The changing face of search
Personalisation of Search
I was recently invited to give a talk at BAFTA, Home of the British Academy of Film and Television Arts and this post is based on just some of the highlights of that talk.
There’s been a lot of talk about personalisation of search results over the past few years but it’s only recently that people have actually started to notice any differences in their search results.
So just how personalised have search results become? Sometimes the differences in results are quite subtle. At other times they are quite extreme and it then becomes clear that Google really has been taking an interest in your browsing habits and that it does indeed know where you prefer to shop and what you like to read or do online.
A quick search experiment
To find out for yourself, here’s a little experiment you might like to try to see just what Google thinks about you.
For the experiment, you will need access to the computer you use regularly to work and search on and a person nearby with similar access.
Ask the person next to you to search for some random search terms; say ‘holidays’, ‘shopping’, ‘shoes’ for example. Each time, compare the search results. You may find, depending on the degree to which your results are personalised, that there is anything up to a 50% or more difference in the results.
Here’s a search I did, alongside a colleague at exactly the same time last week:
The search term was Vodka. We chose a word neither of us had previously searched for. Lets take a look at the results:
My colleague’s results:
My personalised results:
As you can clearly see, above the fold, the results are strikingly different. I am immediately offered shopping results while my colleague isn’t. Coincidentally I was searching from my laptop, from which I often buy online whilst my colleague was using his desktop machine that he rarely makes purchases from.
Additionally – I am served up Amazon results whilst my colleague isn’t. Another coincidence that I regularly buy books from Amazon, so Google sees no reason why I shouldn’t want to buy vodka from Amazon as well?
Other search results yielded similar differences.
So what does all this mean? Well for one thing, there’s an algorithmic gatekeeper telling you what results you can see and what you can’t.
For companies with websites, its a completely different story. As a business owner it has long been good practice to fire up Google and key in the search terms that you want to be found for and hit the button to display the first ten results.
If you were on page 1, fantastic! Assuming of course that your snippet was compelling enough and optimised well enough to attract click-throughs.
Up until recently you could could then go about your business, confident that because you could find your own website on your page 1 of Google, it was almost certainly on everybody else’s page 1 for the same terms.
Fast forward to today and onto the next 18 months; with the steady erosion of the once monolithic set of search results, there is no longer a single page 1. We’re seeing multiple page 1’s. Our own individual page 1’s if you like. This effectively means that you can no longer be absolutely confident that your website is on all of your audience’s page 1’s.
The question is changing from, am I on page 1? To, what percentage of my audience can see me on their page 1?
Personalisation doesn’t just end with Google’s results. Algorithmic filters are everywhere, from Facebook, through to Amazon and multiple online channels, deciding for us what news we see, the content we are fed and the ads that we are served. If I spend any time on my sons computer it’s like looking at the outside world through the eyes of an 8 year old, being served ads for pokemon amongst search results aimed at a school kid.
Where did it all begin?
long before personalisation, in the days when Google proudly displayed the fact that it indexed 25 million pages, search used to be easy – providing you know how to search! The problem started because the majority of people never really learned how to search properly, using inverted commas and searching for specific doc types for example. So as Google’s index grew, searching for many was a bit like trying to take a drink from a fire hydrant! A bit overwhelming!
So Google tried to help us by starting to serve up the results it thinks are the most relevant to us. Whether you’re logged in to Google or not, Google still tries to help you.
While this may be very handy if I’m out in London, searching for a restaurant on my iPhone, and Google uses my location and penchant for spicy food to steer me to a nearby curry house, I would always prefer the option to choose.
So until Google decides to offer us the option to genuinely choose between algorithmically filtered results or raw results based upon genuine authority, just remember; what you’re looking at is almost certainly not the same as everybody else, even if you think it is!
This story will continue…