The enterprise SEO platform. Redefined.

Toby Shaw

Toby Shaw
I started at Linkdex a couple of years ago after graduating from Imperial College London and since then I've been working on data analysis and product development. When I blog, once a month, I'll be looking at the world of SEO from a scientific perspective.

What Google’s Semantics Update Will Mean For SEOs

A few weeks ago Google started to release information stating that it would gradually move away from the traditional search model, where a query is answered by a list of links, and more into the business of actually giving information to the user, gleaned from the contents of the webpages that would otherwise be getting traffic from the search. For example:

Semantic Search Result

Implications for SEOs

This isn’t an Earth-shattering news – they have long moved away from just seeing search queries as isolated words. As long ago as 2003 they started looking at the actual semantics of search queries and analysing the meaning and synonyms of words typed into the search box. They’ve been giving information such as cinema times, flight details and numerical constants for nearly as long. It’s a bit of a mystery why their hype machine has kicked into overdrive about this particular small update, but it’s a good reason to have a look at semantic search from an SEO perspective.

For most sites, the effect will probably be minimal, at least in the short term. Some sites will be quite heavily hit – those sites that currently perform very well across wide varieties of searches that offer just pure information will not fare well. Wikipedia, answers.com, ehow.com and other sites currently getting very high traffic just based around facts might, to varying extents, be sidelined.

In the longer term this might cause a complete change in the way SEO is done, with sites potentially moving away from targeting the traditional keyword-based SEO and instead trying to become trusted sources of factual information, in order to drive traffic and rankings. In my opinion SEOs should try and make the most of this new way of being featured in Google.

It remains to be seen how heavily Google will credit and/or reward sources, with either a basic link below the information or with increased rankings elsewhere, but they will doubtless be rewarding quality information in some way. At the moment it looks like they are just linking to their sources with a show/hide toggle:

Semantic Search Sources

Along with Bing and Yahoo, Google have launched schema.org, a set of HTML tags that are read by the various search engines in order to increase the efficiency of data extraction. Widespread use of these tags would be massively beneficial to search engines trying to glean all possible information from a webpage.

Why Make Search More Semantic?

The new update seems to be paving the way for Google Assistant, Google’s answer to Siri, which will be released for Android later this year. When it is asked a question, it will have to come back with an actual nugget of information, rather than a list of sites where you could potentially find it out yourself. Google is just integrating this functionality into traditional search.

It is unarguably a good thing for users, who will be able to get their requested information much more easily. I’m slightly struggling to see how Google will manage to increase revenues this way, especially as it gets more and more widespread over the coming years. Google’s monetization model where paid links generate the vast majority of their revenues means that it is imperative for the company to push as much traffic onto external sites as possible, with the hope that a decent percentage of the people searching will be clicking on the paid links. Where is the incentive for someone to do this when they already have the information they need?

But fundamentally Google isn’t trying to directly increase revenue with this update. It is trying to cement its place as the go-to search engine in the face of up-and-comers like Bing. One of the reasons Google has managed to stay ahead of the game for so long is its refusal to focus on the bottom line. Google’s main mission statement has always been to serve the user first and foremost.

Recently, Bing has been making an effort to improve their search engine’s actual understanding of the web, buying Powerset, a tool that sells itself as being able to actually ‘understand’ the content of web pages. Apple obviously already has Siri, a tool approaching the problem that Google Assistant is trying to solve from a different direction. It simply remains to be seen whether Google can hold on to its market share despite the shifting priorities and methods of ‘searching’.

How to get data out of Twitter

Extract Influential Twitter Users

Social NetworkSince I wrote this post Linkdex have released market-first data extraction tools that make it easy to find all the bloggers in your market, discover & analyze networks of Twitter influencers and visualize their relationships. It’s worth reading more about these features to understand the best new ways to get influencer data from Twitter.

But below is my original article explaining how you can extract the number of tweets for any URL. Handy if good blog posts don’t have social share buttons to show you how popular they are! My free tool below is perhaps still the best way to do this.

Retrieve Tweet Counts for any URL

The modern SEO professional needs to be able to do a wide variety of things quickly and easily. Linkdex helps you do a lot of these things extremely well, but one thing we don’t offer (yet) is easy social media integration. To that end, today I’m going to give you a simple way of getting Twitter data yourself as a placeholder for the top-notch social tools we have planned.

Tweet Count for any URL

The number that shows up on the ‘Tweet!’ button on pretty much every website in the world is actually fairly carefully cloaked by Twitter. Trying to crack your way into the Javascript is quite complicated. There is an API call that you can do to get this data in bulk, however, and I’ve written it up into this Google Doc.

All you need to do is copy the document into your own account, input the URLs, and let it chug through them. It’ll then give you an output you can use for whatever you want. It’s easy to customise. If you want to look at Facebook ‘Likes’ rather than tweets, you can simply swap the twitter API URL in column B for the Facebook API URL (“http://graph.facebook.com/?ids=”) and do some jiggery-pokery with the substitutions in column C.

There’s one problem – Google only lets you do 50 API calls on any one spreadsheet. That means, if you have thousands of posts to process, you’re going to need a better way of doing things. That’s where we have to move into actual programming.

This is a Python script. It’s incredibly easy to run – just install Python 2.7 and put my script file together with a CSV file of the URLs you want to check into the ‘scripts’ folder. Then go into the Command Prompt of your computer (run ‘cmd’ from the Start menu) and do something like this:

Command Line

It’ll save a CSV file of the output, which you can then easily get the data out of. Simple – but not as simple as just using the Google Doc, admittedly.

Why This Tool is Useful

Data like this can be used for a variety of things. I used it for some interesting analysis on what kind of posts generate the most tweets and links (that you can find in an earlier post on this blog), which would obviously be useful to any website trying to optimise their content output for the social media world.

Running analysis like this on your own site can help you learn what people enjoy reading about on your blog, and can help you tailor your content towards what will get you noticed in the social arena. Alternatively, you could run the analysis on link prospects – if you want to know which sites it is best to do a guest blog post on, for example, you could find out easily which would be able to get you the most mentions. You can take your link prospect data out of Linkdex and feed it into the Google Doc, and it will give you a great idea of who to try to contact to maximise your post’s impact on the social world (and therefore Google rankings, hopefully!)