Site   Web

August 19, 2009

Searching Social For Search

I recently wrote about how, in future, search could greatly benefit on-demand digital television, but the future of search doesn’t start there and isn’t even that futuristic. I think search is about to undergo a major evolutionary shift that will change the underpinnings of how search works, is used, and is defined.

Advertisers want to fish where the fish are, and the fish are very social these days. Social media sites are coming of age and Twitter search is all the rage. But if social is the new flavor of search, we need to break it down and understand the how and the why.

The first factor in this comes from a cautionary tail of jumping the shark. MySpace did not protect the sanctity of community, which is the single biggest value driver for social. As the community grew, everyone befriended everyone and thus the community lost its value. To avoid a similar fate, Facebook and Twitter must protect their community like the engines guard their algorithms to maintain relevance. At some point, these companies need to start taking some liberties to protect the community for us to ensure long-term value. Twitter, for example, needs to do some serious housekeeping, as there are too many accounts a tad on the shady side, to say the least. If Facebook and Twitter can avoid diluting the community and jumping the shark, they can change the game of search.

To understand how, lets look at why we search. I’m oversimplifying here, but there are three basic need states that dictate our search behavior that I will use to prove out my theory.

  1. Discovery — we search to find new things of interest.
  2. Information — we search to find specific information based on what we discover we like or want.
  3. Navigation — we search to get from point A to point B, because it’s the simplest way around the Web.

If I want to go on vacation, I’ll type in “vacation.” Once I learn a little, then I start searching for “cheap holidays to Spain” or “Bahamas cruises.” After I have compared prices, used multiple engines over multiple sessions, and I have decided what I want, I type in the company or brand that has it and I convert.

Why is this important? Because if community adds value, then some of the aforementioned search activity will start to shift to social areas. I would rather ask my close friends, family, coworkers, or anyone with a shared interest than hunt and peck my way through algorithmic search results. Not only does this have the potential to speed up a consumer’s search journey, but it also comes in the form of trusted sources in real time. You can never underestimate the power combination of immediacy and word of mouth. But given the nature of what social search can be, it will never add a lot of value navigationally — so I see that search activity staying at an engine.

This shift in behavior (if it plays out) has huge implications for our industry. When we do multiclick attribution analysis for our clients, we see about 50% of the value of a last click and conversion being re-attributed to early funnel keywords in the discovery and informational search phases, where the battle for awareness and interest is really occurring. This is because the vast majority of last clicks and one-click search journeys come from navigational searches. More often than not, this is your brand terms. Google right now owns navigational search because it owns the browser-based search box, so using last-click attribution, it gets all the glory. This was part of Yahoo’s search undoing – – our research indicates that Yahoo is typically used in early search phases with nonbranded keyword queries that don’t result in a direct click to conversion. When we changed our attribution weighting we saw that Yahoo received roughly 4% more credit for revenue generated.

If early funnel keyword activity increasingly happens within Twitter and/or Facebook versus at, dare I say, a traditional search engine, then the game has changed in two major ways. One, this shift in behavior represents a sizable amount of monetizable query volume for the social communities. Two, if we attribute roughly 50% of the value of the last click back to earlier exposures, then the price I am willing to pay for that click drops proportionally. Think about it — search under last-click attribution still runs on the 80/20 rule, meaning 80% of revenue comes from 20% of the keywords. So Bing and Yahoo risk a loss of volume and Google risks a drop in click value.

This bodes well for social communities trying to find revenue streams, because early funnel nonbranded keywords, which is where social search can add the most value, come with higher CPCs and scale. So maybe the Google killer is not one engine, but a collective change in behavior where a growing volume of search activity starts happening in social environments. There is a lot of revenue at stake here, especially as advertisers get savvy and move away from last-click attribution.
A change in tracking attribution combined with a more mature social experience is changing search. The questions become: Who buys Twitter, and can Facebook do this on its own?

Rob Griffin is Director of Search & Analytics at Media Contacts, the digital (that’s the frosted) side of Havas Media. Rob can always be reached on his Crackberry at