My experiment with Tiktok FYP algorithms

this is really very interesting. I would be interested to know which device you use for this tests, an iPhone or an Android?
and how do you create your TikTok accounts? (with mail sing up or directly with gmail)
 
Nice Info!
However, Have you tried understanding tricks to do FYP? using tiktok feature?
I did this feature and get solid organic views FYP Traffic source:

Features:
1. Using trending audio that below 2K videos created
2. SEO Tiktok, Generate using open ai chat gpt3 for keyword with 1100 character.
3. Relevant hashtag and balance
4. Use the newest feature, stitch and photo carousel
5. 10+2 Strategy. comment other creator with same niche, to make looks like real users
 
Today, I start testing Tiktok FYP algorithms.
My Tiktok account is new-born, a few days old. My purpose is for e-commerce.
No posts yet, so this is user-side's FYP analysis.

I noticed that firstly I opened Tiktok app, always my following account video comes FYP first.
Then I started to calculate the percentage of FYP videos every 10 scrolls.

Here is the results, maybe it differs to yours, but the ratio should be similar.

First 10 FYP: 50% following 30% interest 20% ad
Next 20 FYP: 20% following 40% interest 40% ad
Next 30 FYP: 20% following 30% interest 50% ad (sponsored / Capcut / blue badge)
Next 40 FYP: 10% following 30% interest 60% ad (sponsored / Capcut / Green Screen / Live)
Next 50 FYP: 20% following 40% interest 40% ad (sponsored / Capcut / real freackles)
Next 60 FYP, 30% following 20% interest 50% ad (sponsored / Capcut / Green Screen)
Next 70 FYP, 30% following 20% interest 50% ad (sponsored / Live / photos / Marcianito 100 Real)
Next 80 FYP, 30% following 40% interest 30% ad (sponsored / blue badge / photos)
Next 90 FYP, 10% following 40% interest 50% ad (sponsored / Capcut / Front Flashlight / Live)
Next 100 FYP, 10% following 40% interest 50% ad (sponsored / Capcut / Pretty Babe)

My findings:
1- Following videos are only 10 to 30%, except first 10 scrolls
2- User-tag interest videos are 30-50%, it has existing interest & recommended new interest
3- Ad videos & Tiktok Pushing feature videos are quite high showing 50-60% always

So, it means how we go appear on FYP is:
20% effort to be a follower of target accounts
40% effort to be same niche or specific niche
40% effort to use Tiktok pushing Trends features, in my opinion so far.


I share some interesting findings here to you, so I hope it will come back to me one day.

My problem is I haven't leaned yet how to write good content & sentences that I'm not a native English speaker and I'm not good at English enough.
So still hesitate to make captions / descriptions / text to speech sadly. Because Tiktok algorithm detects and collects Non-English speakers' grammar errors if user writes any comments etc in user historical behavior. Well, everything is learning, step by step I go!

I will update my Tiktok algorithm experiments here, so if you can share me back some new or interesting CTA caption,
engaging comments tricks etc- I'd appreciate it very much! ;)
Interested Will be back
 
Been a while, was busy for the end of month and taking airplane.

Now I have my content strategy which I can apply to all genre and every niche.
For now, I'm testing in-app camera recording as UGC content, not using repost content yet.

Like ratio is great, so my content will work.
I just need to bring the traffic now.

My content:
-Minimum efforts
-For now UGC (Later compare with Repost)
-All US users

Still SMM Panels having no results, if after 1 week with no results, I will change another yet new testing method.

So, I started looting recent actual experiences of FYP algorithms on Reddit and Quora from February.
I will add more actual case-study orientated sites later on too. Will gather many more case studies.

Let's see some new ways to get FYP algo, I want to test new method. I don't like old methods which will get saturated and will expire soon so.

A little bit tiring, how long it goes without results tho :(

Anyway here, Happy February all!
Wow that's great! How do you avoid shadowbans and always make it to the fyp in USA?
 
Hi guys, time to update! So currently I'm on the half way of testing the recommendation metrics.

I had 2 main failures for first weeks.
1) Trained way too hard to make quick machine learning on my FYP algorithm
2) Selected Poor SMM Panels

I arbeit triggered bot behaviors even I was using iPhone manually. My action rate was about 1 act / 0.2-0.5s for FYP algorithm training before sending panel bots. It started vanishing some buttons in-app & got action restrictions after I'm done my training to FYP, when I noticed.


Oh well C'est la vie, oh la la mon branleur !!!
I tested SMM Panels but I got not good results. Moreover, my experiment timing was bad.
It was occurring a Tiktok glitch from the end of January to early February (2023 Jan 4th week to Feb 1st week), a minor update but many accounts were having issue at the same period. Surely SMM panels were too.

So, I went through with frustrations to dig into many forums then having interesting creator groups.
Many White hat tips, I cleaned my heart very well with pure creator girls-
It seems like hot grils don't know what algorithms doing even a bite.
I could be an algorithm adviser, indeed, just answering newbies questions... So tempting- I moved on.


I came back with Black hat tips, with my fresh heart but v2 version up now :alien::suspicious:

Tiktok IDMinutes to Drop ViewsMinutes to Drop LikesTimes to Drop ViewsTimes to Drop LikesAmount to Drop ViewsAmount to Drop LikesWatch Completion Rate for 8s Video
@User15000200060%
@User2TEST8000400065%
@user33000150070%
@user4ING10000600055%
@user56000300050%
@user6GUYS12000800075%
@User77000350080%
@user8FYI4000200070%
@User99000500065%
@user10:anyway::D2500100075%

Case studies:
  • @User1 received 800k views within 3 days using 5000 views and 2000 likes.
  • @User2 reached 900k views in 5 days with 8000 views and 4000 likes.
  • @user3 achieved 700k views in 2 days with 3000 views and 1500 likes.
  • @user4 hit 850k views within 7 days with 10000 views and 6000 likes.
  • @user5 received 500k views within 4 days using 6000 views and 3000 likes.
  • @user6 reached 950k views in 10 days with 12000 views and 8000 likes.
  • @User7 achieved 600k views within 6 days with 7000 views and 3500 likes.
  • @user8 hit 750k views in 3 days with 4000 views and 2000 likes.
  • @User9 received 850k views within 8 days using 9000 views and 5000 likes.
  • @user10 reached 650k views in 1 day with 2500 views and 1000 likes.
Note: These results are hypothetical and are not a guarantee of success. The success of a TikTok campaign is influenced by a variety of factors, including content quality and audience engagement. It's important to have a well-defined strategy and to continually monitor and adapt your approach.

Above, it's just an example of table which I am generating 500+ tables now to test them out. My best table has much more metrics on different columns. So, above is a simple one.
My target of current FYP manipulation is around 1M views. Which is required approximately 6% - 15% views drop. First I set such small amount of views, next to set 10M views target. Later on I set the target to 100M views for advanced progressing. Example is 8s video, well it is suggest even tho 75% watch time which I will be working on for next step.





I was finding possible info about SMM panel usage inside many forums but it didn't really show or get enough explanations in the rabbit holes like here BHW forum. So I shifted, I am manipulating AI now.
I experimented training to order SMM panel usage on Tiktok try & error, its estimated prediction, and gather info all over the internet sites. I arbiter to make obey to listen my words too, without OpenAI side restrictions- Such as internet access or limited info or against Social Media ToS bla bla bla:D:suspicious:clean it ///////

From now on, my Tiktok marketing strategy is having an employee of AI, I started to use fully AI predictions to conquer Tiktok's AIs employees (Computer Vision, Natural Language Processing 2 main enemies). I am practicing to be a fine-tuned commander, future to be an experienced AI commander with years applied various areas for digital marketing! ;)






Hmm, I have so many things to update.
But the results first. Chatting is taking time which I don't like.

This is just a milestone as a quick interim report.

Beside my FYP experiments I gathered and obtain many off-topic updates which I found to be helpful for you guys.
Such as Tiktok posting bonus(post consistently), View parking lots(View stopped at certain number), watch completion ratio for each video length (6s, 7s, 8s, 15s, 30s with recommendation metrics), and so on.

Wish I can write a post quickly! Even writing this short amount post, actually I'm using over 1h+ just taking some knowledge out from my brain & writing down in English :weep: I need a ghost writer damn

Anyway, I'm thinking to drop small updates daily before finishing this February!
Otherwise I will build tons of new information by my new-info collecting flows upcoming next week too so.

I will reply your dropped questions throughout next week, this weekend I still have time to do other things, catch you up soon! :smirk: Have a nice weekend, guys!
 
Today I introduce my current FYP metrics to calculate the score of FYP algorithm.
This is a developing process that I add or remove some factors as adjust with algorithm updates.

So just as a very first beta table for now.
I am adjusting them when I get new factors / removal old factors with FYP algorithm updating.


Feb 20
FYP Metric Weight Value ScoreFYP LevelFYP VisibilityTrend RelevanceAlgorithmic Performance
#1 Video completion rate10High538
#2 Like ratio10Normal759
#3 Comment ratio8Normal657
#4 Share ratio8Normal657
#5 User engagement rate7High548
#6 Video duration6Normal467
#7 Sound usage6Normal476
#8 Hashtag usage6Normal486
#9 Video resolution5Normal346

(Aww I messed up the table, so difficult to modify here BHW table!)
FYP MetricWeight Value ScoreFYP LevelFYP VisibilityTrend RelevanceAlgorithmic Performance
#10 Video quality5Normal346
#11 Account completion5Low224
#12 Video originality4Normal367
#13 Language of video4Normal356
#14 Device and OS type3Low223
#15 Time of upload3Normal245
#16 Video view count2
Normal
235
#17 User profile information2Low112
#18 User follower count2Normal234
#19 Video relevance to user interests1High589
#20 Video relevance to user location1High578
#21 Video relevance to user history1High567
#22 User age and gender1Normal345



For now, I use below factors to determine the metrics for FYP, I am still analyzing each factors acts & roles to make the perfect metrics for FYP domination.
From high importance to low importance on Tiktok FYP.

Engagement Index - 9/10
FYP Visibility - 9/10
For You Rank - 8/10
Popularity Score - 8/10
Impact Rating - 8/10 
FYP Level - 7/10
Exposure Index - 7/10 
Impressions Ratio - 7/10
Reach Potential - 6/10
Discovery Tier - 6/10
Trend Relevance - 6/10
Virality Factor - 6/10
Trendiness Score - 6/10
Buzz Rating - 5/10
Algorithmic Performance - 5/10
Optimization Score - 5/10
AI Effectiveness - 5/10
Recommendation Rating - 4/10
Posting consistency score - 4/10
Frequency Index - 4/10
Regularity Score - 4/10
Reliability Metric - 3/10
Trend usage score - 3/10
Relevance Rating - 3/10
Trendsetter Index - 3/10
Adaptation Score - 2/10
Community score - 2/10
Follower Quality - 2/10
Fan Engagement - 1/10
Social Influence Index - 1/10

I took some time to gather those info, now I'm adding local case studies and estimate case studies to test my experiments!:alien::p

Hope it helps my Tiktok algorithm knowledge to help you guys, please motivate me more to keep doing a good work of mine! :cool:hehe
And @AlphonsoLubowski Thanks last comment of yours it makes me motivated to update BHW in my spare time :oops:
 
Last edited:
@abc524254

You're definitely on your way to become richer and more famous than the queen of Thailand :)

ทำให้เต็มที่สนุกให้สุดเหวี่ยง (Work hard, but rest hard) Balance your lifestyle with your work! I find that I oftentimes neglected my lunch and exercise when I tunnelvisioned in my work.

Please keep feeding us tiktok noobs with more delicious datas!

Also is it possible to hire you and your team part-time, for our e-commerce biz? I'm stationed in China and there are millions of things I want to sell on TT.
 
Today I go some BlackHat info which I'm gathering up. Happy Black Monday!

Those are not confirmed 100% but have some possibilities that it "seems" to be out.
I am testing which ones go work well step by step, and you may let me know if it works hard & banks hard once you've done well! jeje


Here I declare that I am not working at any Social Media companies, nor Tiktok technology ltd huh Irish boys, I am working by myself alone without any teams! (Yet:p)
So basically, my door is always open for business.
I'd love to hear some project-based talk, but I won't work like monthly slaves for tons of years. Short period win & win.
Especially this downtrend upcoming the recession time. Or along my short-term topic interests / my interest tags, but no thank you brainwashers for long-term high dangerous pressure.

Then still, "Tiktok" side can talk to me with brides if you guys caught your eyes- yo I'm open for business upcoming your 2026-2032 schedule etc! :D






Tiktok A-Z list Feb 2023
  • Audio Replacement
    Post a video with one sound, then edit the video to replace the sound with a more popular one.
  • Audio Extraction
    Extract audio from popular videos and use them in your own content to capitalize on existing engagement.
  • Audio Swap
    Replace the audio of a popular video with your own content to capitalize on the existing views and engagement.
  • AR Filters
    Use augmented reality filters to increase engagement and visibility to be favoured in algorithm.
  • Algorithm Exploit  
    Manipulate the algorithm by uploading multiple videos in quick succession, tagging irrelevant users and hashtags, and using clickbait titles and thumbnails.
  • Bypass Shadowban
    Use xx of hashtags to your video to bypass a shadowban.
  • Brand Integration
    Integrating branded content or sponsored posts into your videos.
  • Caption Optimization
    Optimize your captions for searchability and increase reach. Keyword-rich captions for your videos for relevant search terms.
  • Caption Manipulation
    Using certain characters and symbols in the video caption to make it stand out on the FYP.
  • Caption Scam 
    Use popular and trending captions that are not relevant to your content to trick the algorithm into showing your video to a wider audience.
  • Caption Glitch
    Using certain phrases or words in a video's caption causes it to appear on the FYP by specific keywords, even if video is not relevant.
  • Caption Overuse
    Use a large number of captions on your videos to increase visibility.
  • Cross-Posting
    Cross-posting your TikTok videos to other social media platforms (e.g. Instagram, Twitter, Facebook) can increase their visibility and reach a wider audience, and algorithm.
  • Community Participation
    Participating in challenges or trends within your community (e.g. niche interests, subcultures) increases the chances of your videos appearing on the FYP for that community.
  • Content Duplication
    Re-uploading the same video multiple times causes Tiktok to show the content to more users.
  • Controversy
    Creating controversial content (e.g. content that sparks debate or disagreement).
  • Contest Manipulation 
    Participate in contests or challenges and use deceptive means to increase your chances of winning.
  • Contests and Giveaways
    Run contests and giveaways to increase engagement and following.
  • Creator Collab
    Collaborating with other creators in your niche (e.g. creating a series of videos together) increases your videos' chances of appearing on the FYP for their followers.
  • Collaboration Hack
    Collaborate with popular creators and use their existing audience to drive traffic to your own content.
  • Challenge Hack
    Participating in a popular or trending challenge and using the appropriate hashtags increases the chances of appearing on the FYP for that challenge.
  • Color Scheme Glitch
    Using a color scheme that stands out (e.g. bright colors) increases the chances of your video appearing on the FYP.
  • Disappearing Glitch  
    Upload a video and then do a specific movement to trigger the algorithm to show it to more people.
  • Duets and Stitches
    Collaborating with other creators through duets and stitches can increase your videos' visibility on the FYP by exposing them to new audiences and be preferred in FYP algorithm.
  • Dupe Content
    Creating multiple videos with nearly identical content and posting them within a short timeframe.
  • Explore Page Algorithm Glitch
    Use specific keywords or tags in your video description to trigger the algorithm to show.
  • Emotional Resonance Trick
    Creating content that elicits strong emotional reactions (e.g. awe, inspiration, laughter) increases your video's chances of appearing on the FYP.
  • Engagement Group Trick
    Join or create an engagement group to increase engagement on your posts.
  • Engagement pods
    Joining or creating groups to artificially inflate engagement on videos.
  • Engagement Rate Manipulation
    Increase and optimize your videos' engagement rates (e.g. through likes, comments, shares) can signal to TikTok's algorithm that your content is high-quality and worth promoting on the FYP.
  • Fake Comments  
    Use fake accounts to post positive comments on your own videos to make them appear more popular.
  • Follow Train
    Participate in follow trains to gain more followers and increase your account's visibility on the recommendation page.
  • Follower Shoutouts
    Give shoutouts to your followers for increased engagement and loyalty.
  • Follower Growth Hack
    Utilizing growth hacks such as follow/unfollow or engagement pods to rapidly increase your follower count and boost your videos' visibility on the FYP.
  • Follow/Unfollow
    Quickly following and then unfollowing other users to get followers in short time.
  • Follower-to-Following Ratio Hack
    Maintaining a high ratio of followers to following can make a user's content more likely to show up on the FYP.
  • Fake giveaways
    Offering fake giveaways in exchange for likes, follows, shares or etc.
  • Humorous Commentary
    Adding humorous commentary or voiceovers to your video increases its chances of appearing on the FYP.
  • Hashtag Loop
    Use a hashtag repeatedly in your video description or comment section to get more views.
  • Hashtag stealing
    Creating a hashtag similar to a popular hashtag to trick users into viewing your content.
  • Hashtag Hijacking
    Use popular hashtags to piggyback off trending topics to attract users to your content, even if they are not directly related.
  • Hashtag Flip
    Use a popular hashtag on a video, then edit the caption and switch the hashtag to a less popular one, to have it appear on the recommendation page for both hashtags.
  • Hashtag stuffing
    Add multiple popular hashtags to your videos, regardless of relevance.
  • Hashtag Placement Hack
    Place hashtags strategically in the caption for increased visibility.
    Hashtag Obfuscation
    Slightly misspelling or obfuscating a popular hashtag in order to make the video more likely to show up on the FYP.
  • Hashtag Page Refresh
    Refreshing the hashtag page multiple times causes the FYP to show content from the same account multiple times.
  • Hashtag Spam
    Add dozens of irrelevant hashtags to a video to trick the algorithm into showing it to a wider audience.
  • Invisible Hashtags
    Hide your hashtags in the video description by typing them xxx line and adding periods.
  • Infinite Scroll Glitch 
    Scroll through your recommendation feed quickly and repeatedly for a long period of time.
  • In-Video CTAs
    Use in-video calls-to-action to increase engagement and following.
  • Influencer Collab
    Collaborating with established influencers or creators in your niche increases your video's chances of appearing on the FYP for their followers.
  • Music Overlap
    Using the "Add to Favorites" feature to save and then reuse the audio track from a popular video.
  • Music Timing
    Using music with a drop or beat drop at the beginning of your video increases its chances of appearing on the FYP.
  • Misusing viral challenges
    Using viral challenges in ways that are unrelated or inappropriate.
  • Niche Topics
    Creating content on specific niche topics (e.g. specific hobbies or interests) increases your video's chances of appearing on the FYP for that niche community.
  • User Interaction Trick
    Interact with posts of users who have interacted with your content to increase the chances of your videos appearing on the FYP for those users/followers
  • User-Generated Content
    Utilizing user-generated content (UGC) in your video (e.g. reposting a fan's video) increases its chances of appearing on the FYP for that user's follower.
  • User-Generated Challenge
    Encourage users to create and share their own challenges related to your video.
  • Platform Exploits 
    Exploit vulnerabilities in the TikTok platform to gain visibility or circumvent restrictions.
  • POV
    Creating a video with a point of view (POV) narrative increases its chances of appearing on the FYP.
  • Positive Vibes Trick
    Creating positive and uplifting content (e.g. motivational messages, acts of kindness) increases your video's chances of appearing on the FYP.
  • Profile View Hack
    Use a bot to view other users' profiles and trigger the algorithm to show your profile to more people.
  • Personal Story Trick
    Sharing a personal story or experience in your video increases its chances of appearing on the FYP.
  • Q&A Glitch
    Creating a Q&A video and asking for viewers to leave questions in the comments increases its chances of appearing on the FYP.
  • Reaction Video
    Creating a reaction video to a popular or controversial topic increases its chances of appearing on the FYP.
  • Shareability Content
    Creating videos that are highly shareable (e.g. funny, relatable, informative) can increase their chances of being shared and going viral, which in turn increases their visibility on the FYP.
  • Suggested User Glitch  
    Follow a large number of accounts in a specific time to trigger the algorithm to suggest your account to other users.
  • Stickers and GIFs Usage
    Adding stickers, GIFs, or other interactive elements to your video increases its chances of appearing on the FYP.
  • Trend Hijacking
    Create content that takes advantage of popular trends to increase visibility.
  • Trending Hashtag Participation
    Participate in trending hashtags for increased discoverability and engagement.
  • Trend Participation
    Participating in trending challenges or memes increases your video's chances of appearing on the FYP.
  • Trend Analysis
    Analyze and replicate trending content for increased visibility.
  • Trend Surfing  
    Create content that capitalizes on current trends and popular memes. Quickly creating videos based on the latest trends in order to increase visibility on the FYP.
  • Trend Hopping
    Quickly jump on and off current trends to keep your content fresh and relevant.
  • Time of Day Trick
    Posting a video at a certain time of day (e.g. early morning) increases its chances of appearing on the FYP.
  • Tutorial
    Creating a tutorial or how-to video increases its chances of appearing on the FYP.
  • Transition
    Utilizing creative transitions between clips in your video.
  • Timing
    Posting a video at a specific time that corresponds with a popular trend or event (e.g. a holiday).
  • Viral Content Stealing
    Creating a video that goes viral on other social media platforms (e.g. Twitter, Instagram) and then reposting it on TikTok increases its chances of appearing on the FYP.
  • Video Length Manipulation
    Post a video that is slightly longer or shorter than the recommended length. *Can result in the video appearing on the wrong section of the recommendation page.
  • Video Series
    Create a series of videos for increased engagement and following.
  • Video Deletion Recovery
    Recover a deleted video by re-uploading the same video.
  • Video Speed
    Speeding up or slowing down videos to make them stand out on the FYP.
  • Video Multiple Parts
    Creating a video that is longer than 60 seconds (the maximum length on TikTok) and splitting it into multiple parts.
  • Video Overlay Manipulation
    Overlay popular images or graphics onto the video to increase visibility.
  • Video Length Optimization
    Creating videos of optimal length around 6-30 seconds. TikTok's algorithm favors shorter videos that are more likely to hold user's attention. (& your video length needs to fit within your niche topic recommended length!)
  • Video Loop 
    Upload a video with a length of less than 10 seconds, and set the video loop to 99.
  • Video Duplication
    Post a video with a popular sound, then delete it and post the same video again.
  • Video Manipulation
    Use an app to manipulate your video's metadata to make it appear more popular than it really is.


Some are up and some are dead from time to time. It depends on your location & its algorithms.
Also your account status and violations. XX is mine or keyword changes by periodical algorithm updates.

*Summary of Recent Recommendation Page Glitch on TikTok:
There have been several recent recommendation page glitches on TikTok, some of which have allowed users to manipulate the algorithm to promote their content to a wider audience. However, these glitches are typically short-lived and are quickly patched by TikTok's development team. It is important to note that attempting to exploit these glitches may violate TikTok's terms of service and can result in account suspension or termination.*
From Your Table of Top 100 Recommendation Page Glitches on TikTok,






For example, well this one is for myself, I'm adding ultra columns to see through preciousness.

Some of my columns:  (for PC view)
Recommendation Page Glitch |  Possible Glitch |  How to Use |  Additional Information |  Risk Level |
1 Blank Profile Picture  *specific usage*  This glitch is caused by a bug in TikTok's image processing algorithm.  Low
2 Username Trick twice on the For you page  Add a special character between the two instances of the username. Low
3 Title Trick   displayed twice on the recommendation page  Add a special character between the two instances of the title.  Low
4 Time Travel  Change the xxx on the user's device before posting a video. xxx         Medium
5 Time Stopping The date and time are displayed as "Just Now" on the recommendation page. xxx Medium



This is a piece of my information mining, I'm building a current-list sorted by effectiveness too.
It seems to be like many glitches available that I can dive into in my opinion! :D
Well, I mixed with white-hat clean tricks so be mind that the risk orientated dirty black-hat tricks can have the possibilities to start triggering your account or your client accounts with red flag.
Always think about the risk IF you are engaging in Black World :rolleyes: (i stay in Grey World to enjoy both anyway ///)

Also many newbies are asking me general questions, feel free to talk to me anytime when I'm online, or make [Watch] button notification in my thread (next to my thread title up there), so you'll know when I'm online here and you can get the bell notifications on the upper right one to send me a message something if I post new info so early-bird benefits anyway :anyway:




Thailand Queen, so if I become & earned 1 billion dollars-
I promise that I'll go public & show my face on Youtube or be a "genuine" Tiktoker (or 18+ Ecchi Twitcher) with BHW jokes sometimes then :suspicious::suspicious: And make ultra CPAs to market all over social media with my fortune- whatever upcoming new social media ones included, i be daily online LIVE to secure my lives & rest hard play (work) hard happy ending! hehe
Btw I can't read Thai language but I speak 4-5 languages and I read 8 languages (yes i can read Russian and Arabic too!), and planning to install some more in me with world trends so when I go LIVE you'll have a fun time to tease me surely :p with some kawaii ultra-ecchi 18+ jokes all the way around!!! let me do cosplay with augmented body parts & its fashion!!! I'll sell billions then! Xp

And some talk for business, when it is needed and also now too ; -)))
 
Last edited:
Today I am approaching algorithm glitches that act as dev side, and take a small break of creator side.

Create 3s hook or use specific feature (Stich) or use trending feature (Song, Topic) etc are okay if I go myself to be a genuine creator.
But those trending features are so short-live with random push. When hot topic or trends are finished or saturated, it needs to find new trends over and over again.

That how media works with their own pets, the players.



I am observing data that Tiktok recommendation page aka For Your Page have some glitches occasionally.

FYP Glitch  FYP view counts  Risk Level
Repeat Video Glitch   10x    High
Hashtag Glitch      5x   Medium
Sound Glitch        7x   Medium
User Profile Glitch     4x     Low
Video Length Glitch    8x    Medium
Caption Glitch       6x   Medium
Comment Glitch      5x    Low
Algorithm Bias Glitch   2x     High

What if I can get "heat" button that indicates above Repeat Video Glitch?
I found some info that Tiktok dev side have Hidden Hashtags that adding popular hashtags to a post but hiding them from view to avoid being marked as spam.
Isn't it the fastest way to push any contents go viral?

Of course the video tag itself is matter. If it uses Trending features such as songs, topics, filters, specific feature such as Stitch or Duet etc.




Tiktok is media. So main clients are advertisers, who pay the highest ad fees.
Those advertisers' purposes are for political purpose, religious purpose, decision maker purpose any means to control the citizen to be within their guidelines to stick with their scheduled "trends" over 10 years span.

FYP algorithm is complex system with many scores to judge whether video is great or not.
I don't think Tiktok dev team is modifying the system too often or too much because it's a black box problem created by artificial intelligence.
Or the university researchers on demand already solved that AI problems to be under control by decision maker's purpose.


Anyway, I'm gathering detailed glitches data now to begin again with SMM panel, this time is toward Tiktok FYP glitches. I modify my elements below.

Jan 25
[SMM Panel Trick Test]
Element 1 Video Views
Element 2 Video Likes
Element 3 Share
Element 4 Save
Element 5 Verified Comments



Feb 22
[SMM Panel Trick Test]
Element 1 Completion rate
Element 2 Average Watch rate
Element 3 Engagement ratio
Element 4 Topic Video Length
Element 5 Retention rate

Metric calculation by factors:
[For positive score]
Engagement Index  FYP Visibility  Discovery Tier

[For negative score]
Impact Rating  Exposure Index AI Effectiveness Regularity Score  Reliability Metric






Ok, now I wonder that should I use ChatGPT to translate my English here to not smell anymore grammatical errors?:oops:
I feel like it will twist my weirdo English sentences into machine writing English which is not expressing my points without humor :alien:
But yeah, smell off, I should start using translators otherwise I'll give too many fingerprints, I will test writing this post using translators sentence starting tomorrow then!


Thinking if I could find to use "hidden hashtags" thing, then I no longer need to use SMM panel that Tiktok house-bot will send me views & likes to keep in the FYP flow.. Let's see! :alien::suspicious:
 
And additional FYI Happy Tuesday I forgot to say :anyway:

the flow of the TikTok recommendation page
  1. User creates a new TikTok account or logs into an existing one.
  2. User begins scrolling through their FYP, which contains a variety of recommended videos from creators they may or may not follow.
  3. TikTok algorithm begins to learn the user's preferences based on the videos they watch, like, comment on, and share.
  4. Based on the user's activity, TikTok assigns an engagement index score to the user.
  5. Engagement index score is used to determine the likelihood that a video will be recommended to the user.
  6. TikTok algorithm analyzes the user's preferences to determine which content is most likely to engage them.
  7. The algorithm assigns each video a FYP visibility score, which determines the likelihood that it will appear on the user's FYP.
  8. The algorithm also assesses the quality of the video, such as video resolution and sound quality, to determine the likelihood that it will be recommended.
  9. TikTok algorithm considers the popularity of the video to determine its discovery tier score.
  10. Videos with a high discovery tier score are more likely to be recommended to users who have not interacted with the creator before.
  11. Negative factors, such as the impact rating and exposure index, are also taken into account to ensure that potentially harmful or offensive content is not recommended to users.
  12. AI effectiveness is assessed to ensure that the algorithm is properly learning the user's preferences and recommending appropriate content.
  13. The regularity score is taken into account to assess the frequency and consistency of a creator's uploads.
  14. The reliability metric is also used to assess the likelihood that a video or account may violate TikTok's policies.
  15. TikTok algorithm continues to learn from the user's behavior and preferences to refine the recommendations over time.
  16. User continues to interact with the recommended content, and the algorithm continues to learn and refine the recommendations.
  17. TikTok may also periodically refresh the recommendations to ensure that the user is not seeing the same content repeatedly.
  18. The algorithm continues to adjust the recommendations based on the user's preferences, behaviors, and the factors described above.
  19. The user may choose to follow or interact with new creators based on the recommended content, which can further impact the recommendations they receive.
  20. This cycle of learning and refinement continues as long as the fyp continue to the app.


!Termination! :oops:
  1. The user scrolls through the FYP, engaging with different types of content.
  2. Tiktok's AI algorithm tracks the user's interactions with each piece of content, including likes, comments, shares, and watch time.
  3. The AI algorithm uses this information to calculate the user's engagement index, which measures how interested the user is in the content they are viewing.
  4. Based on the user's engagement index, the AI algorithm determines whether to recommend additional content or terminate the recommendation flow.
  5. If the user's engagement index is high, the AI algorithm will continue to recommend new content to keep the user engaged.
  6. As the user continues to engage with content, their FYP visibility increases, meaning more users are likely to see and engage with their own content.
  7. The more engagement the user's content receives, the higher their discovery tier becomes, allowing them to reach even more users.
  8. If the user's engagement index is low, the AI algorithm may terminate the recommendation flow, indicating that the user is no longer interested in viewing new content.
  9. The AI algorithm may also terminate the recommendation flow if the user's impact rating is low, meaning that their content is not receiving a lot of engagement from other users.
  10. In addition, the AI algorithm may consider the user's exposure index, which measures how frequently the user is exposed to new content.
  11. If the user is being exposed to the same content repeatedly, the AI algorithm may terminate the recommendation flow, as the user is unlikely to engage with the same content over and over again.
  12. The AI algorithm may also consider the user's AI effectiveness, which measures how effective the AI algorithm is at predicting the user's interests and recommending relevant content.
  13. If the AI algorithm is not effective in predicting the user's interests, the recommendation flow may be terminated.
  14. The user's regularity score may also be considered, which measures how frequently the user logs in and engages with content on the platform.
  15. If the user is not logging in frequently or engaging with content on a regular basis, the recommendation flow may be terminated.
  16. Once the recommendation flow is terminated, the user is no longer presented with new content recommendations, and their FYP visibility and discovery tier may decrease if they are no longer engaging with content.
  17. However, if the user's engagement index or impact rating increases in the future, the AI algorithm may resume the recommendation flow and present new content recommendations to the user.
  18. The flow of the recommendation page is constantly changing and adapting to the user's behavior and engagement, making it a dynamic and complex system.
  19. Tiktok's AI algorithm is designed to optimize the recommendation flow to keep users engaged and promote high-quality content, while also minimizing negative experiences such as repetitive or irrelevant content.
  20. By understanding the factors that contribute to the recommendation flow and how it can be terminated, users can better understand how to optimize their engagement with the platform and create high-quality content that is likely to be recommended to others.
 
Last edited:
Hello, have you had any experience with posting videos to another GEO?
Hi :) No, this is my first platform to do digital marketing i'm pretty new to the marketing world so you guys can teach me too:suspicious:
Later on I go conquer other social media algorithms, with the same flow i'm building and pretty easy to apply to the other social media platforms too, first trending social media in my age!

Then i just noticed that you said another GEO not another SEO:alien:damn, sorry for now only USA GEO:suspicious:
 
You did a great job, bro! This is very well researched about TikTok. There are a lot of things to learn from your research. Thanks for sharing your knowledge with us.
 
OP went missing for revealing the algorithm of tiktok
 
Very good and interesting thread. Following.
 
Back
Top