Search engines, especially Google, are getting much more technologically advanced, with the slightly ironic goal of more closely approximating human thought and speech patterns. We predict that 2018 will find SEO experts making the most of new technologies to help “teach” search engines about how humans think. Natural language is the lingua franca of SEO, helped along by voice search’s increasing presence. As of mid-2016, 20% of all searches on Android were made by voice; also in mid-2016, ComScore estimated that by 2020, 50% of all searches will be voice searches. In December 2016, Gartner predicted that by 2020, 30% of all search queries will be made without a screen, through a voice-first device. Because voice search queries are questions in everyday language, search engines have learned how to parse everyday language for searcher intent. Google reported in May 2017 that its voice recognition software can now understand human speech with 95% accuracy—which is also the human threshold for speech recognition. In August, Microsoft reported that Bing’s voice recognition error rate was 5.1%, which puts it almost even with Google’s. The result is a continuation of the years-long SEO shift from bot-oriented optimization to optimizing for humans. Rather than tweaking content to match keywords, content teams should use more natural language (even conversational where appropriate) aligned with searcher intent rather than bot-based best practices. This contributes to a more customer-oriented marketing experience, which will soon be the basis for customer expectations. Position zero matters more than ever, as it’s really the only position in search on voice-first devices, such as Alexa, Echo, Google Home, and the new Apple Homepod. Voice-first searches tend to return only one result, and whenever possible, they prioritize the content of the device’s provider over all other content. Even in searches on non-voice-first devices, the search engine will prioritize its content over the content of others. Fortunately, most SEO tools have become more sophisticated to keep pace with the new reality of search. Many can show daily data regarding which sites currently rank for answer boxes and other universal results, allowing marketers to see trends in what search engines deem worthy of position zero and form strategies to capture that position. As search becomes more personalized, rankings become more difficult to discern, since every search engine results page (SERP) is now a custom SERP. A high empirical ranking for a keyword can be reduced or even negated by a searcher’s location, browser history, and previous search data. In fact, the concept of an empirically high rating is starting to dissolve, as there’s no clear indication of which SERP among the many could be considered “true” or “accurate.” Keyword rankings have been a cornerstone metric for SEO since its inception, but with so many personalization factors at play, the SEO industry will likely need to adapt measurement methods for keyword rankings, possibly using AI and machine learning to assess a visitor’s likely search history and location. Over time, the metrics themselves will change, as new KPIs for visibility evolve. In the meantime, SEO teams can take steps to work with these aspects of personalized search:
- Location – make sure your local SEO is on point, even if you don’t interact with customers face to face
- Browser history – keep publishing interesting, rewarding content so visitors come back often, thus increasing your site’s presence in their browser history
- Semantic connections – you can’t guarantee that your visitors will often search for terms similar to yours, but you can maximize your technical SEO, meta content, and markup to help search engines understand what your pages are offering and tie that back to search queries
- If your site gathers signups for an email newsletter, you’ll likely fall under the jurisdiction of the GDPR. Marketing best practices already dictate that signup forms explicitly request consent from the user and clearly state what the user’s data will be used for; now is the perfect time to ensure that all your site’s signups meet those criteria.
- If you use a third-party analytics platform (other than Google Analytics), be sure it complies with GDPR rules regarding personal and sensitive data. Just about every professional-quality analytics package will have GDPR compliance well in place by May, so a simple check should be all that’s required.
- If you haven’t had your site optimized or at least checked for mobile speed in the past year, doing it now will put you ahead of the mobile-first rollout, so you won’t have to make up for lost rankings. The most popular resource for checking load times on desktop and mobile is Google PageSpeed
- Make sure the same content is rendered on both your mobile and desktop sites. That doesn’t mean all the content has to be visible on mobile, but Google has to know that the content is there. So if you choose to hide some content for page speed or UX purposes, make sure your team is using SEO features to ensure Google doesn’t discount the content or its keywords
- If you publish a lot of content or get most of your traffic from mobile, consider accelerated mobile pages (AMPs). Google hosts these pages on its AMP Cache, so it can pre-load the pages it thinks will be most relevant for the user’s possible searches; it also hosts them so publisher sites won’t have access to user data before the user even makes a query. Later this year, Google will be adjusting their displayed AMP URL to be the URL of the content publisher’s site, rather than google.com/amp/[original URL]