Will The Personalised Web Destroy Discovery?
It’s been a big week for the evolution of the personalised web, led by Mark Zuckerberg and the unveiling of Graph Search. Much as the product has been lauded – it filling a gap in social search that we didn’t even really know existed – it has also been met with some criticism, not least around the fact that it limits the breadth of information you can access. The more tailored you make your search, the more limited the search results will be, with the default option of showing you Bing results really being no different to the current search experience in Facebook.
The biggest question that Graph Search raises – and this applies to the personalised web overall – is whether personalisation will ultimately destroy discovery. The premise behind the personalised web is clear: When our time online is at a premium, coupled with the amount of data about ourselves we are freely sharing, there is a clear opportunity for web services and social platforms to automatically personalise content and show only what’s most relevant to us. Where does this leave spontaneity, and broadening your information sources outside your immediate interests and activity?
Customisation Versus Personlisation
It’s worth clarifying what we mean when we say the personal web. This isn’t about building a bespoke online experience through adding contacts to a social profile, following people you like or subscribing to feeds of your favourite topics. That’s customisation. Building an experience within Facebook, Twitter or an RSS reader that gives you bespoke content that, importantly, you have selected.
What we really mean when we say personalisation is when technology does this part of the thinking for you. By deploying algorithms and tracking you through cookies, it serves you content or even individuals that it ‘knows’ you will like, based on your browsing activity.
The personalised web is by no means a new concept, but with the current data economy and the flow of information being traded between users and digital platforms, personalisation is due to reach exponential levels. The concept of the internet is gone. It is now your internet.
The impact of this can be seen in Google’s new product Now. Running on Android, this essentially lets your mobile do the thinking for you, based on the amount of information you have about your life, through your handset.
It works on a system of ‘cards’ like travel updates as soon as you leave the house, or automatically displaying your boarding card when you enter the airport. This is personalisation at its most efficient and groundbreaking.
But is this taking the concept of personalisation too far? Does it ultimately make us too reliable on the technologies around us, removing spontaneity and just thinking for ourselves? What if developments in technology are ultimately detrimental to our personal efficiency?
This is not to suggest that technology is actually damaging our brains. You would never find me arguing that anyway (and what people who argue that should really say is the ‘application’ of technology as opposed to the technology itself), but it would be hundreds of years before any biological change like that could occur, but the immediate, environmental effects are very real. We are creatures of habit, after all.
The Clue Is In The Title
Lest we forget, it’s the World Wide Web. That is, an infrastructure that connects information to present it to the end user within a range of growing media. Any system that advocates closing that web to some degree (i.e. only showing you the information that has been deemed most relevant to you) fundamentally changes what the World Wide Web can offer. It is neither ‘worldly’ nor ‘wide’ but is a web that pertains to you.
Uniquely individual and unlike the web that anyone else will see. Yet the advent of the personalised web is an unstoppable juggernaut, with organisations increasingly favouring personalised web experiences. Indeed, 52% of marketers agree that ‘the ability to personalise content is fundamental to our online strategy.’
The motivation behind this is clear. That social media has allowed anyone to become a publisher, where we are all journalists in a loose definition of the word, creating and publishing content daily, we now have more of that content to sift through. So yes, technologies that can aide this discovery are welcome, but there is a threat within this that ultimately closes off access to all the information we have a right to see.
If we rely too much on personalisation, seeing what an algorithm or publisher has decided we should see, we risk becoming digitally isolated. When our communities and information sources are so carefully constructed, it thereby follows that you would only become exposed to a certain set of information, within your interests, industry of work or personal behaviour. You can see this digital isolation in action within YouTube’s recommended videos. This is what I see on the homepage of YouTube:
Granted, YouTube is not selling itself as a music discovery tool, but I could probably have guessed for myself that if I liked a video by one artist or user, I probably would like their other videos. It doesn’t really help me at all.
What I would actually like to see are videos that I might not have encountered or that someone has actually selected as being interesting. Compare this to a screenshot of Spotify. My favourite way for discovering new music is through the ‘We Are Hunted’ app, which simply gives you a playlist of the top 100 emerging artists. In no way tailored to me based on my listening activity, but providing a gateway into new music:
And this works because it functions around curation of what’s being said online in general, as opposed to just looking at my personal listening history. We Are Hunted, which also runs as its own site, scans social media for what’s being shared and discussed about emerging talent. They use digital methods such as sentiment analysis within blogs or advocacy to determine what’s actually popular, presented in a playlist that anyone can access. I don’t actually care that the playlist I’m seeing is the same as what anyone else will see.
A service like We Are Hunted works because it is crowd-sourcing the trendsetters and informing you of what’s new or popular right now. And sometimes this is what you want. Discovering personalised playlists based on similar genres to what you’ve previously listened to isn’t necessarily the right way.
There is a temptation for services to take the wealth of information available about you, track your entire online journey and present what they think is right. What about what everyone else is saying? What about when I want to broaden my horizons? Maybe I don’t want to see ads for Italian restaurants following me around online just because I looked at a review of one somewhere. Maybe I want to see something out of the blue because someone, an actual person in the know has determined that it’s good.
There is a need for brands and organisations to remain authoritative bringing you the trends based on their knowledge. The difference now of course is that their sources for discovering those trends themselves have grown. We can crowdsource our trends instead of lying on a single source, and we can do this through public publishing platforms such as blogs or forums.
The Role Of The Curator
There is an emerging role within online media to bring the ‘personalisation’ behind the personalised web. While I may not want to be bombarded with ads for Italian restaurants because I looked at one online, I would actually be very interested in then seeing an ad for a Japanese restaurant on the other side of town, that none of my friends have ever Liked, checked into or reviewed on Tripadvisor, that offers a type of cuisine I’ve never even heard of, but that just happens to be reviewed by a little-known but knowledgeable food blogger whose site I’ve never visited.
Even with the most sophisticated algorithm, there’s little chance I would see an ad for this, other than through successfully matching the keyword ‘restaurant’ or context-matching between thousands of restaurants in London. More importantly, I wouldn’t see it as a result of a review on a food blog that someone has determined to be particularly influential for that cuisine.The point is that if the personalised web continues the way it’s going, the chances of discovering little known gems like this are actually decreasing as the algorithm reigns supreme. What we need are those influentials, whether traditional journalists for large publications or home-grown bloggers with no money but a lot of passion, to inform the algorithms that determine what we see, otherwise we really do risk digital isolation.
And as publishers move towards personalised content, such as the Mirror’s plans to introduce bespoke homepages. Using technology from Rummble Labs, they will deliver ‘targeted’ content based on data gathered from social media. While it’s encouraging to see something like the Mirror, who has been slower to adapt to digital than other papers in the UK, there is a danger in using an automated engine to provide the conduit between the articles that journalists are producing and what users are shown. Isn’t this the editor’s job?
Technology like this will only work if people are actually involved at some point. Twitter’s new search engine is a perfect example of this. They’re developing an algorithm that combines automation with human computation. Once a topic has been determined as popular, based on automation, it’s then sent through to actual people who then analyses it and provides a report to ensure relevant tweets and ads are served around that keyword. In this case, context is king.
Access More Information
While Facebook is developing a search algorithm that, no matter how you look at it, ultimately closes the information gap. You’re not searching ‘everything,’ you’re searching ‘everything that relates to you’ and while there is an argument for quantity over quality, the implications of this must be considered, given how much of our online lives are controlled by Facebook.
Compare it to Google’s new search product which aims to improve information discovery. Coming later this year, The Knowledge Graph is (admittedly like Facebook’s Graph Search’) aiming to provide answers, not just links to answers. But unlike Facebook’s Graph Search, it will provide contextual information around a search query, instead of just a list of results that happen to contain that keyword.
So, you run a search for ‘Da Vinci.’
You then get expanded results next to the usual search results, inviting you to explore further, where you can view related collections in image galleries.
While the personalised web has potential, it will only really work if we have an actual, human element behind it, surfacing the best content or influencing an ad algorithm based on expertise and experience. We need to be exposed to knowledge and the personalised web should work in conjunction with influencers deciding what content is also relevant, looking outside of simply tracking our online behaviour.