Choose the fields you want returned: title, image, URL, category, publish date, and any custom fields.
5. Target
Who sees it, using all standard targeting.
6. Automation
This step is optional. Add rules that alert you, or act, when this campaign’s numbers cross a line. See Automation.
7. Performance
The Performance step reports shown, viewed, engaged (a tap on a recommended item) and your content goal. The engaged rate is measured against the times the recommendation was actually on screen, not every time it was sent. See Content recommendations.
Validating it
- Confirm the device appears in Live visits.
- Open a few articles in the app and check the interactions register.
- Widen the audience, confirm recommendations arrive and are populated.
- Narrow the rules back and confirm the change takes effect.