‘If we’re going to help algorithms understand our content, it needs more context,’ says Adam Edwards, Reprise Digital’s Head of SEO.
With an estimated 50 percent of all searches to be made via voice by 2020, marketers are rapidly prepping search strategies for the interface — but with Amazon and Google still mum about exactly how algorithms drive their respective assistants to give answers, it’s difficult to know which established SEO tactics translate.
But through trial and error over the past two years — Amazon launched the Echo in 2015 in the U.S. and 2016 in the U.K. — SEO experts have ideas. Following his panel discussion at Voice Summit entitled The New Rules For Search and discovery, GeoMarketing caught up with Adam Edwards, Head of SEO at Reprise Digital, to discuss building a voice strategy — and which brands are doing it right.
You’re an SEO expert, and the topic of incorporating or translating digital SEO strategies for use in the voice realm is a hot topic today. What can you tell us about how you are approaching the idea of SEO in voice?
One thing that I think that people forget about is that to be able to improve upon something, you need to measure it. I think a lot of people haven’t even gotten that far yet. They hear about how quickly voice [search] is growing, and it’s like, “oh we’ve got to jump in with both feet.” That idea of measurement — how these searches [are working] and how people are finding businesses — that’s something that we’ve done very well [at Reprise] within Google, to start.
But being able to answer these queries — whatever interface is involved — is really about it coming from some sort of data. You have to build a framework of knowledge, a knowledge base, that can address people’s questions and also just be available as a value to people who happen to be looking online. [What marketers can control] is making sure that knowledge base is there and that these [intelligent assistants] have the essential facts about your business.
How can businesses build that knowledge base, in your view?
Well, thankfully, we do have some visual cues. For instance, the most obvious within Google is you can see the quick answers at the top. So if you do a search, and you see that, where you see a paragraph or some bullet points or maybe even an image at the top of the search results, that’s likely to be a response to a voice query. It’s not 100 percent sure, but that’s how we start: We build out by recognizing why that is doing well — why that’s becoming a featured answer on Google — and then we can inform the rest of our strategy.
Any examples of how that works?
I’ve been doing SEO for such a long time that I can remember we used to tell our clients to have as big a website as possible. You need a page to answer every single keyword. That’s no longer the case, because Google is getting better at semantically understanding the content of a site; it can pick out key points if it’s written well and it’s concise, within paragraphs. If you structure your site well, it’s not just thinking about the user — which is of course paramount — but for search, the key content has to be as close to the homepage as possible.
There’s been a big trend in UX recently (user experience), to talk more about the broad goals that a user might have. That works well, but you also need to keep in mind that if we’re going to also help computers or algorithms understand our content, it needs more context.
Content is king, but context is queen, right?
Exactly.
The title of your panel at Voice Summit is “New Rules for Search and Discovery.” What arethe most important rules?
One thing is that I think too many content calendars — whether it’s for blogs or other aspects — are only focused on trends that they happen to hear about that are big. It’s a bit reactive. Instead, if you measure [how people are finding your site], if you can start to prove an ROI on what is possible, then you can say specifically, “we need these 15 or 30 or 200 pieces of content to actually address what people are looking for.” That’s beneficial for both voice and “traditional” search. It’s about creating new content — and some of that is voice content as well, obviously.
What brands in the space do you think are doing a good job with that?
Johnson & Johnson is one that comes to mind: They’ve tackled a lot of voice skills, and they’re going back to measure things. If you actually try voice searches and see what happens, a lot what happens on the Echo products from Amazon with the Alexa framework, for example, they will suggest Alexa skills to users because out of the box it can’t answer everything.
If you ask about allergies, for instance, it might recommend a skill from Zyrtec, which is one of Johnson & Johnson products. So, again, I do kind of think at this stage it’s about testing and seeing what happens. If you haven’t built a voice skill yet, just ask [the questions people might ask related to your brand] and see what comes back. Trial is probably one of the most effective ways of figuring out what to do.