Bursting Your Bubble

The Basic Argument

Recently a debate has been sparked about the impact of personalized filtering algorithms on the search results different individuals will get from seemingly universal search engines such as Google, Bing etc. The argument made is that these personalized algorithms learn from our online search behaviour and will increasingly modify results to better cater to our preferences, significantly changing overall search results per individual. If all search engines do this, the argument goes, then in the end, all of us are each trapped in a particular information bubble determined by what the filter algorithms perceive as our preferences.

Squirls Or Africans?

Some of the argument is summed up in these two quotes from the main “culprits”:

“A squirrel dying in front of your house may be more relevant to your interests right now than people dying in Africa.” – Mark Zuckerberg CEO of Facebook

“It will be very hard for people to watch or consume something that has not in some sense been tailored for them.” – Eric Schmidt Google Executive

Video Of The Whole Argument

Watch Eli Pariser’s TED Talk below for the whole argument:

For more information on the argument, visit: http://dontbubble.us/ and the (filterless) search engine it turns out it promotes: http://duckduckgo.com/ or the TED Talk including responses by the public: http://www.thefilterbubble.com/ted-talk.

Or Is It The Whole Argument?

Taking into account the effects of social

In fact: you don’t need to have a look at search results your neighbour gets to see how you will each have different results. You could simply move search results up manually in Google to indicate you prefer them for instance and ofcourse you wouldn’t expect the same preference to show up on your neighbours screen.

The argument could be significantly strenghtened if instead of just looking at personalized algorithms, we take into account the impact of the social web. Google has been striving to incorporate social into their overall product mix and strategy and have recently added functionality that will move results up if your friends like them. In addition they have introduced the +1 Button, which similar to Facebook’s Like Button can help you indicate your preferences and follow those of your friends.

Facebook on the other hand has it’s own algorithm to manage messages by your friends to rank them in (or out off) your Facebook page’s feed. See for instance: Facebook Edgerank. What’s interesting is that Facebook explicitly states that it wants to introduce this friends based preference system as a search system for the whole web. Where as Google is responding to this development from a position of optimizing search relevance to one which incorporates friends based preferences. And Bing follows.

Taking into account the effects of Local

At this moment the web is shifting in other ways as well: from a universal world community to a location based and mobile device based series of local communities. This means that those of your preferences that succesful web companies cater to, will both in information and in advertising also be increasingly local. They will focus more and more on you as an individual moving around your local community. You will receive more ads from local stores (think of Foursquare and Groupon) and less information about global or even national events. You will receive more information and ads on local events and offerings that match your preferences and less of these that are more distant from- or less pleasant to you.

Better Algorithms

The problem we’re dealing with here could also be phrased as a technological one: what is potentially hindering us is that the algorithms are imperfect. Their imperfection misunderstands our complexities as multifacetted individuals and potentially steer us towards limitations we don’t want. In order for these algorithms to be more perfect we would have to give up even more privacy. But that’s another problem altogether.

A Larger Perspective

Particular versus Universal, Wants versus Needs

The whole thing could be illustrated by a philosophical debate between Newton and Leibniz in the 17th century: Newton claimed the world was absolute and universally understandeable (Google) where as Leibniz believed that one’s position in the world and one’s understanding of it was relative to the individual (Facebook).

So: is the web and all the knowledge in it absolute and the same to everyone? Or do we shape our own information worlds (pre-)determined by who we are: seemingly choosing to do so but in fact locked in by our own limitations as expressed in our preferences (or those of our friends)? It’s very tempting to choose ourselves and our preferences as the basis of our perception of and access to information and yet there is a difference between on the one side: our particular point of view and preferences and on the other: our needs and view on other’s perspectives.

If we choose an absolute space of information, equal for all then we must realize that such a space also may be determined by other things. For instance: search results in Google were determined not only by relevance or some “academic standard of excellence” before personalization and “social” came into the equation, but also by money and strategy i.e.: the site with the best internet marketing strategy and budget would prevail over the site with the lesser budget and strategy in the search ranking. I missed this point in the argument made by mister Pariser. On the other hand SOCIAL marketing strategies will also favour volume that can be bought. Coca Cola is bound to have more friends on Facebook then your neighbourhood greengrocer and will have more impact on your social search results.

Golden Age Of Marketing

It’s been said often that we live in a golden age of marketing. Ten years ago the phrase was: we live in a golden age of information. Now you might think: Are you not an internet-marketeer? Yes! So I’m not complaining and perhaps as a marketeer I’m mad to even write this post or hint at the possible short comings of any marketing instrument or -technology. At least Eli Pariser is helping sell an alternative search engine. What am I selling here? Click on some of the adds on this page to find out (haha).

But seriously: marketing is a mature multi-facetted discipline. It’s not all about bombarding people with crap. On the level of long term marketing strategy for instance it could make perfect sense to weigh the pro’s and con’s of particular trends in relation to societal tolerances.

In Conclusion

As Adam Smith famously argued, and I’m paraphrasing here: “Specialists are great for production and commerce; education and a wider range of interests are great for society.” Many of us do prefer what we want to what we need and what’s close to home to what’s happening elsewhere in the world. The web will cater perfectly to those of us. So what the future of social search boils down to is: yes you will be reduced to an information bubble! Especially so if you are a specialist with clear preferences for search topics who is an active social media user with a very particular set of friends who share your preferences. Perhaps not tomorrow but certainly in the longer run. And this effect will be further enhanced by the advertising around all your web activities. Even more so when the effects of personalization and social are enhanced by the maturing of local. But none of us are this one dimensional are we?

Most succesfull web companies need to cater to your preferences in ever better ways not only with information but also, and for the sake of their survival, primarily with commercial offerings. These commercial offerings need you to be interested in what you’re reading or viewing in order for them to be relevant to you. This is necessary for them so that they can be competitive. At the same time it guarantees you get the quality of information that you WANT, and will gradually make you a more predictable consumer of information, and therefor a more predictable consumer of ads. A good investment from a marketing perspective: a sure thing. Visiting the internet will become increasingly like looking into a mirror and liking what you see.

If on the other hand you are generalist with a wide range of friends, you will be less predictable; algorithms won’t be able to cater to your preferences as readily because they are less easily mapped. It might be harder to find what you’re looking for as fast. On the other hand you will have a wider view of what’s going on in the world. You will get less of what you want and perhaps more of what you need. You will be able to take a step back and look at what you’re doing once in a while.