Guides
Part of Video platforms: what to keep and what to drop as of 2027
Video platforms by year: the constraints that lifted, in order
Bandwidth, storage, camera and screen lifted in sequence, and each one left behind a habit around length, format and audience that video platforms still carry today.
What to take away
- Bandwidth 1991 to 2005, storage 2005 to 2008, delivery 2008 to 2012, the camera 2007 to 2013, the small screen 2010 to 2016, live video 2011 to 2016, and selection 2016 onward.
- Marker platformsQuickTime 1991, RealPlayer 1995, YouTube 2005, Netflix around 2008, iPhone 2007, Instagram 2010, Snapchat 2011, Twitch 2011, Vine 2013, Facebook Live 2016, TikTok 2016.
- The order is causal. A cheaper input becomes affordable, a form grows around it, and the form outlives the reason.
- Selection is the exception. It changed nothing about production and everything about who decides.
The order, in one table
| Constraint | Years | Platform that marks the lift | What it forced | What survives after it lifted |
|---|---|---|---|---|
| Narrow bandwidth | 1991-2005 | QuickTime 1991, RealPlayer 1995 | Short, low quality clips | An assumption that attention is scarce |
| Expensive storage | 2005-2008 | YouTube 2005 | Length caps, quality ceilings | Length norms set by an old payment rule |
| Slow delivery | 2008-2012 | Netflix adaptive streaming around 2008 | Download before watching | Impatience with buffering that no longer happens |
| Costly cameras | 2007-2013 | iPhone 2007, Instagram 2010, Vine 2013 | Few producers, high polish | A belief that production value predicts reach |
| Small silent screens | 2010-2016 | Snapchat 2011, Instagram Stories 2016 | Captions, vertical frames, front-loaded hooks | Hooks on work nobody was going to scroll past |
| Live from a phone | 2011-2016 | Twitch 2011, Facebook Live 2016 | An event with a chat attached | Archives treated as a by-product |
| Who chooses | 2016 onward | TikTok 2016 | Immediate performance with strangers | A standing start on every item |
Constraints and what they left behind
The limit
- Narrow bandwidth
- Very short, low quality clips
- Expensive storage
- Length caps, quality ceilings
- Slow delivery
- Download before watching
- Costly cameras
- Few producers, high polish
- Small silent screens
- Captions, vertical framing, hooks
What it forced
- Narrow bandwidth
- Assumed attention is scarce
- Expensive storage
- Length norms from a payment rule
- Slow delivery
- Impatience with buffering
- Costly cameras
- Belief production value predicts reach
- Small silent screens
- Hooks on work nobody scrolls past
What survives now
- Narrow bandwidth
- Expensive storage
- Slow delivery
- Costly cameras
- Small silent screens
Reading the sequence for a service you are choosing
You do not need dates. You need the constraints a service still imposes.
- Long piecesa hard cap, a soft penalty, or nothing. YouTube's published limits allow 12 hours or 256 GB for verified accounts, while Shorts stops at three minutes. Treat both as dated figures.
- Your master filelength, aspect ratio and re-encoding decide whether it survives contact.
- The connectiona service built around expensive data behaves differently, and seven notes on video platforms by country explain why adoption diverges.
- Sounda muted default means the work has to survive being watched silently.
- The audiencewhere it comes from decides whether a back catalog is an asset.
Bandwidth and storage, 1991 to 2008
Apple shipped QuickTime in 1991 and RealNetworks shipped RealPlayer in 1995. Both assumed a file had to arrive before it could play. A clip was small, short and heavily compressed.
YouTube launched in 2005 and took the hosting bill off the creator. Its first public upload ran 19 seconds. Every upload was stored several times over in several formats, which is why length caps were normal.
Google bought YouTube in 2006 for $1.65 billion. Vimeo, launched in 2004, sold the other deal: higher bitrates and no ads on paid plans.
Delivery and the camera, 2008 to 2013
Adaptive bitrate streaming, where the player switches quality as the connection changes, removed the last reason to keep things short. Netflix moved to it around 2008 and YouTube followed.
Netflix went streaming-only in the United States in 2010 at $7.99 a month. YouTube moved to 1080p playback the same year and added 4K later.
Apple's 2007 iPhone put a usable camera in every pocket. Instagram launched in 2010 and added 15-second video in 2013. Vine, bought by Twitter in 2012, shipped six-second looping clips in 2013 and closed in 2017.
Which habits still pay is taken up in the video platforms review of what to keep and what to drop.
The small screen and live video, 2010 to 2016
Watching moved to a screen held in one hand, often in public, often with the sound off. Vertical framing, large text, burned-in captions and silent openings answer that. Captions are also an access duty, written down in the United States federal accessibility rules.
Snapchat, launched in 2011, made the vertical disappearing post the default. Instagram Stories copied it in 2016. The opening seconds became an argument for watching the rest.
Twitch launched in 2011 as Justin.tv's gaming channel, and Facebook Live arrived in 2016. Live made video an event with a chat attached and made moderation a real-time job.
Selection from 2016, and the split it created
TikTok launched in 2016 and absorbed Musical.ly in 2017. Musical.ly had launched in 2014 on lip-sync clips and a following requirement, which the For You feed dropped. Instagram Reels followed in 2020 and YouTube Shorts in 2021.
Nothing about production became possible or impossible. What changed is who decides.
Under subscription-led distribution, the audience you accumulate is the audience you keep. Under selection-led distribution, every item competes from a standing start with every other item. The wider arc is set out in how the shape of these services changed.
Disclosure rules followed the money and now cover micro-influencers as well as large accounts, as FTC guidance on influencer disclosures sets out.
Habits that outlived their limits
Length norms formed around a payment rule that changed long ago. Buffering impatience persists where nothing buffers. A belief that production value predicts reach survives from an era when it did.
Test your habits against the table. A convention that costs time and solves a problem you no longer have is worth dropping.
