Google’s ‘Knowledge Graph’ is changing the landscape of the web

Search on different devices

Google is currently rolling out a new feature called ‘Knowledge Graph’ which will enhance our search experience by intelligently guessing what information we are looking for and putting it at the top of the results page. In other words they will be serving information not links. The picture above was released by Google to show how this will look on multiple different devices.

Many of you will have seen vertically focused instantiations of this strategy in your search results already, and if you are like me you will think they are pretty cool. This weekend I was in Munich and when the discussion turned to the relative sizes of different German cities I was able to activate voice search on my Android phone and ask ‘what is the population of [insert city name here]?’ and get a number straight in the results page without having to do any more clicking. Similarly, I like being able to search on ‘weather xx’ and flight numbers and get the results straight on the page.

The Knowledge Graph initiative is an attempt to build publicity around this strategy and broaden it to more areas.

That’s great for the searcher, and it’s great for Google, but it isn’t so good for the owners of other internet sites. Wikipedia is the obvious sufferer from the population examples I mentioned and I bet weather sites are seeing a drop off in their natural search.

There are numerous examples of other initiatives by Google which arguably improve the experience for searchers but disadvantage other sites on the web, particularly if Google is biased in it’s search results. Google+, Google Places, and their acquisitions of Zagat and BeatThatQuote spring to mind.

I think what we are witnessing here is Google acting more and more like a media company rather than a search company. The distinction is important for anyone who wants to build a business that relies on natural search. In recent years many businesses have achieved success with this strategy, but going forward there is an increasing risk that Google will effectively decide to compete with your business. (Note that this risk is in addition to, although not entirely independent of, the way that Google is increasingly demoting the position of thin affiliate sites in search results pages.)

This is another example of the way the internet economy is moving away from an open collection of sites services with tools that allow us to find the best ones for our purposes at any given point to a world where partnerships and playing nice inside someone else’s empire is the best route to success. Today the most important empire owners are Google, Facebook, Apple, and Amazon, with companies like Twitter, Quora, and Pinterest as challengers.

  • Andrew Buckman

    Particularly interesting was Google flexing their power in this domain at the end of last year but restricting referrer URLs (that allow a destination site to see the search term that drove traffic). If they roll this out across the board then the businesses that are built on search and, more recently, those that are built on search data will suffer. Google has yet to use this kind of data in display advertising though

  • Interesting. Thanks Andrew.

  • Definitely agree Nic, look at what happened to Ciao! when Panda hit and Google Products took off…building a brand that people come to direct is the only way to reduce dependency on search (e.g. I find myself going straight to Quora to look stuff up more and more often). However Google makes money by selling paid clicks, so they will have to make a trade off there. Where will they move after flights, products, places, credit cards..?

  • Yup, Knowledge Graph has a lot of potential; especially when opened up to third party sites via activities such as the joint Yahoo/Google/Microsoft See some initial thoughts from me here – – and some non-positive sentiment from the established Semantic Web community here –

  • There was a good article on Search Engine Land covering this:  Amit Singhal suggested that the boxes encouraged more searching. I suspect it’ll come down to the type of content one generates, how rich it is, the degree of interpretation within the information, expertise etc. A lot of information triggers the need to discover more information.

  • Hi James – I’m not sure I buy that argument. I’m sure there are a number of searches that are only after one piece of info – weather forecast, flight status, population of munich etc – these queries will increasingly be satisfied by Google and hence with zero page views and zero ad impressions for third party publishers. In the past we would often have had to click on multiple links to get this data, which was a pain, but helped multiple publishers.

  •  Yes, flight status and population of Munich, I’d agree. For info relating to Van Gogh as per your article, I suspect it’ll generate searches. Thankfully, the numbers will tell us soon enough and we’ll scuttle off and divine the future from that! Hope you’re still on a high from the w/end 🙂