Schema markup for voice search
How an SEO agency (and their clients!) can succeed in voice search…
In our recent article Is Voice Search the Future? we explored the rapid growth of voice search. There are a wide variety of predictions out on the Internet about how this growth will continue, with a much-cited prediction by ComScore suggesting that by 2020, 50% of all searches will be voice searches.
But there are some key enablers that need to be firmly in place for this predicted growth in voice search to become a reality. As well as the growing range of smart devices and improved voice recognition interfaces, websites also need to be prepared for voice search so that accurate search results can be returned.
One of the ways that you can ensure your website is voice search ready is to begin to implement schema markup for voice. This article takes a look at what this means and how to do it.
What is schema markup?
Schema markup – or schema – is a type of structured data markup to include in your webpages. It is a vocabulary of tags (or microdata) that you can add to your HTML. Doing this will provide search engines with a fuller explanation of what your website is about. It will also improve the way your page displays in SERPs by enhancing the rich snippets that are displayed beneath the page title.
So where did schema come from?
Schema is designed to be a universal “search engine language” and provides a collection of shared vocabularies that can be understood by the major search engines. It was jointly developed by Google, Microsoft, Yahoo!, and Yandex. The vocabulary is evolving all the time and vocabularies are developed by an open community process, using the [email protected] mailing list and through GitHub.
What kind of items does it apply to?
Schema can be used to mark up all kinds of items including people, products, events and recipes. A full list of items you can mark up with Schema is available here. Each item has an agreed set of subproperties that will be recognised by all search engines and which provide a fuller explanation of that item.
How can I learn how to add schema markup?
One good way to start is to use Google’s structured data markup helper, pictured below. Simply enter either the URL of an existing page or the raw page HTML, select Start Tagging, highlight parts of the page with important information, then Create HTML to generate the page code – which can then either replace or be included in your existing page.
So what are the implications for voice search?
Schema is particularly valuable for voice search because the structuring of your data strengthens the direct answers that are required for voice search results. Schema also increases the chance of your website appearing as a rich result – for example a featured snippet or quick answer – which in turn are often used as the sources for answers in voice search.
In light of this, Google has recently announced the development of a new Schema markup called speakable. They are still classing it as Pending, but at the same time are trialling it via a beta program with news publishers.
Speakable provides the ability to directly markup sections of content that can be spoken aloud via Google Assistant on devices such as Google Home that use text-to-speech (TTS) playback.
This is a significant step towards full audio browsing. It means that webpages with speakable structured data will be able to make use of the Google Assistant to distribute their content through new channels and reach a wider base of users. This could be a golden opportunity for marketers and SEO agencies.
Speakable markup is currently available for English language users in the US, but Google has said it hopes to launch the capability in other languages and countries “as soon as a sufficient number of publishers have implemented speakable.”
So how do we use speakable?
Speakable markup is currently only functional if you are a registered news publisher using US English in the USA. However, there are examples to see at https://pending.schema.org/speakable if you want to try and get ahead of the game.
The current schema is marked as “TO DO” for the microdata and RDFa, but is available as JSON-LD.
Below is an example of speakable markup using JSON-LD. The section “TYPES…JSON:” shows the text that would display but the remaining sections of code demonstrate the implementation of speakable.
You can see more details of the above example here.
Guidelines for effective use of speakable
Even though it is early days for speakable markup, it is good to plan how to use it in the most effective way. Getting into good habits right from the start will reap dividends in the future. Some guidelines to bear in mind are:
- Make sure that the content to which you add speakable structured data is suitable for voice-only environments, so does not include content such as timelines, photo captions and source attributions.
- Where possible, highlight a brief overview of the content- an executive summary – rather than try to highlight too much and risk getting curtailed.
- Aim for around 20-30 seconds of content per section of speakable structured data, approximately 2-3 sentences.
- Ensure that content has clear headlines that lead users through the information.
- Break content into short, clear sentences as this will read more clearly for TTS.
Speakable markup opens up a landscape of new opportunities. It has wide-ranging implications.
The reach of Google Assistant is immense. According to Google, in 2017 it was already “available on more than 400 million devices, including speakers like Google Home, Android phones and tablets, iPhones, headphones, TVs, watches and more.”
This means more than 400 million devices with the potential to make use of speakable structured data.
As we have said, it is early days for speakable schema and we will return to this topic as it develops further. But now is the time to ensure that it is firmly embedded in your digital marketing strategy so that when it is more fully available you will be ready to take advantage of its enormous potential.