29
- June
2009
Posted By : Adrian Chan
Answer services: satisfying two user experiences

Interesting article about Aardvark in the New York Times over the weekend. In
Now All Your Friends Are in the Answer Business, author Randall Stross covers a few “answer businesses” alongside Aardvark’s recently public IM service. Yelp and Yahoo Answers are mentioned. Mahalo’s Answers is not, but could have been.

The answers space is a difficult one to crack. It continues to attract companies interested delivering a solution to what would seem to be a fairly straightforward problem. Given all the people connected online at any given moment, why can’t I get an answer to my question from a person who knows the answer?

Think about it: Google searches documents. It is up to the user doing the search to qualify search results, and try again if first results aren’t the right ones. In other words, the user doing the search knows the answer, or knows enough to recognize it, or its source. Google doesn’t do well at questions that a) the user can’t recognize a qualified answer to b) involve some further contextualization or “human” assistance.

All questions are not alike. The ones that we pose to search engines tend to be a particular kind of question: a query. Queries can be captured, parsed, and addressed by a combined use of data and algorithms. Matches, made relevant by means of filtering sources and sorting the order of results (popularity, authority, recency, etc). But the user experience of a google search is still more a quest than it is the posing of a question: users are given enough to continue to look through results for what they need.

The problem addressed by answers services is a completely different one. It involves matching a human answerer to the human asker. Two minds, not one. This ought to provide an advantage. Two minds ought to be able to tackle issues that a search engine cannot: whether the question is properly phrased or stated; whether it is properly directed; whether it can be finessed or better constrained. The question can create an opportunity to refine the question. The asker asks the answerer, who in turn asks the asker for more information, and so on.

But this also creates issues. There are now two user experiences to address (and satisfy). Furthermore, the user experiences are asymmetrical: the asker has a clear motive (his or her question, in mind, here and now); the answerer has no motive (until being asked a question, the answerer was presumably thinking about something completely different).

Furthermore, this coupling of asker-answerer creates a dependency.

  • The asker’s question determines the experience of the answerer: a dumb question asked of the answerer does not provide a shot at stardom.
  • And the answerer’s answer determines the experience of the asker: A dumb answer does not satisfy the asker’s pressing need.

Satisfying both user experiences, not just one, is the challenge faced by any answer service.

Which is why many wrap the user experience in a conversational or social practice for better results. It’s a rule of social interaction that participants can use the rules of grammar and linguistic communication or the rules of ritual and etiquette to negotiate an interaction. If the communication fails (question is misunderstood), participants can use etiquette to reframe the matter at hand, clarify intent, need, and so on. Social rules complement linguistic rules — and all social interactions are a fragile combination of the two at work, hand in hand, and generally unconsciously.

The success of an answer service hinges as much on satisfying the answerer as in satisfying the asker. If answerers have no incentive to answer questions, the system fails due to lack of participation. The wrapper used by a Mahalo Answers or Yahoo Answers is social. Both incentivize the answerer by providing a means by which the answerer can build reputation, expertise, points, and social visibility/status.

Aardvark wants to leverage friend relations and use those as constraints on the interaction, presumably avoiding the need to incentivize users socially because friend relations trump social (remains to be seen) relations.

Answers services are a good example of why social interaction design needs a two-sided accounting of the user interaction. They’re an exemplary case, in fact, because answer services are conversational and interactive in their very definition. There’s no clearer example of inter-subjectivity than one subject requesting something of another subject. But the ask-answer situation extends to all social media interactions. Many if not most user activities provide the possibility for response by another user, if not soliciting response then suggesting how a response may be taken. Many, if not most, user activities on social media imply this social dynamic and perpetuate social interaction precisely because users are capable of making claims and responding to claims.

Interactions on social media do not just concern the user’s interaction with the software application. They concern the user’s interaction with another user, mediated by the software application. The social dynamics on which a successful social media experience depend require the satisfaction of interdependent, interacting, and inter-subjective user acts and actions. Answers services are a good example of this dynamic at work, and are an example of how difficult it can be to successfully align and resolve two distinct user interests at the same time.

Comments

Leave a Reply