×![]()
Click anywhere to close
The engine, catalog and visitor profiles behind your on-site widgets can also deliver recommendations where Personyze does not draw the page: inside an email another tool sends, as JSON for your own code, or inside your iOS or Android app. Each of these is its own campaign type, with its own wizard. This page is the map.
Pick the surface
| Where the recommendations go | Create it | Wizard steps | Guide |
|---|---|---|---|
| Products in an email sent from any email platform (Mailchimp, Klaviyo, HubSpot and the like), picked for each recipient when the email is opened | New campaign → Email & SMS → Product Recs in 3rd-Party Emails | Catalog, Interactions, Look, Recommendation, Code, QA, Automation, Performance | Open-Time Email Recommendations |
| Articles, guides or videos in such an email | New campaign → Email & SMS → Content Recs in 3rd-Party Emails | Catalog, Interactions, Look, Recommendation, Code, QA, Automation, Performance | Content Recommendations in 3rd-Party Emails |
| Products as JSON, for your own backend, app or custom layout | New campaign → App / API → JSON API — Product Recs | Catalog, Interactions, Recommendation, Data in Response, Code, Automation, Performance | JSON API — Product Recommendations |
| Articles as JSON | New campaign → App / API → JSON API — Content Recs | Catalog, Interactions, Recommendation, Data in Response, Code, Automation, Performance | JSON API — Content Recommendations |
| Products inside your native iOS or Android app | New campaign → App / API → Product Recs for Apps | Set up, Catalog, Interactions, Content, Target, Automation, Performance | Product Recommendations for Mobile Apps |
| Articles inside your app | New campaign → App / API → Content Recs for Apps | Set up, Catalog, Interactions, Content, Target, Automation, Performance | Content Recommendations for Mobile Apps |
| A decision your own backend acts on (a ranking, a price, an API response), not only recommendations | New campaign → App / API → API & SDK Personalization | Set up, Target, Content, Automation, Performance | Server-side personalization & A/B testing |
A banner rather than products is a separate campaign type: Personalized Banners in 3rd-Party Emails. For emails Personyze sends itself, see Personalized Re-Targeting Emails.
How the three families differ
| 3rd-party emails | JSON API | Apps | |
|---|---|---|---|
| What Personyze hands you | A snippet of HTML. Each product is an image, rendered for whoever opens the email. | A set of request URLs. Each answers with a JSON array of the fields you chose. | What the SDK returns: a rendered widget for a WebView, or a JSON feed your app lays out itself. |
| Who the picks are for | The recipient, named by your email platform’s merge tag (email or your own user ID). Or a product ID, for picks that are the same for everyone. | Whoever you name in the URL: an email, your own internal ID, the Personyze user ID, or an item ID. | The app user the SDK reports. |
| Who gets it | No Target step. Everyone the email goes to. | No Target step. Whoever your code asks for. | A Target step with the usual rules. |
| Checking it before launch | A QA step that shows the images any recipient would get. | No QA step. Request a URL and read the answer. | No QA step. Check it in a build of your app. |
What they share with your on-site widgets
- The catalog. One per site, kept in Settings → Recommendation setup → Product Catalog (or Article Catalog). Every wizard’s Catalog step shows the same one. See Setting the Product / Content Feed.
- The interactions. Views, add-to-carts and purchases on your site, set up in Settings → Recommendation setup → Product Tracking (or Article Tracking). What happens inside your own layout or app is something you report yourself. Each guide says how. See Product Interaction Tracking.
- The algorithms. The email and JSON wizards open the same algorithm picker as an on-site widget. Only the Page row is different. An email or an API call is not on a page of your site, so the row offers what the person did instead: General, Items they viewed, Items in their cart and Items they purchased. See Recommendation Algorithms, Filters and Fallbacks.
Measuring it
| Campaign | Where | What it counts |
|---|---|---|
| 3rd-party emails | The campaign’s Performance step. For one campaign or all of them: Analytics → Off-site campaigns → 3rd-party emails (the Get Code statistics screen). | It starts at Viewed, because Personyze does not send the email. Then Clicked → Added to cart → Purchased, with attributed revenue. For articles it is Viewed → Engaged → Converted. |
| JSON API | The campaign’s Performance step (JSON API statistics). | A Viewed for each recommendation served to your application, and a click each time someone goes through the click_url returned with an item. Then the same funnel as the emails. |
| Apps | The Performance step’s Recommendations tab, the same as an on-site widget. | Shown, viewed, clicked, added to cart and purchased, with attributed revenue. For articles: shown, viewed, engaged and your content goal. |
On the email and JSON screens, every metric card has a Users … link that lists the people behind the number, and the list can be exported to CSV. Figures appear 1–3 hours after a session ends. What each number means: Email, SMS, web push and 3rd-party emails and Product and content recommendations.
Related
- The Recommendations hub, for everything else about recommendations.
- Targeting & Personalization for Mobile Apps, for installing the SDK.
- REST API