I’ve been thinking about trailers - as you do.
(Well, actually, if you’ve got a feature film coming out that took four years of your life to make, yes, you do!).
Now, I and other Machinima creators have tended to go one of two ways with trailers:
Release them well before the release of the film, usually whilst you’re still in production or,
Don’t have one.
Option 1) is definitely a bad idea. For starters, if you’re still developing the look of the film, your trailer won’t be as slick as your finished product - and your trailer should ideally be the slickest element of your film package. Secondly, no-one ever gets their time estimates right on a Machinima product, for the same reasons that no-one ever gets the time estimates right on software development - too many variables, too much variance. That means that your trailer will come out months and months before your film, in all likelihood - giving people time to forget it.
(I’ll have a piece up on scheduling and estimating Machinima production at some point, but the summary is “don’t if you can help it”.)
At best, a trailer beforehand might help build up some anticipation, but it isn’t guaranteed - and worst case, you’ll actually convince people that your production quality will be lower than it is.
(We speak from experience. The BloodSpell trailer was not well timed.)
Option 2) is a common one, but it’s also not a good idea if your film’s more than about five minutes long. Just as there’s a steep cut-off point for download sizes (a 10 Mb download will attract 5-10 times the number of downloaders as a 100 Mb download), I’m fairly sure that there’s also a cut-off point beyond which people won”t experiment with a long film. I’m not sure what it is, yet, and I think that it’s longer than you might expect, but it’s there - particularly with an unknwon quantity.
And yet, the reluctance to download or watch goes away if you’ve got a track record. Brandon “Oxhorn” Dennis, for example, recently released a 12-minute film, the third in his “Inventing Swear Words” series - longer than I’d normally expect a peak-popularity film. However, he’s getting remarkable figures - 83,000 views in the first week.
So why trail? Because it gives people a “teaser” work to get some idea of the quality they can expect. Given that even if you’re a prolific Machinima creator, there will always be new people coming to your films, having some kind of short work is a good way to go - and having a trailer, particularly if you update it for each part in a series, say, is a good way to ensure that the first part of your work that they watch is the best (a critical problem that we had with the original BloodSpell series - Episode 1, being the first one we made, was by far the weakest episode).
But again, when to release the trailer? Well, taking everything into account, it’s seeming to me that the best time to release a trailer is at exactly the same time as you release your film. It provides a “proof of concept” as mentioned above. It provides a quick way for lazy or harried media types to get a sense of whether you’re worth bothering with. And it hopefully ensures that curious people arriving at your site but not wanting to dive straight into a long piece will look at a recent work full of your best efforts, rather than hunting around for a shorter film you’ve made (which was probably earlier and therefore worse).
We’re going to be trying this approach for BloodSpell. We’ll see how it goes.