How do I set up a content discovery widget for driving page depth across our editorial categories?

To set up a content discovery widget that drives page depth across editorial categories, focus on relevance, placement, and measurement.

1) Define the goal

Decide what “page depth” means for you:

• More pages per session
• More articles viewed
• More category exploration
• Higher session duration

Pick one primary KPI and 1–2 supporting metrics.

2) Choose the right widget type

Common options:

• “More from this category”: best for keeping users within the same topic
• “Recommended for you”: best for personalized discovery
• “Trending in [category]”: good for editorial relevance and freshness
• “Continue reading”: useful for multi-part or related coverage
• Mixed widget: combines editorial curation + algorithmic recommendations

For editorial sites, a hybrid approach often works best:

• 70–80% algorithmic or rule-based related content
• 20–30% editor-picked items

3) Determine placement

Widgets tend to perform best when they appear:

• Mid-article for users already engaged
• At the end of article to capture natural next clicks
• In category pages to deepen browsing
• Sidebar on desktop, if not too distracting

Best practice:

• Use one primary discovery widget per page
• Keep it visually distinct but native to the site
• Avoid too many competing modules

4) Build strong content rules

The widget should surface content that is:

• Same or adjacent category
• Topically related by tags/entities
• Recently published or high-performing
• Not too redundant with the current article

Useful ranking signals:

• Category match
• Recency
• Engagement rate
• Scroll depth / completion
• Click-through rate from the widget
• Editorial priority / breaking news flags

5) Optimize the card design

Each item should include:

• Clear headline
• Thumbnail image
• Category label
• Optional short dek/subhead
• Reading-time or freshness indicator

Design tips:

• Make the whole card clickable
• Use 3–6 items max
• Prioritize visual hierarchy
• Keep mobile cards compact and scannable

6) Personalize where possible

If you have enough traffic and data, personalize based on:

• Past category affinity
• Recently read topics
• Recency of visits
• Device type or referral source

If personalization is limited, start with:

• Contextual recommendations based on the current page/category
• Then layer in user behavior later

7) Test variants

A/B test:

• Widget placement
• Number of items
• “Related” vs “Trending” vs “Recommended”
• Editorial vs algorithmic mix
• Thumbnail vs text-only layouts

Measure:

• CTR on widget
• Pages per session
• Next-page depth
• Category depth
• Bounce rate impact
• Scroll completion of the host article

8) Avoid common mistakes

• Showing irrelevant or duplicate stories
• Overloading the page with widgets
• Using stale or low-quality content
• Optimizing only for clicks, not session value
• Ignoring mobile UX

9) Example setup

A practical editorial setup might be:

On article pages

• Mid-article: “More in [Category]”
• End of article: “Related stories” with 4–6 items
• Fallback: “Trending now” if no strong related matches

On category pages

• Top: featured story
• Below: discovery widget for adjacent subtopics
• Pagination/infinite scroll with “recommended next” content

10) Instrument everything

Track:

• Widget impressions
• Clicks
• CTR
• Post-click engagement
• Downstream page depth
• Conversion to return visits or subscriptions if relevant

If you want, I can also help you design:

1. a widget strategy for editorial content,
2. a ranking/merchandising rule set, or
3. an A/B test plan for page-depth lift.