After reading a number of complaints about the quality of results and how often Google+ pages were featured in the Top 3 positions in the new personalized search, I reached out to the SEW authors community and had a bit of fun testing out Search Plus Your World. In our admittedly small and completely unscientific test, we found that personalized search seems to be missing a few key ingredients we’ve come to expect from Google: relevance, content quality, freshness, and context. This is largely due to the fact that Google+ results outrank those with more of the qualities we expect in organic search.
It seems possible to rank top 3 for whatever you want in the personalized search results of people who have Circled you on Google+. In its current state and with the volume of content on Google+, Search Plus Your World reminds me more of Alta Vista’s Scooter, scouring their own network for broad keyword matches, than an evolution of Google search.
This could be a great thing for marketers, until it changes, as it inevitably will. Right now, though, early Google+ brand page adopters and those who survived the Google+ business account shutdown can rank for pretty much anything they choose in the logged in searches of the people who have Circled them.
Quality of Results for Personalized Search
First, I searched myself. Logged in as myself, I would expect a search for “Miranda Miller” to show results pertaining to… well, myself. Surprisingly, 3 of the top 9 results were for other Miranda Millers – granted, there are quite a few of us. But if Google can’t figure out which ones are actually me, how can anyone else? Of the 6 relevant results, 2 were Google+ pages or posts, with the #1 result for Miranda Miller being my Google+ profile page. This listing dominated the space above the fold, with my last 3 activities listed in a large box at the top.
Searching for “flights to San Jose,” the 5th organic spot goes to the only person in my network, apparently, who has mentioned “flight” and “San Jose” in a post. His post is from September 2011. He’s not talking about flights to San Jose, but that his building is in the flight path in San Jose. Relevance? Context? Freshness? All absent. Broad keyword match is all it takes.
Ranking in Organic Search Within Your Network for Whatever You Want
Seeing this, we tried to rank for terms that have absolutely nothing to do with anything we’re involved in. AOL Search Engine Optimization Director Simon Heseltine chipped to see if he could rank for my terms, as a person in one of my Circles.
The results of this (really small, completely unscientific) experiment were surprising because they showed again that currently, there seems to be absolutely no relevancy or context required to rank in the top 3 results in a personalized search. Heseltine posted a repetitive string of “hire a rocket scientist” in a public G+ post, and within two minutes, ranked #2 in organic personalized search results for the term to myself and others who had him in their Circles.
Now, you can’t Circle people yourself and appear in their personalized search results – we tried. They have to choose to be connected to you. Once they are though, you can rank for whatever you want in their searches, provided they are logged in to personalized search. We saw the same results with “selling puppies” and the more competitive “flights to Jamaica.”
This is big news for brands, who can rank at the top for anything they choose right now to users who follow them and are signed in while searching (AKA those 5 to 25 percent of users you’re not getting keyword data on anymore!). This might depend on the topic and the volume and frequency of posts on it, but that it works this way at all runs counter to all of the improvements Google has made in regular search in the past several years. You don’t have to be a topic expert. The freshness and quality of the result doesn’t matter. The context of the query doesn’t even appear to make a difference.
It’s hard to imagine this improving personalized search results if or when, as Google plans, Google+ is widely adopted and the volume of content increases, with more to pull from their social network on a greater variety of topics. Even in early days, it seems to add a lot of unnecessary noise to organic search.
In 2009, when Google integrated Twitter into regular search, Rae Hoffman found that she could tweet her way to the top of the SERPs with Viagra and Miley Cyrus spam. One obvious difference here is the scale; Google+ users can only reach people within their network with personalized search, not the general public. However, the very idea of personalized search, the reasoning behind so many of the changes over the last year and the steps Google has taken to this, were all to improve search for users. If what we’re seeing isn’t fresher, higher quality, more relevant, or exactly matched to what I was looking for, where is the improvement? And what is it doing at the very top of the results?
Matt Cutts took to his personal blog to “share a search story” in an effort to show that Google’s personalized search isn’t favoring their own social results. He writes, “I hope that helps to make my point. Search plus Your World builds on the social search that we launched in 2009, and can surface public content from sites across from the web, such as Quora, FriendFeed, LiveJournal, Twitter, and WordPress.”
His argument misses the mark, though. Whether or not other results are there, they certainly don’t appear to be weighted evenly. If the most relevant, trusted results rank tops, that needs to apply to those coming from within Google’s own properties as well as everyone else’s if it is to offer any benefit to users.
I imagine we will see personalized search being tweaked and improved upon on an ongoing basis. However, this initial version really looks and feels like a completely stripped down version of search we might have expected 10 years ago. Full social search has to start somewhere and this initial offering leaves the kinds of holes Google spent years stopping up in regular search. Do with this knowledge what you will.