AI personalization explained

rasa.io doesn't send the same newsletter to everyone. Instead, our AI selects different articles for each subscriber based on what they're likely to engage with. This article explains how that works.

The big idea

Traditional email blasts send the same content to everyone on a list. With rasa.io, every subscriber gets a personalized version of your newsletter — same template, same brand, but with articles chosen specifically for them.

Over time, the AI learns what each subscriber is interested in, which leads to higher engagement (opens, clicks, time spent) than a one-size-fits-all newsletter.

How it picks articles

The AI considers several signals when choosing articles for a specific subscriber:

1. Topic tags — every article gets tagged with topics (e.g., "leadership," "remote work," "industry news"). The AI tracks which topics each subscriber clicks on.
2. Source preferences — if a subscriber consistently engages with articles from a specific source, the AI gives that source more weight.
3. Custom Tags you add — when you add Custom Tags to an article in the Edit Article panel, the AI treats them as additional topical signals.
4. Featured articles — articles you feature get priority placement.
5. Engagement history — what the subscriber has opened and clicked in past newsletters.
6. Recency — newer articles generally outrank older ones, unless the subscriber's history suggests otherwise.

What you control vs what the AI controls

You control:

• Your sources (where articles come from)
• Filters that block low-quality or irrelevant content
• Featured articles and featured sources
• Sections and section limits
• Min/max article counts
• Custom Tags

The AI controls:

• Which specific articles each subscriber sees, within your constraints
• How articles are ranked for each subscriber
• Subject line article selection (when set to "Selected" or "Featured")

This is the right division of labor: you set the strategy, the AI handles the personalization at scale.

Topic tagging

Every article that comes into rasa.io is automatically tagged by topic. These tags are visible (read-only) on the Edit Article panel.

If an article is in a language rasa.io's AI doesn't support, tagging may fail and the article could end up Inactive. See Languages for the list of supported languages and how to enable them.

Custom Tags

You can add your own internal tags to any article via the Edit Article panel. Custom Tags:

• Are internal (subscribers never see them)
• Are treated as Topics by the AI for personalization purposes
• Are useful for grouping articles by an organizing principle the auto-tags might miss (e.g., "member spotlight," "board recommended," "campaign content")

Featured content

Featured articles and Featured Sources help you put a thumb on the scale:

• Featured Source — every article from this source is automatically featured
• Featured Article — a specific article gets priority placement

The AI still personalizes within those — for example, of three featured articles, one subscriber might see article A near the top while another sees article B — but featured content as a group gets priority.

How long does personalization take to "kick in"?

The AI needs some signal to personalize. For brand-new subscribers, early newsletters will be more generic (often featuring popular articles or your featured content). As a subscriber clicks and engages, personalization sharpens.

Most newsletters see noticeably better engagement within 4-6 sends.

Common questions

Does every subscriber get a different newsletter?

Yes — different combinations of articles, ranked differently. Same layout, same template.

Can I see what a specific subscriber received?

You can see overall analytics and per-issue stats. For a specific subscriber's version, contact support@rasa.io.

What if I want everyone to see a specific article?

Feature it. Featured articles appear for all subscribers (subject to your Article Counts settings).

Can I turn off personalization?

The AI is core to how rasa.io works — turning it off would essentially make rasa a regular email tool. If you want fully consistent content for everyone, use a Blast instead of a newsletter.

What's next

• Curating content — managing your Content Pool
• Sources — where articles come from
• Analytics overview — see how personalization is performing