If there’s one thing that’s made me borderline OCD (happy now Michael?) would be the world of personalized search. Probably more so, behavioural data, is an area (both explicit and implicit) that seems to be part of any likely evolution of search engines (and by extension SEO).

Now, I am not alone in this. The folks at Google have also been more than a little motivated as well. Or as Marissa Mayer said a while back, “(…) what we really want to do is search it as each individual user sees the web”. I mean, think of that for a moment. A granular level that is truly personalized. If what moves up in the SERPs is the pages and related types of pages, that you like. Through interactions, direct and indirect, results are re-ranked. Would spammers click and favourite their own sites all day for the lonely glory of only seeing it rank on their own particular computer? Fun, but not profitable.

Are we there yet? Not really. From what I understand and believe, it is still more of a user categorization than a truly granular implementation. But it is coming IMO and it’s worth looking at.

Personalization data

While there is no end to what can be tracked by a search engine, this particular Google patent (awarded last week) offers up the following;

Scalable user clustering based on set similarity (also interesting is; Personalized search methods)

Types of data that may be tracked -

• demographic data
• psychographic data
• personal interest data
• personal activity data

And what exactly are those? It is interesting to break them down. It gives you a better idea the types of things that Google (or any search engine really) might be tracking to perform personalizations

Demographic data includes;

• age
• gender
• geographic location,country of origin,
• country, city, state, zip code,
• income level
• height, weight, race, creed,
• religion
• sexual orientation
• political orientation
• education level
• criminal history
• health.

Psychographic data is any data about;

• attitudes
• values
• lifestyles
• opinions derived from demographic data

Personal interest data includes items such as interests,

• hobbies
• sports
• profession or employment
• areas of skill
• areas of expert opinion
• areas of deficiency
• political orientation
• habits.

Personal activity data includes data about;

• past actions of the user
• such as reading habits
• viewing habits
• searching habits
• previous articles displayed or selected
• previous search requests entered
• previous or current site visits
• previous key terms utilized within previous search requests,
• time or date of any previous activity.

This gives us some insight into how the search engine might try to better understand the user in hopes of better categorizing them. I say that because in most implementations so far they do, as stated previously, look to categorize searchers for personalization. If you fit into a given profile, then grouping those interactions can make for better results and re-ranking mechanisms.

By extension we can see the implications of the social graph to this process as well. That though, is another discussion altogether.

Is there a need for SEOs to be concerned?

This of course is the core question one needs to ask at this point and the answer is two fold. Yes… and No. HA! Take that!. Ok, but seriously, the last round of personalized search research we did here on it seems to suggest that a lot of the personalization, in relatively new query spaces, are not that personalized. Before the user has actually been identified in an area, there is less of a chance of the standard results being heavily modified (that may have changed as Psearch for all has propagated). To that end, I wouldn’t be overly concerned (think of more generalized ecommerce areas as opposed to a specialist informational site).

On the other hand, we can also make the leap that personalization will only get deeper and having strong grasp of your demographic and encouraging engagement will be paramount. I would suspect that this is more important for some query spaces over others as inferred above.

At the end of the day, there will ultimately be a premium on truly making the most from a user experience perspective.

A potential framework of SEO for Psearch

The next thing we need to consider is what adaptations would one make to best position themselves to get the most from personalized search implications? This is a tough one and we don’t want to put the cart before the horse while maintaining our need to future proof our SEO.

• Demographics – know your market
• Relevance Profile – build strong themes
• Keyword Targeting/Phrase Strategies – tighten up
• Quality Content – DUH
• Search Result Conversion – work the meta data
• Freshness – if your market demo requires it
• Print Page/bookmark me/social – seriously, can’t hurt.
• Site Usability – keep it user friendly.
• Analytics – start to measure users more (SEO not just conversions)

To quote myself, not because I am lazy, but it seems to capture it, (ok, maybe just a little lazy);

“If you are already knee deep in Content Strategy, incorporating Social and otherwise maxing out engagement levels, I can’t see that you have much to be concerned about. Using implicit data gives greater insight into user satisfaction (on paper at least) and those already moving in this direction should be fine. If your still pushing borderline MFA sites and link spamming for love; it may be time to reconsider the approach. ”

There is every chance that your SEO strategies may not need a complete over-haul.. just some conscious efforts to expand what you’re already doing.

What else should you be doing?

Work on engagement strategies – it seems that we would also want to not only work on getting the right users to the site, but also engaging them once they’re there. This can include implicit actions (scrolling, page depth, click data, dwell time etc..) or explicit (adding to favs, emailing, social voting/sharing etc…).

Universal elements – another interesting aspect of personalization is in the universal SERPs. As some secondary anecdotal evidence has suggested, there is also an element of personalization taking place in the when and where of universal elements in the SERPs, (local, shopping, video, news etc..). This also does seem to indicate the being more intimate with your query space and demographic will be important in a granular Psearch world. Know what they like and give it to them.

Don’t forget Android – we might also want to start paying a little more attention to elements outside of the standard web search. In particular the world of Google Android which is powering not only the mobile world, but Google TV as well. This means a larger behavioural data set and new ways to engage users.

Qualitative Research in SEO – given the premium on better understanding your target market, it may also be advantageous to do some qualitative research into the target group(s) that you may be after. Find ways to better understand your demographic and mould your content programs accordingly. (learn more about that here).

Don’t discount PPC efficacy – another twist we might throw into the mix is that users actions with PPC adverts could also be a factor. To that end one would want to ensure good engagement with any ads being run as it may just affect which user types are shown

the Caffeine Connection

And one last thing before we go; deeper personalization would require more power. Or at least that has been my belief for many years. With the prospect of getting more power, more specifically through the recent Caffeine infrastructure update, we may be one step closer any of this really being of importance to us. One can’t help but look at this, coupled with the interest in the social graph, to see that this is now, more than ever, an area SEOs should be seriously considering.

How deep and how much personalization is having will be something that requires more research. For now it may be prudent to start programming with the concepts in mind where they over-lap good webmastering, marketing and optimization techniques already in play. Future proofing your SEO is part of the art. Learn it well.

/end time travel experiment

Resources;

I would be remiss if I didn’t leave the adventurous with a whack of extended studies;

On the Trail;

• Personalized PageRank: a user sensitive affair
• Why Googles behavioural targeting should creep you out
• Is Google REALLY using bounce rates as a ranking signal?
• Your favortite sites could rock the SERPs
• Yahoo, Page Rank and Teleportation Oh My!
• What every SEO should know about Personalized Search
• I have seen the future and it is VERY personal
• Why Google Personalized Search is Important to You
• Personalization through application focus
• Query analysis confirmed as Google ranking signal
• Taking off my tin foil hat

Marie Claire-Jenkins recently posted; 10 research papers on Personalized Search

From Google

• Personally Speaking 
• Personalized Search and Privacy
• Google Puts users in charge
• Talk on your personalized Home Page
• More of the world in your pocket

Patents

Google Patents –

• Collaborative Search Results
• Methods and systems for personalized network searching
• Detecting anomalies
• Domain-Specific Sentiment Classification
• Rank-adjusted content items
• Annotating webpage content (Site Wiki?)
• Systems and methods for analyzing a user’s web history
• Systems and methods for modifying search results based on a user’s history
• Systems and methods for providing subscription-based personalization
•  Systems and methods for managing multiple user accounts
• Methods and systems for personalized network searching
• Personalization of placed content ordering in search results
• Systems and methods for providing a graphical display of search activity
• Systems and methods for providing subscription-based personalization
• Systems and methods for analyzing a user’s web history
• Systems and methods for combining sets of favourites
• Systems and methods for modifying search results based on a user’s History
• Variable personalization of search results in a search engine

Microsoft Patents –

• Presenting result items based upon user behaviour
• Ranking and Providing Search Results Based In Part On A Number Of Click-Through Features
• Dynamic client interaction for search
• System and method for ranking search results based on tracked user preferences
• Specifying relevance ranking preferences utilizing search scopes
• Constricting web query hierarchies from click-through data
• Using related users data to enhance web search
• Adaptive dissemination of personalized and contextually relevant information
• Implicit links search enhancement system and method for search engines using implicit links generated by mining user access patterns
• Intent aware search
• Query classification based on query click logs
• Search system using user behavior data
• Identifying relevant information souces from user activity
• Dynamic search with implicit user intention mining
• System and method for personalized search
• Method and system for adapting search results to personal information needs

Yahoo Patents -

• Shared influence search
• Customization of search results
• User entertainment and engagement enhancements to search system
• System and Method for Creating and Applying Predictive User Click Models to Predict a Target Page Associated with a Search Query
• Method and systems for using community bookmark data to supplement internet search results
• Personalization of search
• Collaborative-filtering contextual model based on explicit and implicit ratings for recommending items
• Systems and methods for collecting user annotations
• Trail Identification in Search Results
• Monitoring user interactions with web pages

• Tweet
• Buffer

• Bio
• Latest Posts

David Harry

at SEO Training Dojo

David Harry is an SEO and IR geek that runs Reliable SEO, blogs on the Fire Horse Trail and is the head geek at the SEO Training Dojo.

Latest posts by David Harry (see all)

• Google Social Search; the Lost Update - July 5, 2011
• Weekly Search & Social News: 05/17/2011 - May 17, 2011
• Weekly Search & Social News: 05/10/2011 - May 10, 2011