Filtering Discover by platform, pillar, and outlier

This article covers each filter, what it does, and the workflow that uses them well.

The filters available

Pillars

A pillar is a top-level topic ("Productivity", "Fitness", "Marketing", "Cooking"). Eden ships with a fixed set, and during onboarding you pin the ones you want in your feed. They show up as a horizontal chip strip above the search input.

How the chips work:

• All chip on the left shows posts across every pillar you have pinned.
• Click any pinned pillar to filter the feed down to that one pillar.
• + Add lets you pin another pillar to the strip. Only pillars you have not already pinned show up in the picker.
• Hover a pillar and click the small × that appears to remove it from your strip. Your pinned pillars persist across sessions.

Pillars are the single most useful filter day-to-day. Most of the work in Discover is "look at what is hitting in my pillar this week," and the chips are the fastest way to get there.

Search

The search bar above the chips runs a keyword query against post titles and bodies. Eden uses a semantic search under the hood, so "morning routine ideas" finds posts about waking up early, journaling, and cold showers, not just exact substring matches.

Search composes with every other filter. You can search "ai writing" inside the Productivity pillar, on YouTube only, with a 10× outlier minimum.

Platforms

The Platforms dropdown lets you toggle which of the six platforms show in the feed: YouTube, TikTok, X, Instagram, LinkedIn, Substack. By default all six are on.

Common ways to use this:

• Studying a single platform when you are writing for it that day.
• Excluding a platform you do not post on so the feed reads tighter.
• Comparing the same topic across two platforms by toggling between them.

Your platform selection is remembered per workspace, so you do not have to reset it each session.

Follower range

The Audience filter narrows the feed to creators in a given follower range. Presets:

• Any. No lower or upper bound.
• 1K – 20K. Micro creators.
• 20K – 100K. Small creators.
• 100K – 1M. Mid creators.
• 1M – 8M. Large creators.
• 8M+. Mega creators.

Or set a custom min and max in the input fields if the presets do not match what you want.

This filter is most useful when you are deliberately studying creators at a specific size. A solo creator at 5K wants to see what is working for other creators at 5K to 20K, not what a 5M-follower account is doing. A 500K creator wants to study the next size up to figure out what changes there.

Outlier minimum

The Outlier filter sets a floor on the multiplier. Presets:

• Any (no extra outlier minimum you set; the feed can still apply engagement floors and, when inventory is sparse, show broader matches below your usual bar)
• 3× or more
• 5× or more
• 10× or more (default on first visit)
• 20× or more

Any does not add an extra minOutlierScore on the request. 3×, 5×, 10×, and 20× send that floor to the server. Move the threshold up when you only want posts that genuinely broke through. Read reading the outlier multiplier for what these numbers mean.

A higher threshold is good for studying. Any or 5× is better for trend spotting, since rising posts often start at lower multipliers and grow over a few days.

A workflow that uses the filters well

A workflow that works for most creators:

1. Click the chip for the pillar you write in.
2. Set platform to the one you are working on that day.
3. Leave Audience on its default and Outlier on 10× or more (or choose Any / a lower floor) unless you are looking for something specific.
4. Scan the feed. Save anything worth studying onto a board.
5. When you sit down to draft, open the board next to your draft.

The default view is the right view most days. Filters narrow the field quickly when you are doing a specific job, like "find ten 10×+ Substack notes about productivity from creators under 100K so I can study how small accounts are breaking through." Pillars and search will take you most of the rest of the way.

Where to go next

• For what the outlier number means and how to read it, read reading the outlier multiplier.
• For platform-specific patterns, read the platform pages: YouTube, TikTok, X.