enewton@brightedge.com

M Posted 8 years 1 month ago

t 9 min read

Google announced a number of AI-driven product enhancements at its annual Google I/O conference on May 10. It is delivering materially on its positioning as an AI-first company. To Google that appears to mean injecting AI into all products and services to make them smarter and more useful. Just before the keynote, Google announced it is rebranding its Google Research division to Google AI. The move further emphasizes how Google is focusing R&D on computer vision, natural language processing, and neural networks. As part of this process, it is also subtly expanding the scope and the possibilities of Search.

Google Assistant Gets Really Helpful

Google Duplex is taking voice and AI to a new level as Google CEO Sundar Pichai demoed the Google Assistant calling a hair salon and making him a hair appointment. Some people declared the demo passed the Turing Test, where the AI is so good that the person does not know they are talking to a computer. Google will make conversation more natural by not requiring the "Hey Google" or "Ok, Google" and using more natural timing and cadence. While this technology may not have a direct relationship to Search on its surface, on closer inspection it does bring up some interesting questions. Voice is already one of the most significant transformations to the way people search online, often presenting direct verbalized answers to users' questions and bypassing the SERP altogether. Voice already represents a full 20% of Google searches made on mobile and Android devices as of 2016, and one in six adults in the United States possesses some form of "smart speaker" device as of 2017. It's clear that this is a technology set to become a potential tidal wave. That said, Pichai's product demo is another reminder that as search marketers we must continually question whether we're looking at SEO for Voice the right way, as we do for any other type of optimization. If the future of Voice is moving from queries like "which hair salons are open on Tuesday?" to directions like "make me a haircut appointment on Tuesday," how will that affect SEO? If a user issues a direction to buy batteries, but doesn't indicate a brand or ecommerce platform preference, how will the voice assistant decide which options to present first? Will it present options at all? It is an opportunity for brands to take advantage of even more certain and commercial search intent.

Gmail Starts Suggesting What to Write

In addition to the existing smart reply feature, Gmail will soon be able to suggest full sentences for you while you write with its new Smart Compose capability. Machines are already writing more than most people realize, playing a role in sports, weather, and simple news updates. Companies that use machines to write, do not usually publicize it.

Google Maps Will Augment the Reality of Your Pedestrian Navigation

Google Maps is getting more friendly and social with its For You tab where it will feature new and interesting restaurants and businesses in your neighborhood. And it will soon combine street view and camera to add augmented reality directions that will help you navigate the real world with a layer of video and audio. As we discuss in more detail below, Google is expanding aspects of its search experience - in this case local search results - and surfacing them areas other than a basic search results page. Someone walking down the street in an unfamiliar city may not have the time or inclination to navigate away from GPS navigation in Google Maps to look up "restaurants in my area," so now they could have the results of that local search query tiled over an augmented reality viewpoint. This opens up another avenue for local SEO to get in front of potential customers at exactly the right moment. Additionally, this move is a reaction to localized vertical search engines like Yelp and FourSquare. By answering the need to present a competing feature set to these VSEs, Google indirectly acknowledges that VSEs are now a significant force in Search.

Google News Will Let You Personalize Your News Without Becoming too Insular

Google News will also being using AI to find and curate news customized for the user. Additionally, it is designed to provide you "a range of perspectives" to expose users to ideas outside their core. Google seems to be doing some beneficial social engineering to provide personalized content while keeping society from becoming too fragmented and insular.

Google Lens Will Help You Make Physical Virtual

Google Lens will help you take real-world text and make it editable and available to use inside your phone. And it you can point it at a product and it will help you find a place to purchase it online. Google is in a continual process of refining its search experience to present users with the exact content that they need in response to the exact query or intent that they have. This was the rationale behind all of the named search algorithm updates as well as the introduction of the customer micro-moments model. By introducing contextual ecommerce listings to Google Lens, Google is applying the same philosophy horizontally across its technologies. An outgrowth of the same strategy can be seen in the recent addition of ecommerce aspects to Image Search via structured data tags. Google is trying to get even more dialed in on micro-moment targeting by expanding the breadth of where search results-style listings are presented to the end user. Previously these listings may only have been surfaced to the user through the traditional SERP - but is that where every digital audience is beginning its customer journey now?

You have heard of voice search, but have you heard of thought search?

As futuristic as it sounds, an MIT student has a working proof of concept of his AlterEgo product that allows him to think a search, search Google, and get the answer silently vibrated into his ear. Arnav Kapur was featured on “60 Minutes” and did calculations, found trivia information, and ordered a pizza using only his thoughts. The device is not connected to his brain, just worn on Kapur’s head. When the user thinks and internally vocalizes a specific command or question, it is conveyed to a computer and Google — sort of like silently Googling something in your head. Electrical signals that the brain normally sends to the vocal cords are intercepted and sent to a computer and that information is then communicated to the user’s inner ear via vibrations. Google is expanding not just the places where search happens, but the ways in which we conduct searches altogether.

How does BrightEdge use AI for SEO success?

BrightEdge Insights uses deep learning and big data to recommend the critical actions that will drive the biggest impact for your business. Insights gives you the confidence that you are focusing on the content or optimization issues that will drive the biggest impact today. Insights prioritizes all of the actions in one simple feed, placing the most critical actions at the top. Instead of spending time on data extraction, manipulation, and syntheses, you will get back hours each week by taking recommended actions from Insights that yield quick wins in SEO. With artificial intelligence, Insights does the heavy lifting to uncover the most relevant actions you should take. Insights looks at millions of web pages, examines thousands of changes each week, and boils them down to a few key findings that will move the needle for you.

Takeaway

What do these advances in AI mean for marketers? The advice remains consistent: build relevant, quality content that provides a great user experience. As Google uses RankBrain to understand intent at higher specificity, more detailed and specific (longtail) content is likely to be in first position. Additionally, watch for developments in technical SEO and markup that will make your content more discoverable and desirable to the AI-savvy bot and algorithms.

ksikora

M Posted 8 years 1 month ago

t 9 min read

Last April, Noble Performs attended the Share Global Insights Conference in London, where I was delighted to speak on a panel, discussing the roadmap to digital success.  During the panel, we took some time to reflect on the trends and strategies that are driving SEO and content marketing at the moment. During our panel, the subject of keyword clustering (or topic clusters) came up on several occasions. Keyword clustering has been a key component of our SEO approach for some time, so we’re pleased to see other industry players getting on board too. If you’re not already in the know about keyword clustering, also known as topic clusters, we’ve put together a brief guide to help navigate a shift towards this strategy. By its very nature, Search Engine Optimization is constantly evolving. You can pretty much bet that the SEO strategies and tactics you’re using today will be in need of a refresh in six months. New technology, changes in consumer behavior and alterations to search algorithms are among the factors that influence updates. Constantly monitoring industry trends, then learning and testing techniques in the real world requires a huge investment of time from a digital performance marketing agency like ours but is wholly necessary to provide the best possible outcomes for our clients. As a result, coming across a strategy such as this one that has the potential to transform the SEO industry for years to come- has gotten us pretty excited!

Keywords: The good, the bad and the ugly

At the core of SEO is the concept of the keyword or search term. SEO is formed around the premise that someone is actively looking for information on a specific topic; that they have intent. This differs from the traditional advertising and marketing model, where a company creates a message based on what they consider to be in the interests of their target audience. For most people, search starts by asking Google to look for information on a particular topic. The precise words used in such a search have always been crucial for practitioners in SEO. As marketers we would consider:

• “Is this the way my target audience is really searching for what we offer?”
• “Does this method of search show an intent to purchase our product or service?”
• “Would an alternative word work better to describe what we offer?”

In order to get an answer we have to make a number of assumptions. On the plus side, as experienced SEO professionals, we’re able to make a lot of accurate speculations. We’re able to define when a customer is looking with "research intent" versus when they’re looking with "purchase intent." We’re also able to optimize for plural and singular keyword searches. Yet no matter how good we are at this, we still miss a large portion of our potential audience. But why? The answer is that there are many reasons. One is that we have our own personal language biases and our way of speaking has an influence on framing search. Additionally the sheer number of people from a variety of backgrounds who are scanning the web makes it a virtually impossible task to imagine each and every potential search term they could use. To illustrate a simple example, what happens when someone is researching a holiday with their family? Let’s consider some of the possible terms they could use in a search engine:

• “family holiday ideas”
• “family friendly holidays”
• “holidays with kids”
• “things to do with kids in Bristol”
• “best family friendly destinations in South West England”
• “travelling with children”
• “holiday ideas with kids”

It’s quickly evident that there are a number of ways a person could search with the same purchase intent. If you then actually run a search for each of the above search terms you’re then likely to see significant variations in the results that Google displays. One of the major reasons for this is that companies will have selected variations of possible search terms and then committed their optimization strategy to these terms. Essentially, it’s a self-limiting approach.

A search for “things to do with kids in Bristol” on Google If you compare the results of the two searches in more detail, you’ll easily see that the layout of the SERP is very different. This is due to Google attempting to provide the most relevant content for the person doing the search, as well as company’s strategy of choosing just a handful of keywords to earn rank for. Ultimately, this is a rapidly outdated approach to SEO, and it’s time for a more modern, effective method.

Keyword clustering for better SEO performance

To define keyword clustering in the simplest terms, a topic cluster is a way of thematically grouping content. They are formed of a core topic or “pillar” page with related subtopics or “cluster” pages. The cluster content links back to the pillar page, as well as to the other cluster pages. Generally, a pillar page would comprise of a long-form content page that’s rich in information on an extensive topic, while the cluster would be a more specific look at topics that relate to the pillar page. This video from the content marketing platform, Hubspot, is worth a watch. At Noble Digital Performance, we've been successfully implementing keyword clustering for our clients since 2015, working with enterprise SEO platform, BrightEdge. Until this year we’ve been internally referring the to the concept as content siloing but effectively it’s built on the same principles. Through creating a semantic relationship and hierarchy between the individual content we’re able to consistently provide an improved user experience and stronger signals to the search engines on which topics our client’s sites have authority in. Our idea for content silos (or topic clusters) was inspired by our experiences working in the travel industry for several top-tier destination marketing organizations (DMOs). Our observation was that while our clients tended to show a decent footprint and solid rankings in the first two pages across search engines, hitting the top spot for a range of topics still eluded them. To take an example; if you were searching for dining options you would be likely to find that top rankings would be held by sites such as Yelp, OpenTable, and the like who focus on restaurants and dining-related content rather than the higher quality of content available on the DMO sites. Through detailed analysis of site structure, internal linking and the quality of content we established several factors:

• The best topic-related content wasn’t always housed in the main directory, often it was located in the blog
• Related content wasn’t linked to their primary page or together
• Several pages were competing for rank on the same keyword
• Often the pillar page had lower quality content than that of the related content.

A pattern began to emerge, so we continued to analyze and compare for other industries too. Ultimately it became clear that while these sites had gained a large footprint from having a large amount of content (breadth), they invariably had little topical depth. Adopting the keyword clustering model therefore enabled both.

Machine learning and why topic clusters work

While it’s fairly intuitive that topic clusters seem to be a brilliant way to group content and organize concepts, the reason that it’s so effective for SEO goes beyond that. Back in 2013, Google’s Hummingbird update challenged traditional keyword-focused SEO with a crossroads. The search algorithm began parsing out phrases instead of focusing on keywords. In the SEO community, the Hummingbird update is widely acknowledged ad Google’s official switch from keyword to topic focus. SEO professionals were given a further major impetus to move towards topic-based strategies in 2015 with Google’s RankBrain update. This development of Google’s machine-learning algorithm was created to better interpret the content of people’s search queries. Using previous searches as a basis, RankBrain pulls out similar themes and multiple phrases and keywords to provide the best results for that query. When you add a further trend for voice-assisted search on mobile devices (Google attributes more than 20% of mobile search to this), it becomes increasingly obvious that topics are rapidly replacing keywords.

Putting it all together

As digital marketers this means that we need to be increasingly attentive to creating the most relevant content resource around a topic instead of creating a strategy driven by optimizing pages based on a series of chosen keywords. In effect, Google’s advances mean that it can now interpret the intention behind a query and match it to sites that produce the most credible results for that intent. While this requires a fundamental shift in SEO approach, the end result is a far better performance by the search engines, and a more satisfactory experience for site visitors. Rather than simply intercepting keyword searches, topic clusters drive understanding of the buyer’s journey through responding to their questions more effectively. The result is that your content gets the exposure you desire and wins you loyalty too. Keyword clustering is  also an  ideal content creation framework and allows you to leverage the clustered content beyond just SEO, but into an integrated approach. For instance, content clusters can inform your site extension strategy within AdWords tying together your SEO and pay-per-click strategy.

About the author

Kate Sikora Client Success Director Grace under pressure. That’s how others describe Kate Sikora’s ability to do so many things well – without going mad. But with her sanity fully intact, Kate serves as the Client Success Director and overseer of daily operations for Noble Digital Performance. This Manchester native supports new business opportunities whilst working on Noble’s U.K. marketing efforts. An entrepreneur who has successfully run her own business, Kate also helps Noble with integrated digital strategies, finance, recruitment and mentorship of the U.K. team. An avid skier and paddleboarder, Kate moved back to England in 2017 after living several years in the United States. Noble Digital Performance Like peanut butter and jelly (or jam), Noble Digital Performance is made of two things that are better together. It’s the across-the-pond partnership between Noble Studios, a leading-edge digital marketing agency from the US, and the team behind Mr. B & Friends, a full-service brand agency from Bristol. From their home in Reno, Nevada, Noble Studios brings 15 years of digital success with international clients, a contagious can-do attitude and Silicon Valley smarts. With Mr. B & Friends, we sit at the intersection where technology, creativity and ambition meet; where they connect and become buddies (or chums). It means Noble Digital Performance has a simple goal. To harness the power of digital, to consistently and continuously improve what we do for our clients. To combine vision with insight and delivery with results. Together, we’re better.