The shift from keyword search to a queryless way to get information has arrived.
Google Discover is an AI-driven content recommendation tool included in the Google Search app. Here is what we learned from the data available in the Google Search Console.
Google introduced Discover in 2017, and in September 2018 it claimed that there were already 800M active users consuming content using this new application. Back in April 2019, Google added in the Google Search Console statistical data on the traffic generated by Discover. This data is meant to help webmasters, and publishers in general, understand what content is ranking best on this new platform and how it might be different from the content ranking on Google Search.
What was very shocking for me to see, on some of the large websites we work for with our SEO management service, is that between 25% and 42% of the total number of organic clicks are already generated by this new recommendation tool. I did expect Discover to drive a significant amount of organic traffic, but I totally underestimated its true potentials.
In Google’s AI-first approach, organic traffic is no longer solely dependent on queries typed by users in the search bar.
This shift has a tremendous impact on both content publishers, business owners, and the SEO industry as a whole.
Machine learning is working behind the scenes to harvest data about users’ behaviors, to learn from this data and to suggest what is relevant for them at a specific point in time and space.
Let’s have a look at how Google explains how Discover works.
[…] We’ve taken our existing Knowledge Graph—which understands connections between people, places, things and facts about them—and added a new layer, called the Topic Layer, engineered to deeply understand a topic space and how interests can develop over time as familiarity and expertise grow. The Topic Layer is built by analyzing all the content that exists on the web for a given topic and develops hundreds and thousands of subtopics. For these subtopics, we can identify the most relevant articles and videos—the ones that have shown themselves to be evergreen and continually useful, as well as fresh content on the topic. We then look at patterns to understand how these subtopics relate to each other, so we can more intelligently surface the type of content you might want to explore next.
Embrace Semantics and publish data that can help machines be trained.
Once again, the data that we produce, sustains and nurtures this entire process. Here is an overview of the contextual data, besides the Knowledge Graph and the Topic Layer that Google uses to train the system:
This research is limited to the data gathered from three websites only. While the sample was small, few patterns emerged:
Google tends to distribute content between Google Search and Google Discover (the highest overlap I found was 13.5% – these are pages that, since Discover data has been collected on GSC, have received traffic from both channels)
Pages in Discover have not the highest engagement in terms of bounce rate or average time-on-page when compared to all other pages on a website. Yet, they are relevant for a specific intent and well-curated.
Traffic seems to work with a 48-hours or 72-hours spike, as already seen for the top stories.
For news websites and, generally speaking, for websites with a high frequency of publishing, it makes sense to filter Google Analytics results to track Google Discover traffic in real-time. Follow Valentin Pletzer’s instructions to learn how to do it. It is not trivial and he has everything you need to get started.
To optimize your content for Google Discover, here is what you should do.
Would you like to receive this checklist directly in your inbox?
This article is veeery long, but it's also super useful. What about getting this checklist emailed to you to have a reference in the future?
Just fill the form with your email and we'll send it to you right away!
Thank you! The checklist is on your way!
1. Make sure you have an entity in the Google Knowledge Graph or an account on Google My Business
Either your business or product is already in the Google Knowledge Graph, or it is not. If it is not, there are no chances that the content you are writing about for your company or product will appear in Discover (unless this content is bound to other broader topics). I can read articles about WordLift in my Discover stream since WordLift has an entity in the Google Knowledge Graph. From the configuration screenshot above, we can see there are indeed more entities when I search for “WordLift”:
one related to Google My Business (WordLift Software Company in Rome is the label we use on GMB),
one from the Google Knowledge Graph (WordLift Company)
one presumably about the product (without any tagline)
one about myself as CEO of the company
So, get into the graph and make sure to curate your presence on Google My Business. Very interestingly, we can see the relationship between myself and WordLift is such that when looking for WordLift, Google also shows Andrea Volpini as a potential topic of interest.
2. Focus on high-quality content and a great user experience
It is also good to remember that quality is essential, both in terms of the content you write (alignment with Google’s content quality policies) and user experience. A website that loads on a mobile connection in 10 seconds or more won’t be featured in Discover. A clickbait article — with more ads than content — won’t be featured in Discover. An article written by copying other websites and patently infringing copyright laws is not likely to be featured in Discover.
3. Be relevant and write content that truly helps people by responding to their specific information need
Recommendations tools like Discover only succeed when they can entice the user to click on the suggested content. To do so effectively, they need to work with content designed to answer a specific request. Let’s see a few examples “I am interested in SEO” (entity “Search Engine Optimization“), or “I want to learn more about business models” (entity “Business Model”).
The more we can match the user’s intent, in a specific context (or micro-moment if you like), the more likely we will be chosen by a recommendation tool like Discover.
4. Always use an appealing hi-res image and a great title
Images play a very important role in Google‘s card-based UI as well as in Discover. Whether you are presenting a cookie recipe or an article, the image you chose will be presented to the user, and it will play its role in enticing the click. Besides the images’ editorial quality,I suggest suggest you follow the AMP requirements for images (the smallest side of the featured image should be at least 1.200 px). You also want to make sure Google has the rights to display your high-quality images, and this can be done either using AMP or by filling out this form to express your interest in Google’s opt-in program. Similarly, a good title, much like in the traditional SERP, is super helpful in driving the clicks.
Don’t underestimate the value of Open Graph metadata
Speaking about images, titles, and cards, also consider that the Open Graphmetadata is now playing a role in Google Discover. Google might choose to show a piece of content using the og:title, og:image, and og:description tags for the preview card.
Open Graph Metadata are the same ones used by social media channels such as Facebook and LinkedIn to show the preview of a web page. So, knowing it or not, you might be using the Open Graph Protocol to polish the social media previews of your web pages, posts, and articles. Always be accurate in filling the editorial information for social media sharing as it might also be displayed on Google Discover.
5. Organize your content semantically
Much like Google does, using tools like WordLift, you can organize content with semantic networks and entities. Semantic enrichment allows you to: a) help Google (and other search engines) gather more data about “your” entities b) organize your content the same way Google does (and therefore measure its performance by looking at topics and not pages and keywords) c) train our own ML models to help you make better decisions for your business.
Let me give you a few examples. I provide, let’s say, the information about our company and the industry we work for by using entities that Google can crawl. Google‘s AI will be able to connect content related to our business with people interested in “startups”, “SEO” and “artificial intelligence“. As we usually say, machine learning is hungry for data, and semantically rich data is what platforms like Discover use to learn how to be relevant.
If I look at the traffic I generate on my website, not only in terms of pages and keywords but using entities (as we do with our new search rankings dashboard or the Google Analytics integration), I can quickly see what content is relevant for a given topic, and I can improve it. I can also decide to plan more editorial content about the most popular topics.
Here below a list of pages, that we have annotated with the entity “Artificial Intelligence“. Are these pages relevant for someone interested in AI? Can we do a better job of helping these people learn more about this topic?
Ready to transform your marketing strategy with AI?Let's talk!
6. Experiment with Web Stories
A Web Story is a visual storytelling format that immerses the user in a tap-through full-screen experience. Web Stories can appear in Google Search, Google Images, in the Google app, and, also, in Google Discover.
As few publishers are using this format, it’s good to experiment with Web Stories to get a Google Discover spot.
Web Stories include visual narratives, engaging animations, and tappable interactions. Therefore, you can use them to engage users on Google and then bring them to your website, inviting them to learn more.
Be careful here: Web Stories should not have more than one outlink or attachment per page.
To help Google better understand your Web Story, you can add structured data to your Web Story. With structured data, Web Stories can also be eligible for other types of rich results (for example, the Top stories carousel or a host carousel).
7. Create social buzz around your blog posts
If you think about how Google Discover works — selecting content based on a user’s interests — you will notice that it works pretty much like a social feed. Therefore, it should be no surprise that Google favors articles and blog posts that proved to be engaging on social media.
Look at this experiment by JR Oaks, who made it to Discover with a little more than one hundred retweets from the SEO community.
Would you like to receive this checklist directly in your inbox?
This article is veeery long, but it's also super useful. What about getting this checklist emailed to you to have a reference in the future?
Just fill the form with your email and we'll send it to you right away!
Thank you! The checklist is on your way!
Learn more about Google Discover – Questions & Answers
In the next paragraphs of this article, you can find a list of questions that I have been able to answer as data from Discover was made available in GSC. I hope you’ll find it useful too.
How does Discover work from the end-user perspective?
The suggestions in Discover are entity-based. Google groups content that believes relevant using entities in its Knowledge Graph (i.e., “WordLift”, “Andrea Volpini”, “Business” or “Search Engine Optimization“). Entities are called topics. The content-based user filtering algorithm behind Discover can be configured from a menu in the application (“Customize Discover”) and fine-tuned over time by providing direct feedback on the recommended content in the form of “Yes, I want more of this”, “No, I am not interested”. Using Reinforcement Learning (a specific branch of Machine Learning) and Neural Matching (different ways of understanding what the content is about), the algorithm can create a personalized feed of information from the web. New topics can be followed by clicking on the “+” sign.
Topics are organized in a hierarchy of categories and subcategories (such as “Sport”, “Technology”). Read more here on how to customize Google Discover.
How can I access Discover?
On Android, in most devices, accessing Discover is as simple as swiping from the home screen to the right.
Is Google Discover available only in the US?
No, Google Discover is already available worldwide and in multiple languages, and it is part of the core search experience on all Android devices and on any iOS devices with the Google Search app installed. Discover is also available in Google Chrome.
Do I have to be on Google News to be featured in Discover?
No, Google Discover also uses content that is not published on Google News. It is more likely that a news site will appear on Google Discover due to the amount of content published every day and the different topics that a news site usually covers.
Is evergreen content eligible for Discover, or only freshly updated articles are?
Evergreen content that fits a specific information need is as vital as newsworthy content. I spotted an article from FourWeekMBA.com (Gennaro’s blog on business administration and management) that was published was first published in 2017 under the entity “business”.
Does a page need to rank high on Google Search to be featured in Discover?
Quite interestingly, on a news website where I analyzed the GSC data, only 13.5% of the pages featured in Discover had received traffic on Google Search. Pages that received traffic on both channels had a position on Google Search <=8.
How can I measure the impact of Discover from Google Analytics?
A simple way is to download the .csv file containing all the pages listed in the Discover report in GSC and create an advanced filter in Google Analytics under Behaviour > Site Content > All pages with the following combination of parameters:
Discover is another important step in the evolution of search engines in answer and discovery machines that help us sift in today’s content multiverse.
Google’s decision to shut down the Structured Data Testing Tool to enhance Rich Results Test usage has raised an engaging debate in the world of SEO. Is this good news for experts around the world or it’s time to look for better alternatives? We might have the answer to this question.
On July 7, 2020, Google announced the upcoming shutdown of the Structured Data Testing Tool, an instrument widely used to date by SEO experts to verify the correct implementation of the structured data on a webpage. The decision is closely linked to the announcement of the release from the beta version of a new, more effective, testing tool, Rich Results Test. As Google explains:
“Rich results are experiences on Google Search that go beyond the standard blue link. They’re powered by structured data and can include carousels, images, or other non-textual elements. Over the last couple of years, we’ve developed the Rich Results Test to help you test your structured data and preview your rich results.”
To announce the transition from one tool to another, Google has also added a new message to the closing tool.
As you can read in the official documentation, the new tool brings more advantages to the analysis and more targeted advice on improving structured data, including:
Showing which Search feature enhancements are valid for the markup you are providing
Handling dynamically loaded structured data markup more effectively
Rendering both mobile and desktop versions of a result
It is fully aligned with Search Console reports
The tool can be used to test both code snippets and web page URLs and provides users with errors and warnings. The errors prevent a page from being displayed with the multimedia results in SERP, while the warnings indicate that one or more elements concerned will not be shown in the rich results. For example, as specified in the documentation, if there was a warning for a missing image property, that page could still appear as a rich result, just without an image.
As announced by Google earlier this month, Rich Results Test is finally out of beta and fully supports all Google Search rich result features. The tool was born in 2017 as a solution to test rich snippets, rich cards, and all other multimedia search results. When it was launched, however, it only supported four types of structured data: recipes, job listings, films, and courses. It has now been updated and finally supports all types of structured data that can be seen in SERP on Google.
Rich Results Test: is it the best solution to test structured data? What are the limits of the tool?
Rich Results Test is ready to replace the old Structured Data Testing Tool. Is it good news? For now only in part, as the international SEO consultant Aleyda Solis points out in a tweet:
When Aleyda Solis wrote this feedback it was clear that Rich Results Test wouldn’t support all types of structured data, but only those that trigger Google Rich Results. Turns out she wasn’t the only one raising the issue and this week Google’s John Mueller said that the company heard the feedbacks and that “we are planning on expanding the Rich Results Testing Tool.”
As he explained, the original idea was to simplify the job for those who were only interested in the “types of structured data that actually have an effect in search. And that’s why we focus on the Rich Results Tool which focuses on the things that we would show as Google in the search results.” But SEO experts want it all. We’ll stay updated to learn more about future improvements to the new testing tools.
However, the launching of a new tool is always an exciting time to discover new features and understand how they can help us improve our content to win the front row seats on Google Search, especially if we are talking about Rich Results. But it also opens up an important question: what if there are other, better, tools out there? Let’s take off the tooth right away and find out the best alternatives outside the Googleplex.
Top Structured Data Testing Tools
First of all: do you really need to use a structured data testing tool? Absolutely yes. These testing tools are extremely useful as they give you a lot of important information on the deployment of the structured data in your web pages, providing insights about how the search engines read these data and if they are eligible for Rich Results. Of course, each testing tool is different and can help you improve your structured data through several features. Let’s take a look at the most interesting structured data testing tool out there!
SEO SiteCheckup is a website analysis tool that contains more than one tool, including the “dear old” Structured Data Testing Tool, in one-window service. All you need to do is paste the URL of the site and click Checkup to validate the structured data, check the schema usage, monitor your website SEO, and display any issues that need to be fixed such as page load speed, URL redirects, and mobile responsiveness.
If you think you’ll miss the Structured Data Testing Tool, Yandex Structured Data Validator is a suitable alternative as is very similar to the Google tool. Along with check the markup on your site, this site helps you monitor how the structured data is processed and “seen” by search engines and whether the crawlers will be able to extract the information present in the structured data.
RDF Translator is a multi-format conversion tool for structured markup. The main value of this tool is that, unlike most other free tools out there, it supports data formats such as XML, N3, and N-Triples. Along with the use of RDF Translator to validate your structured data, you can also incorporate the tool on your website, as it comes with REST API for developers.
It comes by itself that JSON-LD Playground is the best tool for validating JSON-LD structured data format. The use is quite simple: you just have to enter the markup code with <script type=’application/ld+json‘> or the URL of the remote document and wait to get a detailed report.
Bing Markup Validator is a part of the Bing Webmaster Tools that also includes SEO Analyzer and Keyword Research Tool. This tool is particularly useful to verify your webpages markup and get an on-demand report that helps you validate different types of structured data such as HTML Microdata, RDFa, JSON-LD, OpenGraph, and Schema.org.
Structured Data Linter is a pretty minimalistic tool that helps you verify the structured data present in your web pages by simply pasting the URL of a page or a code or by just uploading a file. It supports RDFa and JSON-LD but at the moment does not support microformats.
We’ve seen the best alternative to Google’s Rich Results Testing Tool, but what about data quality monitoring?
Ok, at this point you have an overview of the new Rich Results Tool and of the most suitable alternatives out there that will help you check the markup in your web pages. But is that the best you can get? Our answer is simply: no.
As avid structured data users ourselves, having developed a powerful AI SEO tool that relies on data quality in order to enhance the content of a website and make sure that connects in the right way with search engines, we decided to build our own testing and monitoring tool.
Yes, you heard right! We think to know exactly what you need not only to validate structured data and find any error but also to do it in a smart, time-saving way. How? Take a look at the most relevant features of our tool:
UPTIME. Test your structured data availability automatically worldwide
VALIDATION. Ensure that data is always valid. We alert when something breaks, or if Google’s rules have changed
ALERTING. Get alerted by WordLift when errors or warnings are found
GUIDES. Learn how to improve your website rich result’s performance
Testing is crucial, but what about monitoring? Our new WordLift tool not only gives you an exhaustive report to constantly keep control of your quality data but also alerts you when you need to intervene, making your job easy and your markup secure.
Uh, didn’t I tell you? You can also take advantage of our dedicated technical support!
In several cases you might need to mix structured data using different formats like microdata and json-ld; in this article we review the do’s and don’ts for these edge cases.
Can I mix microdata and json-ld?
Yes,it is totally fine to use both syntaxes side by side on the same page but Google will not be able to merge attributes for the same entity using the item ID unless you are using json-ld ONLY.
Let’s get into the details:
I can have on the same page both syntaxes (microdata and json-ld); for instance I might use microdata to render WebPage and use json-ld for Organization;
I can also merge attributes related to the same entity when all the data is available in json-ld but …
I cannot combine information related to the same entity by item ID when this information is written in microdata and json-ld. While this is possible in principle, and a pure RDF application would be able to do it, Google does not support it, which means properties won’t be merged and, most importantly, this won’t satisfy the Rich Snippets‘ requirements.
This topic is particularly relevant as microdata remains today the most widely used format for structured data (see data below collected by Aaron Bradley from the 2019 Common Crawl’s sample) and there is a huge demand to improve structured data to gain additional visibility on Google’s SERP.
Before engaging with the community we created two examples HTML pages:
json-ld + microdata: here is the result validated with the Google Structured Data Testing Tool (where you will see the “Unspecified Type” error since GSDTT cannot merge the two syntaxes);
json-ld + json-ld: here we can see that GSDTT supports the merge by type ID when data is written in json-ld
Interesting enough the first example would be properly rendered by the Structured Data Linter: a tool designed to help webmaster validate structured data markup. Here follows the information from the Twitter thread and the messages by Dan Brickley and Jarno van Driel:
in general you can use both syntaxes side by side, but you won’t get the fine-grained merging of triples by ID that a pure RDF application might expect
The way we communicate and interact online is constantly changing. Users have come to expect a much more personal and tailored experience, the type that can’t be provided using traditional ways of interaction.
When looking at the words conversational marketing, some people might be wondering what exactly that is. Well, it basically is a strategy that gives customers the personalized value they are looking for and allows businesses to scale while saving time and resources. We found out that through conversational marketing and therefore through live chats, chatbots, and social monitoring it’s possible to promote genuine conversations and real relationships. The goal here is, of course, to enhance the user’s experience while minimizing friction.
Long gone are the days when consumers were passive recipients of marketing messages who had to be bombarded with a blatantly pushy sales pitch in order to be convinced to make a purchase. New, interactive technologies enabled them to break the fourth wall and have their say about how they feel about brands and what they expect from them. This means that the time has come for brands to learn how to listen actively while their customers do the talking. Marketing is a two-way street, and that’s the essence of conversational marketing.
What’s Conversational Marketing?
Unlike traditional marketing which heavily relied on TV commercials, billboards, newspaper ads, direct mail, and similar methods which customers learned to ignore successfully, conversational marketing enables brands to have relevant, meaningful, one-on-one conversations with their audiences across different channels of communication.
Live chat and chatbots are the first things that come to mind when it comes to conversational marketing. However, this strategy is much more than these two tools, and it can be extended to social media, phone calls, SMS, and IMs – pretty much any channel that your customers prefer.
Some of the benefits of such an approach include:
Being available 24/7. This is something that your customers will appreciate as you’re putting their needs first, and override your regular working hours which are somewhat limiting. AI-powered bots can answer customers’ questions in real time, be it 7 a.m or midnight. No wonder that by 2020, more than 85% of customer support interactions will be handled by chatbots.
Getting to know your audience on a more profound level. These chats and conversations are a gold mine of customer information, and they can help you understand your audience better and start using their language in your messaging.
Humanizing your brand. By combining live chat, bots, and social media, your outreach will be much more natural, and you’ll avoid using generic request forms which your customers don’t consider particularly promising in terms of providing them with timely responses.
1. Sephora’s Virtual Artist
The upscale beauty retailer stepped up its marketing game by introducing the Sephora Virtual Artist feature in their Facebook Messenger bot.
This innovative AR functionality allows the brand’s customers to “try on” makeup by uploading their selfies and applying different lipstick shades, eyeshadows, and false lashes.
Besides being fun and making it easy for customers to share their makeover photos with friends in order to get valuable feedback or add them to Facebook Stories, Visual Artist offers something much more important – a try-before-you-buy experience without having to visit a physical store.
What’s even better, once a prospective shopper makes their purchasing decision, they can order the products they want directly from the thread, which additionally streamlines and improves the customer journey. The brand reports that Sephora Assistant, a similar Messenger bot for booking makeovers in one of its stores, accounts for an 11% conversion rate increase.
eBay’s Google Assistant App tremendously facilitates browsing through the company’s vast online shopping inventory and lets customers start their search by saying “Ok, Google, ask eBay to find me…”, and this smart app will ask you additional questions in an attempt to narrow down your search and provide you with the most relevant results. Once it finds the best deal, the chatbot will ask you whether to send the results to your smartphone so that you can complete your purchase.
Given that Siri, Alexa, Amazon Echo, and other voice-based assistants are increasingly popular, it’s clear that implementing such a tool can significantly boost customer engagement.
This widget comes after the online retailer’s Facebook Messenger ShopBot, which uses AI and Machine Learning in order to personalize the shopping experience based on a deeper understanding of customer intent.
Planning and executing such an effective conversational marketing strategy can be a complex endeavor, which is why it’s a good idea to consult experts from digital marketing agencies and see what the best approach will be for your company and how to make it work within your budget.
3. Domino’s AnyWare
Domino’s wants to make the process of ordering pizza as easy as pie.
Back in 2015, the company encouraged its customers to tweet or text a pizza emoji and have a pizza sent their way.
This concept evolved further, so that now with Domino’s AnyWare it’s possible to order your favorite items from their menu through a number of available options – Google Home, Alexa, Slack, Facebook Messenger, Twitter, or even a Smart TV. This versatility and abundance of different channels of communications is something that’s of vital importance to today’s picky customers, and Domino’s does everything g to meet its patrons’ preferences.
Again, personalization and an in-depth understanding of customers needs is exactly what helps Domino’s build loyalty thus making sure that its clients will come back knowing that they can easily reorder their favorite item from the menu with a single click, tweet, or word, as well as track their order and see when it will be delivered.
4. General Motors and Social Media
Although conversational marketing is mostly related to innovative chatbots powered by the latest tech, social media is another tool that can make this strategy work.
One of the best examples of this approach is General Motors and the way it dealt with the 2014 ignition switch recalls, a crisis which threatened to ruin not only the company’s finances but also its reputation.
Over the course of several months, more than 30 million cars worldwide were recalled, while the switch ignition flaw resulted in the deaths of 124 people. G.M. was transparent about the issue and owned it, raising the bar on customer support and experience along the way.
Customers flooded the company’s social media channels with distressed comments and negative feedback, and the auto giant had its customer support reps address each and every individual complaint and offered to help on the spot.
Some customers got loaner cars until their problems were solved while others were given a refund for the travel expenses caused by the malfunction of their vehicles. Instead of trying to hush things up and switching to traditional tactics such as emails, phone calls, and other more private communication channels General Motors chose to listen to their customers, hear their objections, and proactively handle this huge blunder in the public eye.
It’s time to jump on the conversational marketing bandwagon, if you already haven’t, and take a cue from these companies who mastered the art of customer experience and satisfaction with the help of this powerful strategy.
Nina is a technical researcher & writer at DesignRush, a B2B marketplace connecting brands with agencies. She loves to share her experiences and meaningful content that educates and inspires people. Her main interests are web design and marketing. In her free time, when she's away from the computer, she likes to do yoga and ride a bike. You can also find her on Twitter.
Conversational AI at the center of your content strategy!
2019 will be the year of voice search. This article explores how conversational AI and voice queries are driving the top SEO trends for this year.
What are the top trends for SEO in 2019?
The trends for SEO in 2019, one way or another, are linked with the growth of conversational AI, voice queries and the emergence ofknowledge graphs. Here are the top 5 trends you need to watch in 2019:
We’re already living in a world where 13% of Google searches are voice queries (here is a good article to read by Rebecca Sentance on what the future of voice search holds for us). The very enthusiastic supporters of the voice search bandwagon might even say that by 2020 one every two searches will come from voice but predictions, as we know, are very hard to make especially in the marketing world.
What we have seen for sure in 2018 is that voice search is shaping the entire SEO industry. As SEO Expert Aleyda Solis pointed out recently, voice search is driving “a bigger shift”, from specific “results” to “answers” that become part of a continuous “conversational search journey”.
Artificial Intelligence, though, is the real enabler behind this transition. It was back in December 2016 when, Andrew NG Chief Scientist at Baidu at the time, predicted that speech-recognition accuracy going from 95% to 99%, would have moved the needle in terms of mass adoption of voice interfaces.
As speech-recognition accuracy goes from 95% to 99%, we’ll go from barely using it to using all the time! https://t.co/TfjqJLDTPJ
As of today machine learning and semantic networks are being constantly used to provide a personalised user experience across a multitude of channels, to guide the user across different tasks (from driving directions, to cooking, from podcast discovery to news reading and “listening”) and to help us find what matters the most.
In 2018 we witnessed knowledge graphs entering the Gartner hype cycle as emergent technology forming the bridge between humans, knowledge and conversational AI. In 2018 not only researchers and universities, but also industry companies have been heavily investing in knowledge graphs and not only Amazon with its product graph, Facebook with the Graph API, IBM with Watson, Microsoft with Satori and Google with its own Knowledge Graph but also Internet startups like Airbnb, Zalando and many others have committed resources for the creation of functional knowledge graphs meant to support general-purpose reasoning, inference and above all an improved search experience.
In conclusion, we don’t expect the entire world to shift to voice but we can predict that conversational AI and a combination of voice and touch interactions will drive SEO in 2019.
In 2019 SEOs and marketers have to prepare for a more human-driven and conversational web. Voice search will not disrupt every business but it is a driving force in the entire content marketing sector. The 2019 trends for SEO, one way or another, are linked with the growth of conversational AI and the emergence of knowledge graphs.
1. Conversational user intents and long-tail keywords
Focusing on the user intent is going to be strategic. We expect that the user search intents will be likely expressed in a more advanced range of sophisticated conversational queries. Focusing on long-tail keywords that target a specific user (in a specific context) will be simpler (and wiser in most cases) than going after a broad keyword.
When and if, we decide to go broad and to target a more general intent (ie. “business model“) we shall provide enough structure (and data) to help users (and machines) find the winning answer by further refining the initial request as if they were having a dialogue with the website (“Are you interested in the business model of Apple or of a startup?“).
Imagine preparing content as if the user would be always asking their questions to a Google Home or another voice-first device. We need to prepare all the possible answers that a conversation around that topic might trigger. We need to guide the user from the initial request to a deeper discovery of the available content and we need to match the format the user is looking for (some might like to activate a video, while others might prefer a long-format article).
2. Featured snippets and answer boxes
As a result of the information overload, and due to the growth of queries carried out via smart speakers (statistics talk about 26.4 million daily voice queries) machines will constantly need to sum up a vast amount of information, find what is really relevant for us and provide a decent speakable version of our content. While parsing language remains one of the grand challenges of artificial intelligence good results are today being achieved by Google Featured Snippets, Bing Expanded Answer and the alike. Visibility across the SERP and throughout the AI-first user experience is very much dependent on answer boxes and featured snippets. Once again, this is currently an aspect of SEO that involves mobile as well as the desktop but it is fueled by the growth of conversational AI. We shall prepare content that can be easily summarised and read aloud; we also need to leverage on structured data to help machines disambiguate the context and to support the meaningful summarization of it.
3. Structured data and knowledge graphs
Linked data, knowledge graphs and schema markup helps us connect content with a specific search intent using usage patterns that are now embedded in the search experience. A vast variety of Google SERP features are already dependent on structured data and more will come in 2019. Here below a quick overview highlighting in blue the SERP features that are linked with the use of structured data.
Google Search Features (the original list is from BrightEdge) in blue items that have some connection with schema markup.
Structured data is foundational in the conversational era as it provides well-defined information for a wide range of encoded user intents. Machine learning needs to be trained across a vast amount of semantically relevant datasets. This is true for commercial search engines, for smart assistants and for our own internal user experience – once again our website needs to become capable of answering to specific intents by guiding the user where it matters the most.
Knowledge Graph Entities are being used by the Google Assistant for answering simple questions.
4. Google News optimization and content discovery
Pre-emptive knowledge delivery and content discovery have been a trend for a few years now. This basically means helping users discover content in a serendipitous way and without searching for. In September 2018 Google introduced the Google feed “to surface relevant content […], even when you’re not searching for”.
Being able to predict the information need in a queryless way is a major focus in Google’s future of Search and it will be strategic for all major consumer brands (Amazon, Facebook, Microsoft and Apple). If we focus on Google alone we can see that, content being proposed in Google Discover is made of three types: youtube videos and fresh visual content, evergreen content like recipes and news articles. As we have seen in the checklist to optimize for Google Top Stories being present in Google News has become an important asset since the explosion of the fake news scandal (Google News content represent a selection of somehow validated and authoritative content that Google – and other players – can trust). While not all websites are eligible for Google News, if you are producing fresh content for your industry this is definitely a time to consider Google News as a new distribution channel. This becomes even more strategic these days since Google announced an upcoming voice-driven version of Google News for the Google Assistant.
5. Technical SEO for less complicated and faster websites
In 20 years we failed in building a simpler Web. Websites are becoming more and more complicated with an endless number of (sometimes useless) routines that run every time a browser hits the first page. I am not going to get into the topic but there is a brilliant article that you should read to learn about the importance of simplicity and why (the article is titled The Bullshit Web and is by Nick Heer a front-end developer from Canada) we now have to invest on technical SEO. Once again the key aspect in technical SEO – in relation with the conversation AI – is speed.
As highlighted in a study done this year by Backlinko – content that is brought into the Google Assistant is only the content that renders super fast.
A Voice Search study by Backlinko on the importance of page speed.
Page Speed has effectively become a mobile ranking factor as announced by Google this last July and will continue to impact on SERP features (featured snippets, top stories etc.) and voice search in 2019. Another major aspect that we expect to keep on driving organic growth is the support for Accelerated Mobile Pages.
Search in 2019, besides these key 5 trends, will be once again about building relationships, providing valuable answers to people needs and keeping the audience at the very center of the content strategy.
By creating exceptional content and using artificial intelligence tools for SEO you will keep your website ahead of the curve!
Still have a question? Want to go in-depth with more insights and tips? Book a call with us and get ready to dominate SEO in 2019!
Bing is starting to provide, across the world, a brand new Intelligent Search features for its SERP, powered by AI, to provide immediate answers with a new and comprehensive look and feel. In this article we’re presenting few tests of the new search capabilities and some guidelines on how to improve the visibility of your content on Bing.
Bing, with the help of machine reading-comprehension and deep neural networks, is aggregating facts from well-known data sources to provide end-users with enough confidence on the information being displayed on its search results.
Let’s start with an example based on a super simple Ego Search about myself. This panel has been around for sometimes in the US but it is now richer and it can be accessed also (when the language is set to English) from other countries.
In this specific example, the data is sourced from LinkedIn, Crunchbase and the good old Freebase. Now these are exactly the same webpages (and the datasets in the case of Freebase) that I reference in the entity about myself that is used to annotate articles on this blog. This is why, I assume that Bing, is using structured data to detect relevant data sources around entities.
Here is how Bing can help you boost your personal branding
Here is how Bing can help your customers find out more about your company
Below the knowledge panel that Bing has created around the entity WordLift.
Bing is using a machine reading comprehension technology, backed by what they call Project Brainwave, to generate the equivalent of Google’s Featured Snippets by analyzing billions of web pages to provide users with the answer they are looking for.
Let’s try with a couple of queries to see what Bing knows about WordLift. Let’s ask in the first place – “What is WordLift?”
Instant Answer on Bing for “What is WordLift?”
…and then let’s get even more specific with a query like “What is an entity in WordLift?” As you can see, results – on these very narrow queries – are indeed very impressive!
Instant Answer on Bing for “What is an entity in WordLift?“
More helpful in understanding facts about the world
Bing is also providing more ways to read facts about the world. We saw in December last year the new Perspectives Answer Box around highly debated topics like Coffee as well the new Question & Answer panel.
Now, as announced by Bing a few days ago, answer boxes also feature a descriptive tooltip for complex terms that appear in the text of the answer. Have a look at the example below where the term “Liter” is explained when highlighting the word “microliter”.
Intelligent image search
A lot has been done also to improve the image search of Bing that now uses a built-in object detection algorithm or let the user pick up a specific detail of the image with a manual crop. This really makes images way more interactive. See below an example of a photo where Bing is highlighting the two subjects.
The automatic object detection of Bing for images
2018 is definitely the year when publishing data becomes a business imperative as search engines become truly capable of providing direct answers rather than a set of web results. In this context, Bing is bringing significant innovations in the search industry by leveraging on machine reading-comprehension and deep neural networks.
Needless to say we’re particularly keen on following how search engines are starting to use artificial intelligence and how semantic rich structured data help them improve their services and in return, can help publishers improve their online visibility.
How to optimize your content for Bing’s Intelligent Search Features
Now let’s have a look at what we learned from these experiments to help you get the best out of Bing’s latest update.
1. Start using Bing Webmaster tool
It has been significantly improved and there is a lot that you can do to ensure proper indexation from Bing, to measure search traffic and even to improve the user experience on your website by using Bing-powered interactive widgets. Bing, with these widgets, works quite similarly like WordLift. It uses NLP to analyze the content of your webpage and adds an interactive widget using data from its graph. It’s a ground breaking feature and I’ll get deeper on it in the next blog post. For the time being you can preview it by visiting this example webpage.
2. Curate your entities
In the web of data, information is scattered across multiple websites and it can be analyzed and reconciled to provide a more comprehensive overview of a person, a company or a product. By curating your profile on LinkedIn, for instance, or on trusted websites like Crunchbase, GitHub or Stackoverflow you are actually publishing relevant data that Bing can effectively re-use in its knowledge panel.
3. Use structured data to help Bing reconcile content with data
As seen in these initial experiments we’re conquering a significant estate on personal keywords, branded keywords as well as questions related to our product. As algorithms start to analyze more in details the content that you have published using linked metadata and the schema.org vocabulary you can help search engines properly disambiguate these entities to find relevant information across multiple data sources and websites. By publishing articles under my name with a direct reference – in the metadata of these articles – to my LinkedIn profile I am helping Bing (and the other semantic search engines) reconcile and connect the content I write with the data that describes me.