Amsterdam Affiliate Conference > Hunting the Long Tail

Stuart Shaw 9 years ago

Our Head of Strategy James Perrott took to the stage at the Amsterdam Affiliate Conference to talk about how Google is changing search forever thanks to Hummingbird. The session covered how we are seeing a shift from head terms, to the rewards that long tail searches and search personalisation now offer.

Below you can find the slides and full transcript from James’ presentation at the event:

The presentation slides

The transcript

Historically, all you'd need to do is create a piece of content or even point links at a commercial page for it to rank. This meant it was a case of who could build the most links, not who provided the best user experience and content.

Vanity keywords was the sole focus of a lot of online businesses due to the instant search volume they are associated with. Also from a vanity perspective, brands wanted, and still want, to rank for the most competitive terms in their niche. Long tail wasn't a 'thing' until recently.

Create a page that targeted a vanity keyword and use old unethical and unnatural link building methods and you would see your page rank well for highly competitive keywords. There was no reward for simply having great content.

But, Google got better. Websites that had used these unethical methods got caught out and businesses/websites suffered massively as a result. A large scale clean up operation would be required.

How did Google get better at catching these methods? Google Penguin. Penguin looks at the legitimacy of a link and deems it to be either natural or unnatural. If you're entire link profile is based on unnatural links, you'll be impacted. Penguin does not refresh regularly, which makes recovery hard.

But... this is a whole different topic. We're specifically looking at content and the algorithms important for this is Panda, including their manual actions.

Example 1: Interflora

A manual Panda action from Google impacted them for the use of advertorials on mass scale. Incentivising newspapers and bloggers to take part in a review prior to Mother's Day was deemed unnatural by Google and impacted them significantly. They overcame this quickly due to it being a manual action and removing the content.

Example 2: eBay

eBay mass launched a significant number of /bhp/ pages that were created to drive a better user experience for popular products. The problem; no content was on the page and the structure was identical but on the scale of hundreds of thousands. Due to the exact same structure and no copy, a manual action was applied and the impact was said to be around $200 million. Our job? To identify those pages penalised and help create content to aid their recovery and drive that better user experience they were seeking.

Hummingbird has completely changed how content is measured and rewarded by Google due to the shift in how users search. Conversational search has significantly increased due to mobile usage and in turn user metrics such as bounce rate, time on site, pages per visit etc are suddenly all crucial metrics.

So how is it now?

Google has changed how it sees queries, they're broken into two elements; the implicit search and explicit search. Explicit search is the actual search query and implicit is the data Google has collated about you - you're on your iPhone, in London and are 32 for example.

Because of this mass shift in the algorithm, long tail keyword opportunity holds MASSIVE potential.

This graph shows how search queries are shifting away from just vanity terms into long tail and the importance of creating a content strategy to capture this.

The initial reaction was for people to mass create 'how-to' content due to the long tail search opportunity. Google has quickly quashed this with a specific update to cull the amount of thin how-to content on the web. Only good content is rewarded.

Everyone has micromoments. A micromoment is a moment of need when a search is required to answer your problem, and quickly.

A few examples:

• I-want-to-know
• I-want-to-go
• I-want-to-do
• I-want-to-buy

Micromoments have surged in the last year especially due to the usage of smartphones increasing exponentially.

This increase has seen Google react with releasing a mobile specific algorithm and penalising websites who deliver a poor user experience. Having small font, links too close together etc are a few of the performance indicators it uses to determine whether or not a website is mobile-friendly. There is talk of a entirely separate mobile index being created, but the resource investment to see this through is massive and will take time, and may not happen at all.

Due to the increase in smartphone usage and the average Joe becoming much more tech savvy, the demand for relevance is high. If a user does not find what they're after instantly, they will look elsewhere. We are that impatient now due to the access we have.

If you capture this traffic when demand for relevance is high and patience is low, you will attract the right type of audience/customers. A study by Microsoft has shown that the average attention span has gone down from 12 to 8 seconds since 2000, this is how demanding we have now become.

Another micromoment is in-the-moment, we use our phones for purchase decisions, unexpected problems, the pursuit of big goals in small moments (such as saving, using a mortgage calculator etc) and in routine moments (such as for seeking different hairstyles whilst doing your hair each morning etc).

This information discovery largely sits between the ages of 18-34.

There are even more micromoments to consider for your users:

• Is-it-worth-it
• Show-me-how
• Time-for-a-new-one

So, how do we ensure we capture these micromoments for your users/audience?

We have a tool called SCOT which measures functional content (on-page) to ensure it's delivering what Google wants in order to perform well. Here are a few examples of how it measures npower.com's funciontal content. This looks at vanity and long tail and shows us how well and how bad a piece of content is optimised for a particular keyword.

SCOT shows us that the better optimised a page is between 80-100, the better the average ranking. This is not a coincidence and it allows us to instantly identify good and poor performing pages to focus our attention.

Roadmap is another tool we use to continually check on ranking signals across 100,000 keywords in fiercely competitive niches.

The creation of personas through data insight is one of the first stages to identify who you're actually creating content for.

Social data is awesome for this. I'm going to run through an example that we've done recently on GoDaddy's social audience.

After looking at the audience's gender, age, bio, work, hobbies, relationship status and other likes, we're able to create four personas which accurately represent the GoDaddy audience.

From this, we create a moments map for each individual. This includes all of the micromoments I mentioned previously. What moments are your users going to go through and what will they be searching for, map this out and you will be there for your existing and new customers.

Mapping the moments and personas is one piece, then follow this up with manual and keyword research. Look for existing search queries that are popular and a Google Answerbox returns for - more on this later.

Let’s run through how a typical user journey looks today. For example, recently I wanted a new TV.

The first moment I go to is the is-it-worth-it and i-want-to-know moment when searching for the latest technology - 4K.

We can see that for the search 'what is 4K TV?' a Google Answerbox returns. This takes click throughs away from the first search result and it's not the first article to organically rank for this keyword... odd?

You're probably thinking - how do I do this?

Google uses human manual reviewers to ensure the best content returns in these answer boxes, otherwise it wouldn't be informative, which is Google's thought process behind the introduction of these.

What does it look at? Does the title of the content correlate with the user's search? Is the content structured well with heading tags that answer more questions on the same topic?

We can see that Which have done that perfectly. The h1 tag matches the question and the first paragraph answers the question succinctly. This is what's included in the answer box.

I then decide that 4K technology is worth it and search for "4K TVs".

This Curry's result returns and does not try to instantly sell me a 4K television, it educates me even more with the option to see their products as a small call to action at the bottom. Brilliant.

I'm convinced, I-want-to-buy a 4K television.

I search for "Samsung 4K TV" as I'm a bit of a Samsung fan in regards to electrical equipment. I found this brilliant article on tech radar and more importantly, a TV within budget and it's Samsung!

I then specifically search for this model and Richer Sounds returns with an insane amount of offers. All are great offers and the price matches competitors.

Purchase made.

You'd think that as a brand, the journey ends. It doesn't.

There are plenty of i-want-to-do moments with my new TV. I want to connect my Macbook to the TV and again, an answer box returns.

If you identify these moments at the beginning of your content strategy and follow the right content structure, you have a great chance of a) being included in the Google Answer Box, b) ranking well for this keyword and c) delivering a great user experience.

In conclusion, the Google search journey I performed has completely changed and you have to change with it in order to capture the opportunity.

1. Have a mobile-friendly and technically healthy website
2. Create personas
3. Map their moments
4. Create relevant, informative content that is mobile-friendly and structured correctly

Thanks!