Mastering Online Video Before YouTube Existed

There is no single video hosting platform to upload to in 2004. YouTube doesn’t exist yet, and won’t for close to another year. Anyone doing serious online video work today is dealing with fragmented playback formats, real bandwidth costs and a distribution problem that has to be solved manually. Here’s a practical playbook for working within those constraints instead of waiting for them to go away.

Step 1: Accept the format fragmentation and plan for it

Viewers are split across RealPlayer, Windows Media Player and QuickTime, and there’s no dominant standard forcing convergence. Producing a single file and hoping everyone can play it is a recipe for a large share of your audience hitting an error instead of your video. Budget the extra encoding time to export in at least two of the three major formats, and test playback on each before launch, not after.

Step 2: Study the precedent that already worked

BMW ran a series of short branded films called “The Hire” starting in 2001, directed by well-known filmmakers and starring Clive Owen, distributed entirely online rather than as a traditional television spot. The films reportedly drew tens of millions of views over their run, an enormous number for the era, by treating the content as something worth watching on its own merits rather than an ad interrupting something else. The lesson for anyone planning video this year: if the content only works because someone is forced to sit through it, it will fail online, where nobody is forced to do anything.

Step 3: Design for forwarding, not for interruption

The clearest proof of this principle in 2004 is Burger King’s “Subservient Chicken” campaign, launched in April, a simple website featuring a person in a chicken suit responding to typed commands. It wasn’t a polished film. It was built to be forwarded by email, one person to the next, because it was surprising and fun to share. Before you finalize a script or concept, ask honestly whether a viewer would send this to a friend unprompted. If the answer is no, the concept needs work regardless of production budget.

Step 4: Budget for bandwidth, not just production

Video files are large, and hosting bandwidth in 2004 is expensive enough that a video going unexpectedly popular can produce a real, unplanned bill rather than just a marketing win. Before launch, get a clear answer from whoever hosts your site on what happens if traffic to the video spikes tenfold: does the site slow down, does it go offline or does the bill simply grow. Plan a fallback, whether that’s a content delivery arrangement or a secondary host, rather than discovering the answer during a traffic spike.

Step 5: Build a distribution plan beyond your own homepage

A video sitting on your website waiting to be discovered won’t be. Identify the handful of communities, forums and enthusiast email lists relevant to your audience, and seed the video there directly rather than only linking from your own site. Email is still the dominant sharing mechanism for online video at this point, so make sure the video is easy to describe and link to in a short email, not buried three clicks deep in a site navigation.

Step 6: Measure what you can

You won’t get sophisticated analytics here. Track raw view counts, referrer logs showing where traffic originated and any qualitative signal you can gather, such as forum threads discussing the video or unsolicited emails from viewers. It’s imprecise, but a rough measure taken consistently is far better than no measurement at all.

None of this requires a television-sized production budget. It requires treating video as something people choose to watch and forward, planning around real technical constraints instead of ignoring them and having a distribution plan that doesn’t depend on people simply finding your site on their own.