When someone likes¹ or favorites¹ an article online, the platform often uses tags² to decide what else shows up in their feed³. These tags are meant to act as signposts, guiding the algorithms⁴ toward content that may match the reader’s interests. In theory and at the core, this makes sense. Heck, I even use tags on my blogs or online posts while learning a great deal researching the content for this blog.
The Conundrum
If someone enjoyed a post about “prison dog training programs that rehabilitate inmates”, they might want to see more content about “dogs”, “training”, or “rehabilitation”. If they liked an article about “electric car manufacturers adopting Tesla charging ports as a standard”, the platform might think they want more news about “charging”, “adopting”, or “electric”.
The problem is that tags are often misused and are certainly too simplistic. Instead of being accurate descriptors, they are treated by authors and publishers as bait for algorithms. Writers know that trending⁵ tags attract attention, so they attach them even when the connection is weak or superficial. This creates a discovery⁶ experience that feels broken because the system assumes relevance based on tag overlap, not on tag context.
Why This Matters
Misused tags don’t just create clutter within one’s feed with useless content; they distort the way information spreads online in general. When tags are applied broadly or generically and the algorithms amplify content that share those tags, even if the subject matter is unrelated. This means one’s feed becomes a mix of relevant posts and irrelevant noise. Over time, this can lead to frustration, disengagement, and even misinformation when sensational or viral content hijacks popular tags.
The Trade-Off
Here’s the hard truth: Fixing this problem comes with a cost. If platforms enforce stricter tagging rules or implement smarter algorithms, authors and publishers might lose some of the extra clicks they currently get from the wild west of tag-based exposure. Viral content thrives on broad tags because it reaches audiences who weren’t specifically looking for it. Narrowing that reach could mean fewer accidental views and less engagement for trending topics.
For readers this narrowing is an absolute win. The readers will see more relevant content within their feeds; they will experience less clutter and have a better discovery experience.
For publishers, this is an unwanted refining and challenge to their livelihood. They would need to focus on quality of content, not exploiting tags for visibility, rather providing clarity. This shift could reduce the dominance of viral posts and make niche or specialized content harder to push into the mainstream.
Why This Happens
Tags are often applied broadly to maximize reach. Writers and publishers know that certain tags trend, so they attach them even when the connection is weak. Algorithms then assume relevance based on tag overlap, not on meaning.
This creates a feedback loop:
- You engage with content using certain tags:
Like, favorite, or click an article - Algorithm assumes you want more of those tags:
It prioritizes quantity over quality - Your feed fills with loosely related content:
Instead of depth, you get noise - Frustration sets in:
You complain about irrelevant or conflicting content - You disengage:
Stop liking or favoriting, leaving only passive views for the system to measure - You engage with content with what you think is caution and measured precision:
Like, favorite, or click an article - It all starts over again…
Why We Still Need a Solution
Despite the trade-offs above, the current system prioritizes quantity over quality, and that erodes trust. Readers want feeds that reflect their interests, not a flood of loosely related articles. A solution that balances relevance with reach is essential. Smarter tagging, user-driven controls, and algorithmic improvements can create a healthier content ecosystem. Perhaps even one where discovery feels intentional, not accidental?
What can be Done in the Future
Platforms could improve by more carefully allowing the tags on articles or posts and perhaps deprioritizing tags when creating a feed.
- Contextual Tagging: Use AI to understand the actual topic of the article or blog, not just keywords.
- User Intent Filters: Let readers specify why they liked something (e.g., for its technical depth, human-interest angle, author, publisher, sources, or policy analysis).
- Tag Quality Scores: Penalize articles that misuse tags or reward those with accurate tagging.
For now, Readers can
The best we can do is curate aggressively. Unfollow tags that lead to irrelevant content and use custom lists or bookmarks to keep our feeds clean.
My Proposed Solutions
Make Tags in an Article One Complete Marker
Instead of treating tags as individual signals, platforms should group all tags from an article or post into a single unit called a Marker⁷. A Marker represents a fuller context of the article by combining its tags into a Marker. To qualify as a match for a Marker, a new article must include at least 75% of the tags within the Marker from the original article.
If a new article only matches 2 tags of 10 tags (like “tesla” and “cars”), it won’t flood into the feed because it doesn’t meet the threshold. This approach ensures that recommendations are based on strong contextual similarity, not just one or two overlapping tags. It reduces irrelevant content and makes your feed more precise and meaningful.
Suggestion
- Original Article: training, adoption, rehabilitation, prison, dog, inmates, benefit, charity
- Matching Article with 6 of 8 tags the same: rehabilitation, dog, training, adoption, benefit, charity
- Article that would have been noise with 4 of 8 tags the same: prison, rehabilitation, inmates, benefit
More Complex Tag Requirements of the Author
Another method to improve the quality of recommendations is to require authors and publishers to use tags more thoughtfully. They could also be required to provide two or more words within each of their tags, ensuring that tags are written by relevance, not just simplicity and reach. This would help algorithms make better decisions and reduce the amount of irrelevant content in your feed.
Suggestion
- Poor Tagging Example: training, adoption, rehabilitation, prison, dog, inmates, benefit, sentenced, life
- Helpful Tagging Change: dog training, rehabilitation of inmates, dogs rehabilitate inmates, benefit of dogs
Suggestion
- Poor Tagging Example: table, tennis, badminton, pickleball, indoor, game, ping, pong, ball, racket, international, challenge, Olympics, China, gold, medal, award
- Helpful Tagging Change: table tennis, ping pong, Olympic gold medal
Limit Tag Options
Instead of allowing a free-for-all of individual words or just popular trending tags, platforms could set stricter guidelines. For example, authors could be limited to only selecting a limited number of tags that truly reflect the article’s main topics. This limit could be five total tags from a curated list provided by the platform, not the author. This would ensure the author does their darndest to only note what is pertinent from that narrowed list of options they can select.
Suggestion
Curated list of Sports Tags: table tennis, badminton, pickleball, baseball, softball, Olympics, soccer, football, hockey, polo, water polo, downhill skiing, cross country skiing, basketball, volleyball, tennis, golf, rugby, cricket, lacrosse, surfing, snowboarding, skateboarding, swimming, diving, marathon, half marathon, triathlon, cycling, mountain biking, wrestling, boxing, MMA, gymnastics, equestrian, rowing, sailing. Archery, shooting, fencing, curling, indoor, outdoor
Allow the Reader to Favorite Tags, not the Article
Another solution is to give readers more control over their discovery experience. Instead of only liking or favoriting entire articles, readers could favorite specific tags. If I am interested in “rehabilitation” but not “inmates,” I can favorite the “rehabilitation” tag and mute other tags life “inmate.” This would allow the platform to fine-tune my recommendations, showing me more of what I actually want and less of what I don’t. It shifts some of the power from the algorithm to the user, making a feed more personal and relevant.
Suggestion
- All Tags from an Article: training, rehabilitation, prison, dog, inmates, benefit, sentenced, life
- Helpful Tags Liked: training, dog, rehabilitation, benefit
- Muted Tags: prison, inmates, sentenced, life
Allow the Reader to Write the Tag for Themselves
Finally, platforms could let readers add their own tags to articles they like. If a viewer reads a story about NASA’s Artemis program and wants to see more about “space exploration” but not “astrology,” they could tag it themself. Over time, the system would learn from their custom tags and adjust their feed accordingly. This approach recognizes that readers often understand their interests better than algorithms or publishers do. By letting them define what matters, the platform can deliver a discovery experience that matches their intent, not just the author’s or the algorithm’s guess.
- All Tags from an Article: NASA, astrology, Artemis, program, space, moon, solar, system, exploration, mission, takeoff, launch,
- Tags Written by the Reader: NASA Artemis Program, Space Exploration, Mission Launch
Examples of the Clutter
(A) Dog Fighting
Tags: training, rehabilitation, prison, dog, sentenced, lives, life, inmates, transformation, programs
Article liked: How prison dog training programs transform lives of inmates
Articles shown instead:
- Man sentenced to 475 years to life in prison for dog fighting,
- Georgia dog fighting trainer sentenced for taking the lives of his fighting dogs
Both of those articles share the tags, but the context is completely different. The reader was interested in rehabilitation programs, not criminal dog fighting cases.
(B) Electric Car Charging
Tags: electric, cars, tesla, chargers, charging, EV, manufacturers
Article liked: Non-Tesla EVs manufacturers gain access to Tesla Superchargers Stations
Articles shown instead:
- Tesla Supercharging Stations targeted by Antifa protesters
- Tesla controversy about now charging for Autopilot in their EVs
Again, the tags match, but the intent does not. Someone wanted technical updates about EV infrastructure, not political drama.
(C) Plant Based Meals
Tags: nutrition, healthy, recipes, fasting, fast, vegan, plant, food, meals, busy
Article liked: 10 Easy Plant-Based Meals for Busy Weeknights
Articles shown instead:
- Celebrity vegan diet secrets revealed
- Healthy fasting trends mentioned on TikTok
They were looking for practical meal ideas, not fad diets or celebrity gossip.
(D) NASA Artemis Program
Tags: space, NASA, missions, landing, stars, program, planet, solar, system, lunar, landing
Article liked: NASA’s Artemis Program Prepares for Lunar Landing
Articles shown instead:
- Alien landing conspiracy theories gain traction online
- Star and planet related astrology predictions for your zodiac sign
You wanted science and technology updates, not pseudoscience or sensationalism.
Definitions – Help with Terms
#1 What is a “Like” or “Favorite”
When you interact with content online, you often have quick ways to show appreciation or interest without leaving a comment. These actions are commonly called likes or favorites, and they come in different forms depending on the platform.
- A thumbs up icon usually represents a like.
- A heart icon often signals a favorite or that you love the content.
Whether it’s on a news article, a blog post, or a social media feed, these simple gestures tell the platform you enjoyed what you saw. Behind the scenes, they do more than express approval. They influence algorithms, shape your recommendations, and determine what fills your discovery panel.
#2 What is a “Tag”
A tag is a keyword attached to an article, post, or piece of content to describe its topic. Tags act like labels that help organize information and make it easier for algorithms to recommend related content. When you click “like” or “favorite” on an article, the platform often uses its tags to decide what else to show you.
Tags are powerful because they connect similar topics across different sources. But they only work well when they are accurate. If tags are too broad or misused, they can lead to irrelevant recommendations and cluttered feeds.
#3 What Is a “Feed”
A feed is the stream of content you see when you open a social media app, news platform, or blog site. It’s designed to keep you updated and engaged by showing posts, articles, videos, and updates in one continuous scroll.
Your feed is usually personalized. Algorithms decide what appears based on signals like:
- What you like or favorite (thumbs up or heart)
- Who you follow
- Topics or tags you interact with
- Your browsing and reading history
The idea is to give you content you care about. In practice, it often becomes a mix of related posts and irrelevant content, especially when tags are misused. Your feed is not static in any way. It changes constantly as you interact with content, making every click a signal that shapes what you’ll see next.
#4 What Is an “Algorithm”
An algorithm is a set of rules or instructions that a computer follows to solve a problem or complete a task. In the context of your feed, an algorithm decides what content you see next. It takes signals like likes, favorites, thumbs up, and hearts, then combines them with other factors such as tags, engagement trends, and your past behavior.
Think of it as a recipe:
- Ingredients: Your interactions, tags on articles, and overall activity.
- Steps: The platform processes these inputs to predict what you might enjoy.
- Result: A curated feed that feels personalized — at least in theory.
The challenge is that algorithms rely heavily on tags as their only engagement signals, which can lead to mislead guidance. This is why your feed sometimes shows content that shares keywords but misses the context you care about.
#5 What Is “Trending”
Trending refers to the content that is currently gaining the most attention across a single platform or the Internet. It’s the material that algorithms push to the top because many people are interacting with it, by liking, favoriting, sharing, or commenting.
When you see a trending article or post, it usually means:
- High engagement: Lots of thumbs up, hearts, and shares.
- Rapid growth: The topic is spreading quickly across feeds.
- Algorithm boost: Platforms prioritize it because they believe it will keep users engaged.
Trending is not always about quality or depth. It’s about momentum. A post can trend because it sparks controversy, uses popular tags, or taps into a viral moment. This is why trending content often dominates discovery panels, even if it’s not what you truly want to see. Remember algorithms chase popularity, not precision.
#6 What Is a Discovery?
A discovery section is the part of a platform designed to help you find new content beyond what you already follow. Unlike your main feed, which focuses on updates from your connections or subscriptions, the discovery section surfaces articles, posts, and videos based on your interests, engagement history, and trending topics.
It works by analyzing signals such as:
- Tags attached to content you’ve liked or favorited
- Topics you’ve searched for or interacted with
- Popular subjects gaining traction across the platform
The goal is to introduce you to fresh material that feels relevant. In theory, this should expand your horizons and keep your experience engaging. In practice, when tags are misused, the discovery section can become cluttered with loosely related or sensational content. Instead of finding meaningful new ideas, you often end up scrolling through noise.
A well-designed discovery section should prioritize context and intent, not just keyword overlap. When it fails, it turns what should be a helpful feature into a frustrating one.
#7 What is a “Marker”
This is not a term you’ll find in any textbook or tech glossary. It’s my own creation, basically coined by me Gregory Scott Wall, PMP. I call it a “Marker”, and it represents a new way of thinking about tags and content relevance that I came up with one early morning while writing this blog.
Definition of Marker
A Marker is a complete set of tags taken from a single article or post, treated as one unified context signal. Instead of breaking tags apart and letting algorithms match them individually, a Marker groups them together to capture the full meaning and intent of the original content.
Why Markers matter?
When platforms rely on individual tags, even one matching keyword can pull in unrelated articles. That’s how you end up liking a piece about prison dog training programs and then seeing stories about dogfighting crimes. A Marker solves this by requiring strong contextual alignment
Here’s how it works:
- Every article has a Marker made up of all its tags.
- When you like an article, its Marker becomes the reference point.
- For another article to appear in your feed, its Marker must match at least 80% of the original tags.
Example:
- Original Marker (group of tags): electric, cars, superchargers, EVs, manufacturers
- Match requirement: 4 out of 5 tags must align for the new article to qualify.
- Positive Result: One’s feed shows articles with the following options
- Electric, cars, EVs, manufactures
- Cars, superchargers, EVs manufactures
- Electric, cars, superchargers, manufacturers
This approach ensures recommendations are based on meaningful similarity, not superficial overlap. It reduces irrelevant content, improves feed quality, and gives readers a discovery experience that feels intentional rather than random.
Markers are my solution to a real problem. They prioritize context over chaos, and they put precision back into personalization.
Conclusion or Conundrum?
The way we interact with content online is shaped by tags, algorithms, and engagement which is signaled by likes and favorites. While these tools are meant to help us discover what matters, they often fall short when tags are misused or applied too broadly. The result is a feed filled with noise, displaying articles that share keywords but miss the context and intent that drew us in.
The solutions outlined here require more thoughtful tagging from authors, let readers favorite tags or write their own, and group tags into complete Markers for better matching offer a path toward a more meaningful online experience. By making these changes, platforms could shift from chasing popularity to delivering true relevance, giving users more control and satisfaction in what they see.
Ultimately, the goal is not just to fix a broken feed, but to create an online environment where discovery feels rewarding, not frustrating. As readers and creators, we can advocate for smarter systems that respect our interests and time.