I just read a post by Matt Locke, formerly of C4, that makes a lot of assertions regarding particpatory narratives in general, and alternate reality games in specific, that I find... puzzling.
It feels sometimes that designers of interactive story projects spend all their energy on Beginnings and very little on Endings. Most ARGs have very few users who make it through to the end, either by accident or design, and so the cathartic experience of coming to the end is not all that common in participatory stories.
A few things come to mind here.
First, the flip answer: If you don't spend a lot of energy on your beginning, you won't be hooking anyone into your ending, no matter how spectacular it is. Spending less attention to rabbitholes is a great way to shrink the size of your audience.
And that's not even going into the analogy with the magic trick he makes a bit later; a magic trick is something you do to an audience. Participatory narrative is something you do with an audience. Any comparison is inherently flawed.
But digging for a more substantive answer, as they say in Wikipedia: Citation needed. Particularly on this part: "...very few users make it through to the end." I beg your pardon?
It's true that in an ARG, the amount of people who make it to the end are fewer than the number that see the beginning. But this isn't because we're somehow bad at telling endings; it's because there is no cost to trying it and deciding it's not your thing. The number of people who pick up a book and put it down without buying is inevtiably much larger than the number of people who buy the book. This is also true of television and video games; there is no social or economic penalty to quitting.
It also misleads into the idea that ARG players tend to follow along and then drop away in the middle, which is very much not my experience. While I'm not at liberty to divulge numbers for many of the projects I've worked on, the pattern I see isn't that users come in at the beginning, get bored midway through and give up. Attrition happens early. By and large, if a player is with you during your midgame, they will stay through to the ending (unless you do something appalling.)
And without exception, every game I know of was still collecting new players through to the finale, to such an extent that working out how to catch up new players to the story so far is a widely acknowledged structural challenge in ARG design.
And the endings themselves? Well, for a marketing campaign, often those endings are the thing being marketed. For 2012, for True Blood, for The Dark Knight, for A.I., the entire game was a huge windup ticking inexorably toward that big ending, making the entire film or TV show all payoff. Secrets revealed, punchlines told, explanations proffered.
For standalone content, Routes had a finale where a team of players broke into a lab while other players watched the stream and assisted from online. Perplex City had a kidnapping, tortured revelations, and a breathless treasure hunt in the woods. And while these are both my projects, I am confident that my work is nothing special in terms of exciting conclusions.
Now, I'll admit that there are projects out there that end badly. If nothing else, there have been no few grassroots games that fizzled out midgame and were never completed. But one doesn't judge all of literature from unfinished novels; one doesn't judge cinema on student films; and the state of the ARG shouldn't be judged based on hobbyist efforts. Beyond that, I'd really love to know where, exactly, are these projects he speaks of that pay inadequate attention to endings.