The Brief:

Apture was a contextual search startup (now acquired by Google) that allowed users to gain in page, rich media search results through text highlighting. At the time, Apture was just about to release Apture Hotspots and was seeking new strategic directions for its product. I was brought on to provide strategic thinking about potential new avenues for Apture to explore.

The Problem:

Although Apture had great reach, it had to strike a delicate balance between providing utility when needed while also not being annoying when not needed. This tension meant that the designed interaction was so unobtrusive that engagement was low. Apture Hotspots was one new way of bringing the Apture experience to a broader range of people. What were some other ways to balance these two competing needs?

One of the ideas that I explored was in how best to surface ambient behavioral information from people that you cared about. For example, if you were reading about Cleopatra, would it be useful to know a friend had also been researching Cleopatra a few months ago? When & how would this be useful? What would the privacy norms around this be and what would be the best way to present this to the user?

The Solution:

In order to gain insight into these questions, I built Ambient-Wiki, a prototype to be used by internal Apture employees only. Ambient Wiki used a browser plugin to log every Wikipedia page that any Apture employee visited and broadcast this only a shared, ambient display in the office. Simple annotation features were also provided and annotated entries would make an entry extremely prominent on the display.

It was decided that Wikipedia entries were innocuous enough that privacy was not a major concern but interesting enough that they provided an insight into co-workers curiosity.

Findings:

Over the 4 week period that AmbientWiki was running, we discovered that the threshold required to start a conversation was relatively high. However, people did enjoy the voyeuristic element and paid attention to what others were searching. AmbientWiki primarily sparked conversations when you saw someone else searching for something that you have knowledge of (rather than the other way around as we had thought). Unfortunately, it wasn’t possible to integrate these findings into any solid product direction.

My Contribution:

  • I conducted several user studies on Apture’s existing product
  • I created animated interaction mockups of new interactions for new bottom bar and close window behaviors
  • I coded the AmbientWiki internal app, including a Chrome Extension, a Ruby on Rails backend and a HTML+CSS+JS frontend.