Tuesday, April 25, 2017

Actions for Google Home: Time for brands to get creative

Google’s Home device was launched in November 2016 in the US, and as recently as April 6 2017 in the UK.

As a direct rival to Amazon’s Echo in the battle to gain control of the intelligent digital assistant market, Home has made great strides already. Some sources estimate that Google may already have an installed base one-third the size of Amazon’s Echo, which launched in late 2014.

Ultimately, the more effective and useful hardware will gain the public’s vote. What makes the hardware useful will be the software that powers it – and more specifically, the functionality that it provides.

Google has increased the number of Actions available via Home, and third parties are encouraged to get involved and develop novel uses for Google’s voice-enabled assistant.

It feels as though we are at something of an inflection point for this technology.

As such, it seems timely to take stock of where are, showcase some innovative uses of Actions, and also look at how marketers can start to profit from this largely untapped opportunity.

Google ‘Actions’ = Amazon ‘Skills’

Google Home is powered by Google Assistant, which has recently been rolled out across all Android devices. Assistant responds to voice commands, and can perform an increasing number of actions.

Actions are Google’s equivalent of Amazon’s ‘skills’ on Alexa; the full list of Actions can be accessed and enabled from the Google Home app.

Amazon has undoubtedly stolen a march in this regard, with over 10,000 skills already available. Most observers estimate there to be between 100 and 130 Actions available on Home.

A further 20 Actions were added last week by Google – but we are really just starting to scratch the surface of what this technology can achieve.

Google has opened this up to third-parties and has also provided a comprehensive guide to help developers get up and running.

The aim here is to move from a fairly one-dimensional interaction where a user voices a command and Google’s Assistant responds, to a fluid and ongoing conversation. The more interactions a user has with a digital assistant, the more intelligent the latter will become.

Actions: The fun and the functional

We can broadly separate the list of actions into two categories: the fun and the functional.

Some of the more frivolous features of digital assistants do serve to humanize them somewhat, but their use rarely extends beyond the gimmick phase. Just say “Ok Google, let’s play a game”, and the assistant will tell a joke, make animal noises, or speculate on what lies in your future.

On the side of the functional is an integration with If This Then That, which opens up a potentially limitless list of possibilities.

If This Then That integrates with over 100 web services, so there is plenty of room for experimentation here.

There are also a number of integrations with Google products like Chromecast and YouTube, along with third-party tie-ins with Spotify and Uber, for example.

One new – and innovative – use of Google Actions was released by Airbnb last week. The Airbnb Concierge Action serves as an information repository that is unique to each property.

The host can leave tips or prompts with the Assistant, which will then be relaid on to the guest when the correct voice command is made. Guests can also leave recommendations on local restaurants, for example, for the benefit of future visitors.

Marketers should pay attention to this. This is a clear example of a brand understanding that a new medium brings with it new possibilities.

Simply transposing an already existing product onto this new medium would be significantly less effective; we need to view digital assistants through an entirely different lens if we are to avail of their potential.

We have also seen a novel – if slightly mischievous – use (or abuse, depending on your perspective) of Google Home by Burger King this month. Burger King used a television ad slot to interact with Home and ask about one of its burgers, triggering the digital assistant to list the ingredients in a Whopper.

Although Google have moved swiftly to prevent this happening again, brands are clearly seeing Home as an opportunity to experiment and generate some publicity.

Digital assistants provide fertile ground for brands, as they create a new platform to connect with existing or potential customers. Moreover, with only 100 or so Actions available, there is ample room to engage with this now before the market inevitably becomes saturated.

For marketers interested in playing nicely with Google on this, you can sign up here to be informed of any partnership opportunities.

Monetizing voice-enabled assistants

This task is rather straightforward for Amazon, in the short term at least. Users can interact with Alexa to purchase from a selection of millions of items and have them delivered to their door by Amazon.

For Google, it is more complex. Their money-spinning AdWords business has depended on text-based search and a visual response. That input-output relationship is thrown off entirely by a voice-enabled digital assistant.

However, the smart money is on Google to find a way to integrate paid placements into their Home product, even if it takes some trial and error to find a solution that does not diminish the user experience.

During Alphabet’s (Google’s parent company) fourth-quarter earnings call in 2016, Google CEO Sundar Pichai informed investors, “[Home] is the core area where we’ve invested in for the very long term.”

The significance of those words cannot be understated. Google is, like any privately-held company, under pressure from its shareholders to deliver ever greater profits.

Selling hardware alone is unlikely to bring the profits Google needs to keep growing from its already dominant position, so there are clearly plans to monetize their Assistant in an ongoing capacity.

That level of fierce competition will bring advantages for consumers, as the products will improve and prices may even drop.

The advantages for marketers are potentially even greater, should they be willing to take some risks and work to get the most out of this still nascent technology.

Monday, April 24, 2017

Top Tips on Voice Search: Artificial Intelligence, Location and SEO

By 2020 it is projected there will be nearly 21 billion internet-connected devices, or “things” in the world.

The explosive ubiquity of this mobile-connected technology has led people to depend on these devices more regularly, with 94 percent of smartphone users claiming that they carry their phones with them frequently and 82 percent reporting that they never, or rarely, turn their phones off.

These numbers fall in line with a trend that is longer-standing, with Morgan Stanley reporting as early as 2011 that 91 percent of mobile users have some kind of mobile device within arm’s reach 100 percent of the time.

Corresponding with this increase in mobile device usage is the rise of what is called “voice search,” as well as the increasing prevalence of devices that contain “personal assistant” software like Alexa and Siri. People have become increasingly accustomed to the idea of speaking directly with computer devices and accessing information on the internet wherever and whenever they might need it. Naturally, like mobile usage in general, these emergent technologies have begun to influence search, and the impact will likely become even more apparent as usage grows.

Much in the way mobile devices have disrupted search by bringing on-the-go, local queries and results into the equation, voice search is introducing new methods of query and different results-experiences for users. Now, when a person activates voice search, particularly on personal assistant devices, most personal assistant technology will only deliver what is considered the best answer, essentially reducing the SERP to one result. That means that brands either occupy the first position, or, as far as voice search is concerned, they do not receive any attention at all.

Of course, the single-result SERP isn’t uniformly true for voice search. For voice-activated technologies connected to visual displays like smartphones and laptops, there is a greater possibility for more results. Even so, brands still need to remain focused on appearing in the top results. When someone uses voice search because they are on-the-go or they need an immediate answer, they don’t intend to scroll through pages. Rather, they’re looking for Google rich answers, such as a Quick Answer (which provides a high-quality, immediate answer to a query), Rich Card (information-rich content previews), or other top-featured results.

Google’s new Rich Cards

Over the past few years, we have seen the transformative impact of mobile on search and consumer behavior, including the shift towards the mobile-first algorithm. Voice search is the next major trend that brands will need to focus on to ensure they remain competitive. The more we understand about voice search and personal assistant devices, the easier it will be to optimize for them and ensure that your brand is represented across devices.

The role of personal assistants

As devices with artificially intelligent personal assistance software have become increasingly mainstream, so too has the use of voice search.

According to Google’s Gary Illyes, the number of voice queries in 2015 doubled from the number in 2014. Developers are now beginning to understand there are particular types of search queries people are more fond of using voice for, rather than text. For example “when is my meeting?” Users are 30 times more likely to use voice for these types of queries, rather than text.

These personal assistants, which have been put forth by several different brands, have empowered customers to remain even more connected to the internet at all times, even when engaging in hands-on activities like cooking or driving. Customers can ask about the cook time for chicken, for example, while in the middle of preparing the meat without having to remove themselves from their original task.

Mary Meeker’s Internet Trends Report looked at the reasons why customers use voice search, as well as which device settings are the most popular. The report indicated that the usefulness of voice search when a user’s hands or vision were otherwise occupied was the top reason that people enjoyed the technology, followed by a desire for faster results and difficulty typing on certain devices.

Where do users access voice search? It turns out that, more often than not, consumers are opting to use voice-activated devices is at home, followed by the car and on-the-go.

These personal assistants, along with voice search in general, are creating an increasingly connected world where customers expect search to be ever-present and capable of addressing their needs immediately.

 

How Artificial Intelligence powers voice search

Artificial intelligence powers personal assistance capabilities for mobile users. AI helps voice search and the associated algorithms to better understand and account for user intent. This intelligence, using semantics, search history, user proclivities and other factors, is able to process and understand the likely context of queries and provide results accordingly.

Natural language triggers, such as “who,” “what,” “where,” “when,” “why,” and “how,” for example, make it easier for AI to understand the user’s place on the customer journey and the likely goal of the search. Voice-activated devices can then direct users to where they most likely want to be on the web.

AI is essentially able to sift through voice search queries and identify the most important information, as well as the understand the intent regardless of an array of speech errors. For example, a query that changes direction mid-sentence, such as “How was the… what was the score to the White Sox game last night?” will be correctly answered. This enhances the conversational capabilities of the voice search, understanding the reason behind a query even if it is not asked in a precise way.

Voice search in practice

Voice search makes it even easier for customers to ask hyperlocal queries, which is significant in the context of a mobile-rich environment. Consider how users execute search queries differently when speaking to mobile devices rather than exploring the web via a desktop computer.

Voice searches tend to contain slightly different words, such as “close” or “nearby”, which are not commonly used on desktop computers. Why? Because people tend to use mobile devices to access personal assistance software, and mobile devices are most often employed to find businesses or other locations while on-the-go. The aforementioned language triggers, “who,” “what,” “when,” “where,” and “why,” are also common, setting the context for the query and what the user likely wants to find.

These queries are also most likely to contain longtail keywords, conversational phrasing, and complete sentences. All of these factors impact how brands should optimize their content to maximize its appearance in voice search.

Voice searches have also become increasingly complex. For example, users might ask, “Find a French restaurant near me” and then follow up with, “Call the first one.” The voice search algorithm is able to interpret the second query as related to the first and act appropriately. The ability of the voice search algorithm to understand the related context of these queries enhances user experiences and maintains the conversational tone.

Voice search and local search: How the SEO marketer can succeed

Knowing that voice search is an emergent technology that will impact marketing at large is one thing. Understanding how to take advantage of that fact is another. For that reason, marketers should develop an array of best practices to ensure success in the wake of this incoming trend. Here are some tips to get you started:

Tip 1. Use keyword and intent analysis to better understand the context of the queries. For marketers to be able to accurately create and optimize content for voice search, they need to know the replies that users expect when they make a particular voice search query. Then, tailor the content to meet the needs of the users. Remember to consider synonyms and alternate means of phrasing the same query, such as “How do I get to the store?” versus “Give me directions to the store.”

Tip 2. Incorporate important location keywords into the content that could impact voice search. For example, Fisherman’s Wharf, Pier 39, or Golden Gate Park might all be landmarks that people use to find a suitable restaurant in San Francisco. Incorporating these terms into your content will boost your hyperlocal presence and make it easier for you to rank for voice search.

Tip 3. Use markup to ensure that your content is ready to be displayed by Google rich results. Rich answer boxes, such as Google Quick Answers and the Local Three Pack, play a big role in providing rapid answers to user queries on-the-go. Making sure that all your content is marked up with schema will help ensure that your content is prepared to be displayed in any rich boxes that become available.

Tip 4. Make sure that each physical business location has its own site and that each site is individually optimized. This means you need to do more than just translate keywords to other languages or optimize all sites for the same terms. You need to optimize each site for the context and desires of their specific targeted audience. Learn what interests customers in that particular area through targeted keyword and intent research and make sure that each site is ready to compete within its own local sector.

Tip 5. Since a large part of succeeding with voice search is having a strong local presence, paid search and organic search teams can work together to maximize the brand’s presence. Research valuable keywords for the organization, intent, and how the brand ranks. Identifying the opportunities where having a paid ad would be the most beneficial and where organic search will be able to establish the brand can help organizations maximize their resources.

Tip 6. Do not neglect your apps. Remember that apps dominate a significant portion of the mobile experience. In fact, an estimated 90 percent of mobile minutes are spent on apps. Your data from your research about local search and natural language voice search will help you construct your app to maximize the user experience. Use deep linking within your app to ensure that customers who engage with you through voice search are able to find the content that originally interested them.

Source: Smart Insights

Voice search continues to become a dominant force in the world of digital marketing. Businesses need to be prepared to respond and keep their brands recognizable as people become more accustomed to immediate answers wherever they might be.

Friday, April 21, 2017

Four most interesting search marketing news stories of the week

We’re back with our weekly round-up of the most interesting search marketing news stories from around the web.

I hope you all enjoyed last Friday’s Easter search trivia quiz, and if you haven’t had a chance to test your knowledge yet, be sure to have a go and share your score with us on social media!

This week: a look at the newly-relaunched Google Earth and what it could mean for marketers, and a study has shown that 45% of marketers say their biggest difficulty with Schema.org markup is proving its value.

Plus, Google’s new “suggested clip” feature in search results shows how far its ability to search within videos has improved, and a new menu of Partner-only Features on Google’s Developer Blog hints at some exciting things to come.

Relaunched Google Earth introduces 3D local maps, visual storytelling opportunities

Google has just unveiled a stunning relaunch of Google Earth, with a wealth of new features and information to explore. On Search Engine Watch this week, Clark Boyd gave us a tour of the new Earth, including a look at how marketers can take advantage of the visual storytelling opportunities it presents, and what it means for local search, where “near me” searches will activate a 3D local map featuring business names, photographs and contact details.

45% of marketers have difficulty showing the value of Schema markup

A recent survey carried out by Schema App, a provider of tools to help marketers use Schema markup, has provided some insight into the difficulties that marketers encounter when using Schema markup.

Schema markup is often touted as a killer search tactic which is nevertheless seeing very little uptake among website owners. It can vastly improve the look of websites on the SERP with the addition of rich data, and it is integral to a number of Google features like featured snippets.

But according to Schema App’s survey, 45% of marketers say they have difficulty in “showing the value of doing Schema markup – reporting the impact and results”. Forty-two percent struggle with maintaining the ‘health’ of their markup when Google makes changes, while 40% cited difficulties in developing a strategy around what to mark up with Schema.

Meanwhile, nearly a quarter of respondents (24%) said they had difficulty understanding Schema markup vocabulary at all.

Google shows “suggested clip” feature in search results

Google is continually improving its ability to search within a video, and to surface a particular search result within the content of a video. In a previous search news roundup we reported on the fact that Google’s machine learning technology can now recognize objects within videos, as demonstrated at Google’s Next Cloud conference in early March.

Then this week, Ryan Rodden of Witblade reported that Google is now showing suggested video clips in search results for particular queries:

Image: Witblade

The suggested clip appeared in a query for “blur out text in imovie”, highlighting a suggested clip of 25 seconds in the middle of a how-to video. While it’s unknown how accurate this result was for the query, it shows that Google is making bold inroads into searching within video and is treating video like other kinds of content to be crawled, indexed and presented as a Featured Snippet.

Given the huge rise, and popularity, of video of all forms in marketing, social media and publishing at the moment, it’s a smart move and something we can probably expect to see more of in future.

Google adds extensive new menu of Partner-only Features

Google’s Partner-only Features are a forum for it to debut certain search features to a select group of approved and certified providers, before they are rolled out on a wider scale. Aaron Bradley noted in the Semantic Search Marketing Google+ group this week that Google has just added a huge new menu in the Partner-only Features section of its documentation.

The new menu features eight sub-sections including “Carousels”, “Indexing API”, “Jobs” and “Live coverage”.

All of the links currently lead to a 404 error, but it could be an interesting insight into what’s to come from Google.

Thursday, April 20, 2017

How to use Google’s new demographic targeting for search ads

Through AdWords, Google has given advertisers a lot of control over when their ads are shown, by means of the different match types and using remarketing lists for search ads.

Until recently, however, you were unable to target users based on demographic – a function that has been available for a while now on both Facebook and Bing.

The new feature allows advertisers using Adwords to target users based on:

  • Age
  • Gender
  • Parental status

This feature will be particularly useful where user intent varies considerably based on these variables. For example if you were selling high-end investments or watches, it is unlikely that young people under the age of 25 would have the necessary capital to purchase them.

However when using this feature, it is important to make sure that your conclusions are based on data as opposed to your gut feelings. A study by Google has shown that some of our preconceived ideas about which demographics purchase which items may result in us missing out on a considerable proportion of buyers.

Image: Google

For example if you were running a campaign selling home improvement products and excluded women on mobile devices, you could lose 45% of your traffic.

One thing to bear in mind is that your customer might not always be your customer. For instance, the study by Google showed that 40% of baby products are purchased by households that do not contain parents.

Here you can see that a considerable proportion of some markets are not the consumers themselves, but people purchasing on behalf of consumers.

How to set up demographic targeting in AdWords

The demographic targeting options can be found within the audiences tab alongside your remarketing lists for search ads (RLSA) data. To add bid modifiers take the following steps:

STEP 1. Go to the “audiences” tab and then to the “demographics” sub-tab as shown below.

STEP 2. You can switch between demographic data for “age” and “gender” using the two sub-tabs that are located under the graph.

STEP 3. Bid modifiers can be set within the “bid adjustment” column by clicking on the dashed line.

Once you have done this you should see a popup like the one below where you can enter your bid modifier.

STEP 4. To calculate your bid modifier you should use the following formula: divide the age conversion rate by the ad group conversion rate, subtract one, and multiply by 100.

So for example if the conversion rate for people aged 25 – 34 is 3.52% and your conversion rate for the ad group overall is 2.76%, then your bid modifier would be 28%. Note that you need to round up your modifier to the nearest whole number.

When you are faced with “Unknown” data where Google is unable to match the user to their data, you will in most cases not want to exclude this audience.

In some cases we have found that Google can’t match data to a large chunk of your traffic, which can be frustrating, but if you exclude this you are likely to miss out on a considerable portion of your traffic.

Conclusion

Overall, demographic targeting for the search network gives advertisers another dimension with which to narrow down their audience to target the most relevant people.

Google’s example of baby products being bought by households that do not contain any parents is a perfect example of why it is necessary to follow the data as opposed to your gut feeling when using this feature. Otherwise you run the risk of losing a considerable portion of your audience.

Finally, when you are faced with the dreaded unknown column, think twice before excluding this data. In the vast majority of cases this will account for a considerable chunk of your traffic so it is best not to exclude it.

What’s new with Earth? First impressions of the relaunched Google Earth

Google has just re-designed, revamped and re-launched its Earth product, and it has certainly been worth the two-year wait.

Earth is now built into Chrome, so there is no longer a need to download a cumbersome desktop app to access this global repository of images, videos, and knowledge cards.

The Android app has been updated too, with support to follow soon for mobile browsers and iOS.

So what’s new?

A lot.

First impressions of Earth are simple: this is a hugely impressive feat, one that truly celebrates the world – both natural and man-made – by capturing its farthest corners in finite detail.

So, let’s get started. We begin with a zoomed-out view of the planet, before a short introduction from Google on some of Earth’s upgraded features.

These features work in a cumulative fashion, each adding to the last and building up to three-dimensional, customizable, multimedia experience of our planet.

First up, the search function. The foundation of any great Google product, this deceptively simple search bar leads to any location in the world:

This is given extra potency when combined with Google’s vast inventory of knowledge cards about cities, rivers, buildings, and basically just about any landmark you can think of.

These are typically pulled from Wikipedia and appear as a clickable carousel, although other resources are cited on infrequent occasions.

It is possible to zoom in to the level of Google Street View to get a closer look at the palace in this screenshot, as has been available via Earth and Maps for some time now. This is labeled the ‘Photo Sphere’.

Added to this is the “I’m Feeling Lucky” feature, which takes the user to a random point on the map and works like the button of the same name in traditional Google search.

My first trial of “I’m Feeling Lucky” took me from Lagos to Legoland in just one click. It can be quite a dizzying trip, depending on your screen size and propensity for motion sickness, but the speed of flight can be adjusted in the settings menu.

Layer by layer, this builds up to Voyager, the section most likely to keep users engaged with Google Earth.

Voyager contains a wealth of curated content from sources as diverse as Sesame Street and the BBC, but we should expect many more publisher partnerships in future.

This is significant, as it takes Google into the realm of visual storytelling and opens up a host of new opportunities for publishers willing and able to get on board.

There is already a good variety of content on here, including city guides, nature trails, and the work of specific architects like Frank Gehry. That said, this is an inexhaustible resource that will play host to a lot more experimentation soon.

One highlight is the ‘Revealing the Center of Life’ tour, which takes us on a journey underwater to explore coral reefs.

As an educational center, this offers unparalleled scope for exploration and will undoubtedly spark much healthy discussion. Some of these knowledge cards are accompanied by videos and behind-the-scenes features too, providing further context to the images.

The implications for marketers

Brands should really be thinking about how to avail of the storytelling possibilities that this brings. For travel and tourism companies the opportunity is perhaps a little more obvious than for other industries, but in truth there is an opening here for almost everyone.

The Argentinian artist Federico Winer has partnered with Google to create a photographic series on airports, for example.

There is also a history tour that traces the steps of characters in the novels of Charles Dickens, and another that visits some of Ernest Hemingway’s favorite haunts.

With consumer attention spans at an all-time low, Google Earth should now be viewed as an incredibly powerful, engaging tool, should publishers have the imagination to avail of its potential.

In perhaps more prosaic terms, local search remains just as vital as it has been for some time now – perhaps even more so.

Typical searches like the one below for [book store near me] will bring up an interactive 3D map of the local area with some options, so it is vital to have business names, addresses, opening hours, photos, and phone numbers up to date.

Customizing Google Earth

And it doesn’t stop there. Users can import KML (Keyhole Markup Language) files to overlay images and charts onto Google Earth. Google even provides an example of this in action, with an overlaid image of Mount Etna erupting.

KML is based on the XML standard and provides a few extra functionalities, like paths and polygons, that are particularly useful for Google Earth.

Google provides a sample file and comprehensive guide to get started, although this should be pretty familiar to anyone accustomed to creating custom Google Maps.

In summary

The new Google Earth is more than the sum of its features; at its best, it can both distort and inform our perception of space and time.

A historical echo of this project would be the eighteenth-century Encyclopédie, a Herculean effort by Denis Diderot, Jean le Rond d’Alembert, Voltaire, and many others, to catalogue and categorize all human knowledge.

Combine that persistent thirst for knowledge with the technology at Google’s disposal and the product is something as engrossing and enlightening as the new Google Earth.

It seems fitting to give the final word to Google: