Google implemented 1,653 launches, 9,800 live traffic experiments, 18,015 side-by-side experiments and 130,336 search quality tests in 2016. In 2017, there were five confirmed updates and a bunch, many of which were algorithm updates, have been rolled out in 2018.
Google is rewriting the SEO landscape with every move, changing the way our companies rank and drastically shaping the way we work as marketers and advertisers.
Big Changes Thanks To AI
The numerous changes to search that have occurred during the last couple of years are mainly due to Google’s artificial intelligence-first approach, which they adopted in 2016.
This was when Google purchased artificial intelligence (AI) specialist company DeepMind to fuel their strategy, and first implemented RankBrain, the machine learning (ML) algorithm that analyzes the context of content to create more accurate organic search results.
The company also began using ML to derive performance insights and improve ad targeting in AdWords and Google Analytics.
How RankBrain Works
RankBrain was Google’s third most important ranking signal in 2016.
From the outset, it was essentially tasked with understanding entities (such as a thing or concept that is singular, unique, well-defined and distinguishable) and then working out how they connect in a search query in order to understand that query better and therefore display a set of known good answers to the query.
The system then used its machine learning capabilities to teach itself how to recognize synonyms and instruct other portions of the algorithm to produce the correct search engine results page (SERP).
Now, RankBrain is increasing its efforts in making its software think more like a human, using every possible piece of data—from social and user-generated to browser footprint and click patterns—to match individual searchers to the best answers to their questions. And if it can’t, to take a highly-educated guess!
RankBrain converts the textual contents of search queries into “word vectors” or “distributed representations”, which have a unique coordinate address in mathematical space.
Machine Learning in Google AdWords
Google AdWords also leverages the power of ML. It can be used in a variety of ways:
- You can set up cost-per-acquisition (CPA) bidding algorithms within it to save you time and effort in managing your CPC bids and modifiers for device, location, and interest.
- You can use Ad Customizers to build out large quantities of keyword-specific creative and automate the creation of large numbers of adverts as well as give Google ML algorithms more data to work with.
- You can have them build and continuously refine your “similar audiences” and trail new ones.
- You can take advantage of Google’s use of ML to predict when users are in-market and ready to buy specific products.
- And you’ll soon be able to use Google’s free attribution tool, which will use ML to help suggest cross-channel budget adjustments.
What This Means For Marketers
At this point in time, RankBrain and the other Google updates that affect SEO and AdWords, usually involve machine learning and/or artificial intelligence in some way or other.
However, we often have to deal with the knock-on effects of how they change search, so understanding the general trends that affect ranking is key to understanding the whole picture.
So how is all this mysterious business affecting our lives as marketers, and what can we do about it?
So much! Let’s take a look.
SEO & Voice Search
AI technology was undoubtedly used to discover that 20% of mobile searches are actually voice searches in 2016, and this is a trend that only appears to grow in significance.
In 2017, 60.5 million people used Siri, Cortana or another virtual assistant at least once per month. Voice search trends have massively changed the keywords successful marketers use, as people speak in a different way than they type, using more natural language patterns.
To make Google’s AI happy, your content (Google’s second most important ranking signal) must be based on question queries that use who, what, when, where and how. Use natural, full sentences and write blog posts answering your audience’s questions and concerns.
You can also focus on medium-tail keywords instead of long-tail, as RankBrain understands that two similar long-tail terms often refer to the same thing.
Here’s an example:
The End Of Link Building?
Recently, Google patents have begun to suggest that the giant may start using a tracking system for ranking that pays more attention to “linkless” mentions.
After manipulative link practices of marketers were penalized with the Penguin update (mostly due to the fact that links were number one on Google’s list of ranking signals back in 2016), smaller brands are rejoicing as unlinked brand mentions and the sentiment behind them appear to be replacing linked mentions as a site authority signal.
To make use of these changes, and rank better with search AI, focus on developing a well-rounded online footprint that includes PR-focused outreach methods and a social following that talks about you across the web.
AI Allows For Highly Personalized SERPs
SERP personalization has been around since 2007, but recently, Google’s algorithm has become very sophisticated, serving up results based on local geography, search history, interests and activities performed on other Google apps.
One benefit of personalized results is that if users do click on your site and engage with it, they’re more likely to see you again in their search results.
To make the most of this algorithm ability, use location. Rank Tracker is a tool that customizes rank tracking by location, so you get more “real-time” results for users and target more effectively.
AI Can Rate The Truthfulness Of Your Content
Another reason why AI is changing SEO so much is because of its ability to assess all content based on the accuracy of the facts contained within it. Google calls this its “Knowledge-Based Trust” method.
Google’s truthfulness algorithm is getting smarter by the day, differentiating precisely between fact and fiction. This can be seen in the new “Fact check” tags added to Google News posts, which (finally) seems like it might spell the end of fake news.
Being evermore user-focused and producing high-quality, valuable content will be rewarded.
AI’s New Rules: RankBrain Optimizes For Maximum User Value
Circling back to RankBrain, the ML algorithm that processes search queries, determines a page’s value to searchers and informs Google’s model of SERP indexing… let’s talk about the secret to getting it on board.
RankBrain mainly measures two factors:
- Click-through rate (CTR), or the percentage of times a user clicks on a result
- Dwell time, or the time a user spends on a page before clicking away to the search results to find a different page to look at
This means your content must always be of high-quality and value to users, or you’ll get outranked in real-time by others with better content. Your SERP presentation, including optimized title tags and meta descriptions, increases CTR, so that’s where to focus your energy for RankBrain to play ball.
AI, SEO & PPC Advertising
The type of AI responsible for the success of PPC today is based on statistics and ML, which categorizes everything very efficiently indeed, matches it to a wide range of dimensions like geo-location, hour of day, device and audience, and then uses a neural network to create an impression at the perfect moment in time and space. Then it generates another impression in the form of remarketing.
When prediction comes into play here, things can get even more interesting, and the Quality Score (QS) is a great example of this. This is where Google analyzes users’ historical data behavior using ML to find correlations that help it predict the likelihood of a click and/or a conversion. Now there’s efficiency.
Many of the above points on marketing with AI SEO in mind also apply to PPC, as, at its core, it comes from the same principle: delivering high-quality information, whether that’s an ad or native content. That’s why using your own AI tool can help you improve your PPC results even more dramatically.
Artificial Intelligence Ad Tech solutions like Adext AI, an audience management as a service tool, can increase your conversions intelligently and automatically within Google or Facebook Ads, applying up to 480 budget allocations and audience optimizations in each one of the ads in your campaign every single hour.
But the best thing about this tool is its transparency. You can deploy it in just 5 minutes and start seeing how it makes hourly updates to your budgets live!
And is when you see the way the advertising industry is behaving in this AI-driven era, you realize being an innovator and an early adopter pays off... whether it's for paid or for organic marketing.