When you look at Google’s feature releases over the past few months it’s interesting to notice their attempt to keep users on the site for longer periods of time.

Currently the search engine relies on reading text from relevant web pages and linking to this information through rankings, though they seem to increasingly prefer a model of presenting as much information as possible themselves.

A recent step in providing the best information was the release of the Knowledge Graph. This is further supported by an attempt to integrate the overall online experience with the staggered launches of Search Plus Your World and Google+ Local. When you throw Direct Answers and Product Search into the mix you start to see a clear progression, discussed below.

Of course it’s only natural for Google to make these changes and though they will have affected relatively few people there are some interesting dangers and opportunities businesses should be aware of, especially if Google continues down this path.

Providing Answers

For a long time Google has wanted to do more than refer people to other sites. In May 2010 this process became noticeable when they started to give Direct Answers in the results pages. By gathering facts from trusted websites (Wikipedia, imdb etc.) they could give answers at the top of the results, ‘outranking’ the very pages they use to collect the information.

This was one of the first displays that Google could understand semantic ‘entities’ and answer users’ questions without just being a ‘middleman’. After all, why push people onto other sites when you can help them straight away?

Direct Answer

Then earlier this year at SMX London Amit Singhal again confirmed Google were focusing on semantic search to better understand questions so they could improve navigation between entities. Amit’s vision for the search engine is a Star Trek style database that answers all questions and links all content in the best way possible. But this also raises questions about the impact on linked results when Google does answer questions themselves.

Knowledge is Power

Beyond providing scraped data in ‘best guess’ Direct Answers, Google released the Knowledge Graph a couple of weeks ago on May 16. Acting in right hand panels, the Graph provides additional information and links to related searches. I actually like how it allows you to browse through related content with ease, akin to ‘the Wikipedia Effect’. Click on the image to enlarge:

Knowledge Graph - The Shard

Since the US launch of the Knowledge Graph Google have reported an increased amount of searches taking place (which therefore increases the amount of advertising revenue they receive), though they’ve not said anything about changes in CTR on the results themselves.

Some publishers are concerned that certain websites (especially fact-reporting sites) will lose traffic as Google presents crawled information in the SERPs, removing the need for a click on the ranked pages. Amit Singhal’s rationalization is that more searches means websites have a greater opportunity to capture clicks.

The Knowledge Graph may indeed increase curiosity but if users’ following queries are also answered immediately then it doesn’t necessarily encourage more page clicks.

However, as noted by Danny Sullivan, Direct Answers and the Knowledge Graph affect relatively few sites. Beside these fact-providing websites it mainly draws information from Wikipedia and other large sites which would likely be unaffected.

Having said that, the Knowledge Graph does potentially affect transactional searches. For example searching for ‘Machu Picchu’ presents not only ‘fact’ websites but – further down- tour companies. It also presented me with adverts beneath the Knowledge Graph for tour companies (though when I returned to take the below picture they had gone). There are similar competitive searches in the music, travel, sports and film industries which are now displaying a rather distracting Knowledge Graph. I eagerly await professional studies on the CTR of PPC adverts and rankings when this graph is present!

Knowledge Graph - Machu Picchu

It will be interesting to see how this develops in line with the really competitive terms. Also remember that once upon a time Search plus Your World brought up a similar ‘People and Pages’ pane with key G+ pages (below), potentially distracting from the articles and adverts. But that has since disappeared. Though the Knowledge Graph is by no means final, it does demonstrate that Google are dedicated to providing a ‘semantic search’ panel to encourage related searches.

It also seems like a logical progression for them to create individual G+ pages for each ‘entity’ and allow users to +1 or comment on different topics. What better way to integrate Google+? Remember – you heard it here first! (That means you, Amit).

People and Pages for News

How to Benefit

Not that there’s really anything you can do to fight Google’s changes, but there are a couple of opportunities that come with the Knowledge Graph. They may not be immediately relevant to many websites (since the knowledge graph doesn’t cover car insurance or credit cards!) but it does reassert the importance of some essential SEO tactics:

  • Optimize images

Most ‘entities’ represented in the Knowledge Graph so far have been assigned an image which appears to be sourced from the highest ranked picture in Google Image Search. This brings images into the main search pages and, more importantly, this image is a direct link to the host website.

And this image even appears in the Knowledge Graph for other related searches, so you could potentially have a prominent image-link on a number of related searches that you wouldn’t ordinarily rank for.

It is unlikely you will rank for something in the current Knowledge Graph but image optimization is a good practice which could now become more powerful, especially as Google expand the Knowledge Graph.

Try to ensure you are optimizing your images and attracting links to the image itself and to its host page.

  • Seize previously neglected keyword searches

Say you run a film blog. In the past you might have found that traffic always came via searches for ‘The Artist’, even when users were actually looking for director Michel Hazanavicius (perhaps they couldn’t spell his name). In the past it would have been wise to optimize more for ‘the Artist’ than ‘Hazanavicius’. But now when they type ‘The Artist’ they are presented with the director’s name in the graph. With an easy click they are making the search they always meant to make.

Knowledge Graph - The Artist

This development could well spread importance to previously neglected and even long-tail searches. At the moment this might not be too relevant to your company website, but this innovative feature has some valuable implications that I imagine Google will want to expand upon.

Are there any pages on your site that people navigate to and spend time on but don’t necessarily land on initially? Google have created a system which helps people find obscure keyword phrases they might be interested in but wouldn’t have otherwise used. Whether or not it applies to you now, it’s generally a good tactic to optimize for long tail searches.

  • Optimize for more dedicated searchers

Following on from the above, if Google continue to present information themselves people could actually complete non-competitive, low-value informational queries quickly and continue to search for more valuable transactional queries.

In other words, the research people used to conduct on Wikipedia and other fact sites could occur sooner on Google itself, allowing them to change to transactional queries and click on businesses sites in place of research sites.

As I mentioned, at the moment Direct Answers and the Knowledge Graph don’t appear for many competitive searches, but it doesn’t mean they won’t in the future. And, as with the other ‘tips’, these are all good things to optimize for anyway that might even increase in importance in the future.

  • Implement Structured Data

As I noted in my earlier Schema post, structured data can help Google read your website and present it in the best way possible. The more they move toward semantic search and answering questions up front, they more necessary this will become. Why help Google take your information if they’re going to bypass your site? Because other sites will.

Social and Local

Beyond the Knowledge Graph and Direct Answers, if there is proof that Google wants you to spend more time with them it is in the launch of Search Plus Your World in January and the release of Google+ Local last week.

Search Plus Your World

On January 10th of this year Google released Search Plus Your World (SPYW). The social/search hybrid saw Google+ pages appear in the SERPs and even outrank long established pages. More likely to continually affect the average site owner than the Knowledge Base developments, SPYW provided a double-edged sword. It gave user the opportunity of being listed with a G+ profile and also the threat of being outranked by these Google+ pages.

Last week similar opportunities became available to businesses when the search engine announced an update to Google Places, re-launching it as Google+ Local.

Google+ Local

On the above link Matt McGee suggests that, as with SPYW, new local pages (which are integrated into Google+) are being indexed and will likely appear in the SERPs soon and even outrank existing ‘local’ websites. In the same way businesses and employees considered creating a G+ profile in the wake of SPYW, +Local could offer new ranking and traffic opportunities to some local businesses.

When we consider the appearance of Google+ and Google Local pages directly in the SERPs outranking established websites, it reinforces the suggestion that Google is looking to promote their services, even to the cost of existing sites.

They would argue it increases integration and makes search easier and more helpful, whereas others would suggest they’re forcing their unwanted services onto searchers. Either way, it’s better to look at these new ranking options as opportunities rather than threats.

More Time on Google

Another recent development which nicely ties in with the Knowledge Graph is Google Shopping. Look below at how the new Product search uses a Sponsored box on the top right to display adverts. If users become used to seeing ‘definitive’ answers on the top right panels, this form of sponsored listing might become more prolific and even necessary for big brands.

Google Shoppint - Tent

Here they are displaying a panel for a specific type of tent but they could do something similar for almost anything competitive. Think of a Knowledge Graph or Sponsored Panel result for insurance, foreign exchange or web hosting. A definition, an image, a few sponsored links… the horror!

A successful search engine revolves around releasing the right information at the right time and moving people between the right pages. Google are trying to improve the connection between related searches and captivate people with its newly defined entities.

It will certainly be interesting to see how Google continue to push its products and encourage users to find the answers not in Google, but with Google. In some cases ranked pages could almost become more like ‘additional reading’ to Google’s immediate and definitive answers.

Either way Google are making it hard for businesses to ignore their latest developments. Whether or not you like Google, you have to admit their latest string of changes makes it more important to search with them, set up Google profiles and even advertise with them.

What’s that coming over the hill?

Because so many businesses rely on Google to survive it is easy to overreact to every little algorithm change and feature release. But it certainly does seem worrying that Google would rather satiate users’ appetites without them actually clicking any links. How far they might take this is anyone’s guess.

Going forward you might even wonder, if Google is moving away from ‘strings’ of keywords and towards ‘things’ (or ‘entities’), could we project that the actual search function will also move beyond keywords? For example, ‘how tall is Mount Everest’ and ‘Mount Everest elevation’ essentially mean the same thing, yet they return different results. If there is an optimal set of search results for what I meant, why rely on anchor text, keywords etc. and display two different set of results? This probably doesn’t scale, but it could be interesting.

The SERPs are more complicated now than they ever have been. If you currently rank track keywords it’s no longer good enough to give a number, you need to know when the search is variable and infused with other media. You should be using a comprehensive SEO platform to see what else appears in the SERPs. Images, news results, shopping items, videos, G+ pages from SPYW, new +Local pages, Knowledge Graph listings, rich snippets, adverts: these all affect the appearance of the results pages and, it would be reasonable to assume, click through rates.

But for the meantime it would be wise to cover your Google+ profile, +Local page if you’re a local business, mark up your content with schema, optimize images, consider long tail searches and optimize different web pages based on the likely commitment of the user. At least with these often overlooked SEO practices you can better safeguard against Google’s changes and make the most of the coming developments in semantic search.