By skyforbes 
I frequently use Douyin to watch short videos and occasionally post snippets of my daily life and favourite landscapes. Over time, it has become quite “personalised”, like it knows me interested in sunsets, daily vlogs. But TikTok is different. I rarely use it, so I mostly get global trends and various memes and other entertaining clips. Fun, but not to my taste.
Why do two short-video apps feel so different? This article uses a platformisation lens to show how algorithms and my own habits together shape what I see each day.
Douyin: When My Feed Knows Me
I mainly watch travel shorts and daily vlogs on Douyin, occasionally scrolling through collections of aesthetic photos. I watch, like, sometimes save, and I follow a few favourite creators.
Within a few days, my feed tilts toward that trail. It fills with lifestyle vlogs and vibe photo collections. They all feature similar camera angles, slow music and minimal captions.
One night I watched three travel vlogs in a row and saved two. The next morning my feed opened with near-identical videos. The compositions were comparable, the colour palettes similar and even the audio vibe felt the same. I stayed longer and saved another, tightening the loop again. Sometimes I can even predict the next clip. If I’m into cooking that week, the odds of seeing cooking videos on the homepage are quite high.
Why does this happen? My small actions including finish, linger, like, save or follow are all recorded as data points. The system clusters those signals with “similar users” and “similar clips,” then feeds them back to me. That back-and-forth is a feedback loop; turning everyday behaviour into comparable, calculable data is datafication. Combined, they’re how platformisation shows up in my routine.
It’s comfortable. I get what I want fast. It optimises for my past behaviour, not my next question. But the feed also gathers around one mood and unless I actively adjust my preferences, I barely notice the boundary. I’m a heavy user on Douyin, so the loop stays tight and the feed feels personal. On TikTok, I don’t show up enough for it to really know me.
TikTok: A Window to Trends, Not to Me
I rarely use TikTok. Maybe once or twice a week, for five to ten minutes while I’m in a queue or waiting for a bus. The For You Page is mostly global trends and challenges such as popular dances, comedy edits, sports highlights, street pranks. The tags and sounds are all the rage, but most of it isn’t what I usually want to watch.
Once, I scrolled past three dance challenges in a row using the same sound “Just Now”. The comments were funny and the views were huge. I hit like and kept moving. The next day, I opened the app again and saw the same type of video, just different creators and cities. Fun, but it didn’t hook me. I even tried to train it by clicking “Not Interested” on several trending videos and following two small travel accounts. Then, my “For You” page did surface a couple of niche city vlogs, but the viral memes still appeared. It felt like a new shelf appeared, but the shop itself remained unchanged.
Unlike on Douyin, I leave very few signals on TikTok. I rarely finish, linger, save or follow. The system tends to show me what’s trending first rather than tailoring the feed around my taste. As a result, TikTok feels fresh, varied and fast, but not personalised. With thin signals, the safest bet is trends. Personalisation only kicks in if I keep feeding it.
From the platform’s perspective, this is an exploration versus personalisation trade-off. When user signals are thin, the system starts with what’s currently popular to keep me from getting bored, then watches what I choose. That is the cold-start logic at work. Because I open it infrequently, the loop never really forms, and the algorithm does not learn me well.
TikTok has its own appeal. It shows me that cut across languages and regions, it drops me into other cultures for a moment. But for me, it feels more like strolling through a shopping centre than visiting a coveted boutique.
When Small Moves Matter
I sometimes post on Douyin, mostly everyday photos and landscapes. At first, I posted at irregular times, but later I noticed that views were steadier after 6 p.m. I pick non-intrusive music and add tags that fit the content and are also trending, like #sunset or #Sydney.
I’ve posted a few photo series of daily life. Some picked up quickly, others went quiet. Over time, I noticed several differences: posting time, the first cover photo, and tags. In general, posting around dinner time with a good cover, and clearer tags leads to faster view growth and more saves. But it isn’t a rule. I’ve also had a casual midday post take off for no obvious reason. In other words, these moves improve the odds but don’t guarantee outcomes. The first wave of interactions fluctuates, and even similar posts can be distributed very differently.
This isn’t luck. Platforms organise attention and keep optimising based on signals like completion rate, replays, and interactions. Every small adjustment such as timing, length, audio, tags are all changes that first wave of interactions, and that first wave largely decides which tier of audience a post reaches. The system isn’t a pure black box. It’s feedback sensitive. I feed it signals, it adjusts weights and sends the result back.
This also explains my observation that posts in the evening are more likely to go viral. During peak hours, when more users are online and interactions are denser, early engagement gets amplified more readily. The platform serves users, creators, and advertisers at once. It’s a multi-sided platform. Creators’ tiny choices and the recommender work together to shape the fate of each video.
What I Can Post, What You Can See
Recommendation is only one layer, rules shape the feed too. They dictate what I can post and what you can see.
When I post photo sets, I avoid music with potential copyright risk and use clear, specific tags like locations and theme, not vague words. Sometimes I see folded comments or a post auto-routed into a certain topic tag. It isn’t always “throttling”, but visibility does shift. Once I used an old song to post a photo. After publication, the platform automatically replaced it with a currently popular soundtrack. The rhythm of the clips shifted accordingly, until a friend commented “Something feels off.” I just realised that the content I posted on the platform wasn’t being edited by myself. The platform’s moderators were editing it too.
From a platform point of view, this is governance working alongside optimisation: platforms organise attention while community standards and copyright draw the boundary. Copyright is not merely a matter of legality, it’s a copyright logic that shapes distribution. What music is usable, which clips get muted or restricted are often set by frameworks negotiated between platforms and rights holders. Meanwhile, platform culture and moderation also shape what rises or sinks, some styles are more likely to be amplified or dampened.
For me, that means I’m not only teaching the algorithm to understand me but also learning the platform’s grammar. I choose non-intrusive audio, reusable tags, clear covers, and avoid crossing copyright or platform guideline boundaries. It isn’t flattery, it’s recognizing that governance and distribution are the same system. Whether content reaches you depends not only on what I publish, and on how rules are coded into recommendation. When new-media elites lobby, they help decide how copyright and data rules are written. And then those decisions turn into the defaults we click through. Therefore, many decisions are not mine to make. The platform makes them for me. It’s putting comfort and efficiency first and diminishes opportunities to discover novel perspectives.
Common Claims, My Replies
“Douyin only feels accurate because you use it more.”
Indeed, usage intensity matters. But that’s exactly how platformisation works. The more signals I provide, the faster the system optimises, feeding me similar content. The more I watch, the tighter the feedback loop becomes. A multi-sided platform wants retention and monetisation, so it keeps organising attention around what already works. I’m not only choosing content, but the content is also choosing me.
“TikTok is better because it’s more diverse.”
Diversity is great. I like being dropped into other cultures for a minute. But the default optimisation leans toward quick interaction, so exploration often stays superficial unless I actively train it by following, saving, lingering on unfamiliar stuff. Platforms could surface diversity controls or into the defaults organising of attention. In fact, many defaults make that trade-off for me rather than letting me make it.
I’m not to pass judgement on which platform as ‘good’ or ‘bad’. What matters more is my feed is co-produced by my habits, platform incentives and optimisation, and its governance rules. Knowing that helps me use them with a lighter grip. Lean into comfort when I want it. When I want range, follow outside my comfort zone, save an unfamiliar topic, and sometimes clear parts of my history. This way, I preserve comfort while carving out space for freedom. Online freedom is never absolute. Kelty argues that it’s built into the system, enabled and limited by the same infrastructure that keeps it running.
Beyond My Feed
Looking back, the distinct difference between Douyin and TikTok isn’t just about my taste. It’s how platformisation works at a larger scale. Algorithms, governance and habits knit together to organise attention and culture. For me, this represents a trade-off between comfort and exploration. While for the public, it shapes what we see every day and how we understand one another.
When Douyin holds me close to a familiar vibe, diversity drifts to the side. TikTok drops me into global trends, but often at the surface. The contrast is clear, platforms aren’t neutral tools. Their recommendation mechanics and governance make choices on my behalf.
This prompts a broader question, where should our future trajectory lie? Should we continue relying on personalised recommendations for comfort, or irrelevant items to keep diversity alive? Platformisation gives convenience with one hand and tightens boundaries with the other. Noticing that might be the true purpose of this piece.
I won’t ditch either app. What I can do is hold the wheel a bit steadier. When I want comfort, I let the feed stay close and familiar. When I crave novelty, I nudge it. Following one outside my bubble account, save a post from an unfamiliar topic, hit “not interested” on repeats, and clear a slice of watch history now and then. Small actions can alter rhythm of the feed. Each week I follow at least one account from a category I don’t usually engage with. At the end of each month, I tap “not interested” on the repeats to nudge the pattern. I allow myself to save one post on a topic I know nothing about. They’re small moves, but the feed shifts.
Platformisation isn’t going away. It’s part of everyday life. Seeing how algorithms, governance and my habits co-produce the feed doesn’t kill the fun. It gives me room to choose. If the web should work for people, the opportunity of choice shouldn’t be left to defaults. If there really is a slider between personalisation and diversity, I want to be the one controlling it.
References
Berners-Lee, T. (2020). 30 years on, what’s next #ForTheWeb? Web Foundation. https://webfoundation.org/2019/03/web-birthday-30/
Gray, J. E. (2020). Google rules: The history and future of copyright under the influence of Google. Oxford University Press. https://doi.org/10.1093/oso/9780190072070.001.0001
Internet Society. (2014). Who makes the internet work: The internet ecosystem. Internet Society. https://www.internetsociety.org/internet/who-makes-it-work/
Kelty, C. M. (2014). The fog of freedom. In T. Gillespie, P. J. Boczkowski, & K. A. Foot (Eds.), Media technologies: Essays on communication, materiality, and society (pp. 196–220). MIT Press. https://doi.org/10.7551/mitpress/9042.003.0014
Massanari, A. (2017). Gamergate and The Fappening: How Reddit’s algorithm, governance, and culture support toxic technocultures. New Media & Society, 19(3), 329–346. https://doi.org/10.1177/1461444815608807
Popiel, P. (2018). The tech lobby: Tracing the contours of new media elite lobbying power. Communication, Culture & Critique, 11(4), 566–585. https://doi.org/10.1093/ccc/tcy027
van Dijck, J., Poell, T., & de Waal, M. (2018). The platform society as a contested concept. In The platform society: Public values in a connective world (pp. 5–32). Oxford University Press. https://doi.org/10.1093/oso/9780190889760.003.0002