Andrew Chen

Jim is so smart
Jim Young, a co-founder of HotOrNot, commented on the recent blog I wrote on reward schedules. It’s so insightful that I wanted to highlight it.

Jim writes:

Good insights- you nailed this one
on the head. With regards to HOTorNOT, the primary reward the next
photo to be rated. The post-rating screen communicates the community
aspect- "what others thought", "last checked their score X minutes
ago".

You can also consider the rewards schedule to be variable ratio,
since the next person isn’t guaranteed to be hot. If you have ever
played with any of the HoN clones that only show attractive people all
the time, you will find the experience is less addictive.

The second paragraph is incredibly interesting. I think a typical developer building a HoN clone might think that rewarding the user with the hottest matches constantly gives the best experience, because they look at HoN as a goal-driven website rather than an entertainment experience.

HotOrNot as a functional app
If you think about HotOrNot from a completely functional standpoint, then MATCHING RESULTS are the key driver. The goal is to present the user with hot people, as to reward them as quickly as possible. The best possible experience is to have the user show up, burn a couple pageviews looking at the "best" people, and then leave the site.

HotOrNot as an experience
But if you think of HotOrNot from a user engagement standpoint, then it’s not the matching results that are key, but rather TOTAL # of RESULTS or in other terms, more pageviews = better. This also, of course, aligns with the goal of advertising revenues.

In that case, then you want to throw some mediocre-looking folks in with the good ones, to keep people guessing and clicking away. Instead of showing 10 super-attractive people in a row, you spread them randomly across 50 profile pictures, thus increasing your engagement by 5X.

The entire switch here is to think of these websites not as collections of features or products, but rather manufactured experiences that are designed to be compelling wastes of peoples’ time

Press button to WIN
The entire conflict between engagement and functional usability can be encapsulated by the following statement (which I’m stealing from someone, I don’t remember who):

If you handed a game over to usability folks, they’d make a big red button that said, "Press button to WIN"

That is, if you optimize for helping people accomplish a task as quickly as possible, you start missing out on the entire path of accomplishing the goal, which is what drives engagement. Maybe where I should reference the cheesy self-help quotation: "Life is a journey, not a destination"