Google Search Plus Your World: The Peer to Peer Evolution Continues
This article by Howie Jacobson was originally posted on Search Engine Watch.
Google now sprinkles Google+ results into general search results, in an initiative they’ve dubbed “Google Plus Your World.” Webmasters and SEO experts have been arguing heatedly about why Google is doing this (most point to potential revenue enhancement and the possibility of world domination through the marginalization of Facebook) and what this will mean for the SEO game (treat Google+ postings and pluses as links) and how this will influence PPC (fewer bits of prime organic real estate leading to increased AdWords bidding wars for the top positions – see “potential revenue enhancement” above).
This article examines Google Search Plus Your World from a different point of view. What is driving the integration of social and search? How does it work? And what’s the impact on our perception of Google’s relevance and usefulness?
The Supremacy of Peer to Peer Search
The history of search has been the continual de-centralization of the search algorithm. The founders of Yahoo, Jerry Yang and Dave Filo, started it by creating “Jerry and David’s Guide to the World Wide Web,” a directory of cool and useful sites. The sites that made it onto Yahoo! were chosen by Jerry and Dave, and initially users had no say in the selection or order. In other words, totally centralized control of search.
Google shifted search results prioritization through a peer-based system of voting by links. Sergei and Larry didn’t go through the web and choose the best sites for every search, nor did they hire an army of web reviewers to do it for them. Instead, they instituted the PageRank metric that essentially turns each search result into a popularity contest, with the most popular web pages having the most votes via their outbound links.
But the algorithm was still supreme, and operated independently of any insight into the searcher’s unique needs, situation, or personality. The next phase of search wasn’t via a search engine at all. Instead, it came about via Facebook and Twitter, two places where you could ask a question of your entire social network and get almost instant relevant answers from people you know and who know you.
True, the answers didn’t come back as quickly as Google’s (1,300,000 results in 0.39 seconds). But they were often much higher quality. One reason for the increased quality was trust – we expect our friends to look out for our best interests. But the other was quite simple: instead of entering “search queries” of just a few words, you could embed context and conditions into your search.
Search Queries vs. Real Questions
For example: a search for “raw food cookbook” on Google yields several amazon and Barnes & Noble listings, some product photos of book covers, and a couple of websites dedicated to raw food diets. But what we really wanted to find was, “a raw food cookbook that doesn’t require tons of new equipment like spiral vegetable slicers and dehydrators and juicers, and that uses whole foods and doesn’t rely on tons of oil and salt for flavor. And which works if I don’t eat gluten or soy.”
When I type that search into Google, I find it’s too long to be fully considered. And it gives me pages that include my search terms, even if the context of my query negates those terms (like “vegetable slicers” and the other gadgets I don’t want to buy).
But that same question posted on my Facebook wall or my Twitter feed can generate helpful responses from dozens of people who all get the full context of my question in a way that a computer (even Google’s) can’t.
And since my friends can see each other’s responses, they can iterate their answers and get even closer to my search intent. All in all, social search can be a much more efficient process than Google.
The problem with Facebook search is that it’s not archived for on-demand review. If I want to find a link or a quote that I posted last month, I have to manually scroll through my timeline. I can’t search for “Sacred Economics” or “crying bullfighter” or “Rowan Atkinson standup” to find my own posts, let alone those of my friends.
Google, on the other hand, has an answer ready for me the instant I think of a question. By incorporating my friends’ and acquaintances’ opinions and expertise into the full search experience, GPYW takes advantage of peer to peer content generation while incorporating it into the more expansive algorithmic results.
That mashup not only improves the search experience by adding personal results. It also increases searchers’ perception of the relevance and trustworthiness of the organic listings and ads by association.
Google as Playlist
In The Power of Habit, Charles Duhigg tells the story of a hit song that almost wasn’t. The song, “Hey Ya!,” by the hip-hop group OutKast, showed every sign of being a megahit when it was released in the summer of 2003. Record executives and DJs loved it. Fed into “hit predictor” software that compared a new song’s tempo, pitch, melody, chord progression and other factors with known hits, “Hey Ya!” scored off the charts.
Yet when radio stations started playing “Hey Ya!,” listeners hated it. According to Arbitron, 30 percent of listeners switched stations within half a minute of “Hey Ya!” Fans told the DJs that “Hey Ya!” was one of the worst songs they’d ever heard.
What had gone wrong?
It turned out that “Hey Ya!” was too different from the typical Top 40 hits of 2003. Even though it sounded “great” to music insiders, its combination of hip-hop, funk, rock, and Big Band was simply too strange to be “likeable” on first listen. According to research shared by Duhigg, our brains prefer familiarity to quality.
Even if a new Celine Dion song sounds like every other Celine Dion song I’ve ever heard (and I haven’t liked any of them), I keep listening to it on the radio because it’s what I expect to hear.
Our brains have evolved to distinguish familiar from unfamiliar sounds, and focus preferentially on the former. The habit of attention saves us from having to make decisions, and thus frees up our brain for other tasks.
Arista record executives and DJs eventually turned “Hey Ya!” into a giant hit by sandwiching it between two top hits. Not just any hits, but the stickiest kind of songs that sounded familiar the first time you heard them. Radio playlists are an exercise in risk mitigation. If you play only familiar songs, listeners get bored. If you play only new songs, listeners get annoyed. The function of the playlist is to insert new songs that already seem familiar.
One way to achieve this familiarity is to only serve up tunes that sound like other tunes. The other way, used with “Hey Ya!,” is to trick the brain into familiarity by connecting the unfamiliar with the familiar.
GPYW: Playlist Risk Mitigation
The Google search results page is nothing more or less than a playlist: the “Top Hits” for each search query. One of the ways you can tell Google is working is that it shows you new stuff. Stuff you won’t find otherwise. But total novelty, as we’ve seen with “Hey Ya!,” can be off-putting.
If I search for “how to train for a marathon,” for example, I expect to see references to running experts I already know and trust: Phil Maffetone, Danny Dreyer, etc. When I don’t find them, I get worried about the entire results page. Can the resources listed by Google be trusted? Will I overtrain? Get injured? Experience nipple chafing, or worse?
The addition of personal results to the search results page reassures me about the quality of the entire page. Seeing blog posts and shares by people I already know and trust puts the new, unfamiliar listings (especially the things Google really wants me to click, the paid listings) into a comfortable context. Responding to the Facebook challenge, GPYW has stepped up the integration and customization of the Web to blur the boundaries between the known and the new.