To get the best results from Einstein Article Recommendations, review its data requirements and get tips for testing the feature.
Required Editions
| View supported editions. |
Make Sure You Have Enough Data
Plants need water, support service reps need hugs, and Einstein needs data! Without high-quality historical data to learn from, Einstein can’t make accurate recommendations. Here’s what you need.
| What You Need | Required or Recommended? | Details |
|---|---|---|
| Three or more published knowledge articles | Required |
When your Salesforce Knowledge article coverage is extensive and addresses common customer questions, Einstein can make more useful recommendations. Articles don’t need any special tags, data categories, or metadata to be recommended. Newly published or overwritten articles automatically undergo term analysis within one day to be considered for recommendation. If there are multiple versions of an article, only the most current active version is recommended. Einstein only recommends articles stored in your Salesforce Knowledge base. Articles hosted in an external system don’t count. |
To show recommendations in multiple supported languages, you only need three or more knowledge articles in one of the languages.
Take Einstein Article Recommendations for a Test Run
Over several releases, we’re phasing in article recommendations as active in orgs with Knowledge. If we set it up for you, we supply a pre-trained model that uses generic data and default field and language settings. If your business uses other fields, your service reps can see more accurate recommendations when you accept the terms in Setup, select the right fields, and update the model so that Einstein learns from your org’s closed-case data.
While you can set up Einstein Article Recommendations in a sandbox org, recommendations differ between sandbox and production because of underlying data differences. Because the same results aren’t guaranteed, we recommend turning it on in your production org and giving a small group of service reps access to article recommendations. When you’re comfortable with the way recommendations are working, open them up to your larger team.
If you do decide to use article recommendations in sandbox, follow these best practices:
- Match your sandbox licenses to your production licenses.
- To meet the data requirements and guarantee realistic recommendations, make sure that your sandbox org’s article and case data is similar in quality and quantity to your production org’s.