It became very clear by day three of FASTforward ‘09 that there has been a lot of deep thinking around search experiences. It has been FAST’s commitment to deep thinking (reflected by their support of this blog and the format of the FASTforward events) that has continued to gain my respect for them as a vendor.

On Monday of FASTforward ‘09, some of us got to participate in a thinking exercise expressed via a mindmap. The focus of the discussion was “Search-Driven Experiences”, clearly influenced by the messages from Jared Spataro’s talk to us that morning, especially:

“Turn information into business outcomes through engaging experiences:

• Visual — Help you identify patterns and discover new insights
• Conversational — Change how you interact with information, giving you better answers
• Actionable — Allow you to rapidly turn information into actions.”

I share the map my group did, not because of ownership, but because of familiarity with the journey and the insights along the way. With really great partners at my table to work on this exercise, I had recently started a similar list for “elements of a social economy” and pulled it out to see what might fit with this subject.

While seemingly ‘messy’, we learned the relevancy of the map’s visual context. One colleague tried to write up a summary list to report our findings and realized it would lose relevant meaning.

Reinforced by Wednesday’s presentations (and many great interviews esp. Euan Semple), it was evident that the need to find individuals is as important as finding stuff (the latter being the typical focus of search engines). This is not possible without access to an individual’s “Identity”. We noted that most search results focus on “you tell me” (a one-way broadcast). Telling is not the same as “advising” (allowing for choices), or “heralding” (featuring specific or related content).

A more conversational model would let “me tell you” as well (touched on in Bjørn Olstad’s interview). This goes beyond inference because individuals operate in different roles. Amazon tries to infer “interests” from my behavior, but there are exceptions such as when I’m buying for someone else, which we noted as “surrogacy”.

The individual identity becomes a group identity through “affinity” (the power and potential for collectives to form is the focus of Clay Shirky’s interview).

The ability for collectives to form and dissipate in a continuous adaptive way is critical to any living organism or system that supports living organisms. Adaptation is hindered by ‘fixed’ systems and solutions, which are often linear and “non-dynamic” (the html page paradigm of Web 1.0 is non-dynamic). Related non-organic barriers and gotchas we labeled “the swamp”, which is closely related to a need for “breaking the paradigm” (stepping away from what we traditionally assign to search as a function).

Even though we intentionally focused on “breaking the paradigm”, as the next group presented, we immediately saw some of the ways in which we had been ’stuck’ looking at the problem from one perspective, because we’d not included the significance of “social” in our model (the interview with Charlene Li touches on some of the relevant aspects of this potential).

The existing search paradigm assumes that individuals are “coming to your party”, rather than search meeting an individual where they already are in their normal activities and suggest relevant recommendations. We see such seamless recommendations starting to appear in examples such as the new gmail feature where emails with dates/time elicits an inquiry “Would you like to…add to calendar?”

But a recommendation is not a conversation. Without a  conversation an individual cannot suggest what does and doesn’t work for them — they “can’t say no” — to fine-tune what gets presented, to help suggest their specific “intent”. Having such conversations requires a persistent memory to facilitate an intelligent exchange. Without such, we struggle with repeated “wasted investments”, going down a particular path with a series of “relational operators” only to have to “start over” with another search. One search fails to build on the last and doesn’t facilitate a “series of intents”.

Often a search experience “assumes a perfect” scenario — the results are perfect, the individual is imperfect. It is the individual who is unable to tell the engine what it needs to know — a futile effort for something without a memory, thus no ability to ‘know’ anything. And yet, search engines boast about the “results #” returned or the speed in “seconds #” — celebrating the ability to return more wrong results faster.

We covered pretty deep territory in a very short period of time which makes you wonder why many search experiences are still so painful. Doing a better job of covering all of the points we mapped out (including those of the other groups) can be facilitated by technologies, especially search. Because the people behind the scenes at FAST think deeply about these challenges, their tools are architected to support better Search-Driven Experiences (Mark Stone’s interview highlights related examples), and the FAST team is clearly open and willing to discover more.

A big thanks to FAST for allowing us the time to engage in this interactive exchange of mutual discovery.