The process of keyword research is constantly evolving and refining itself. As Googlebots and operators get smarter, and the way we interact with devices and content changes, so do the best practices for research and optimization.
While there is no shortage of content written about keyword research; from the optimal density, to utilizing LSI keywords, to strategically placing head keywords throughout the content, the strategies that worked like magic one day can easily become obsolete the next.
Here, we will cover five new factors that are often overlooked in undertaking keyword research, yet are on the cusp of radically changing the concept as a whole.
1. Voice Search
Voice search goes beyond the voice-to-text capability – something smartphones have possessed for years. Keyword research in voice works in accordance with the capabilities of the Internet of Things (IoT), which are brought to life through devices like Google Home. With voice search, users are verbally stating their queries while they hear AI-generated responses. This can make it very difficult to rank for keywords on this channel.
Things to Consider
Voice search favors featured snippets. Rolling out featured snippets was Google’s lead up to releasing the Google Home assistant, so those are the only answers Google will give as a voice search result.
Voice search keywords are long tail questions. As what should be no surprise, people talk much differently than they type. For example, where a desktop search would be phrased as “rent background check,” a voice search might be phrased as something like “why do landlords do background checks?” Let’s compare the desktop SERPs for these two queries below to see how different they are:
SERP for “rent background check”
SERP for “why do landlords do background checks?”
These SERPs offer completely different results depending on the phrasing of the keyword. Also note that in both examples, the site with the featured snippet does not hold a position one ranking.
Voice search is important for local SEO. Many people use voice search to look for local information, such as “What is the best Italian restaurant near me?” The convergence of local SEO and voice search requires a lot of behind-the-scenes Google magic. That said, there are many things you can do to help improve your chances of getting ranked highly on a local search. For one, you can include a location within the title tag, like so:
You can also do things like include a click-to-call phone number on your website, post reviews, include photos, and more.
2. Mobile-First Index
The mobile-first index is the version of Google’s index in which the mobile version of a site is considered the primary. Depending on the differences between your site’s desktop and mobile versions, this can affect rankings depending on the device from which a user enters a search query.
Things to Consider
Mobile sites must have the same content as desktop sites. Simple responsive design platforms have made this process much easier to implement. The result is that content automatically reformats depending on the device, like in this example from Rent.com:
Responsive design mostly affects menus and content layout, however, there are still some ways mobile sites need to be optimized differently.
Include shorter but equivalent meta data and subheading tags. Meta descriptions are often shorter on mobile, as you can see in the following SERPs for the search “apartment hunting tips”:
Long headings can sometimes show up poorly on mobile, so try to shorten them as much as you can while maintaining the keywords.
Track differences between mobile and desktop keywords. SERPs can sometimes be very different from desktop to mobile for the same keywords. Since half of all searches are done on mobile, desktop and mobile rankings should be considered equally important.
RankBrain is an artificial intelligence algorithm that determines a searcher’s true intent behind a query. It helps Google rank searches it has never seen before. RankBrain is a complicated and somewhat mysterious element of Google rankings, with Google itself saying very little about it.
Things to Consider
Create high quality content that uses LSI keywords. RankBrain’s effectiveness comes from its ability to determine the context of a query or keyword. This has made Latent Sematic Indexing (LSI) even more important for keyword research and content creation. Keyword stuffing has not been effective for a long time, but RankBrain truly makes it obsolete. With a focus on user intent, RankBrain evaluates content based on what will be the most helpful to the searcher, which means well-written content that incorporates the natural use of keywords and their variations.
Freshness is more important than precise keyword matching. Since RankBrain can differentiate between semantically-related search terms, trying to rank for a specific variation of a query is no longer that important or effective. Instead, capitalize on keyword search trends in your industry to create content around what is popular. Similarly, if you intend to rank for a constantly changing query, like “best movies to see this weekend,” that page must always have up-to-date information to consistently rank well.
4. Personalized Search
Since 2009, all Google’s search results have been affected by a number of personalized factors, including location, search history, and browsing history.
Things to Consider
Target local keywords. Location will always affect search rankings, even if users are searching in private or incognito mode. Incorporate your business location into your content and meta data, and filter your search volume results in AdWords by location to leverage the keywords that people in your target location are frequently searching for.
Page one rankings still matter. Search history personalization works by ranking a user’s most-visited websites higher, even if a brand new SERP would display the rankings differently. From an SEO perspective, this can be frustrating. A top result may not appear at the top to all users if they have previously preferred a competitor. But, the top spot increases your chances of becoming the preferred resource for new searchers, and can increase your rankings in later searches where a competitor may outrank you for the same query.
For instance, let’s go back to the earlier example of searching for “rent background checks.” TransUnion’s SmartMove grabbed the top spot, so if a searcher showed a preference for that result, later searches for “credit checks” or “employment background checks” could rank TransUnion’s other services higher for that particular user, even if an unaffected SERP wouldn’t rank quite so high.
Things to Consider
Use historical data to time seasonal content. Seasonal content should be rolled out when the search volumes for those queries will be highest. By looking at several years of data on Google Trends, you can determine the peak of those queries and publish as the volume is increasing, rather than decreasing. For holidays, this is fairly simple: don’t write about Christmas decorations in June. Take a look at the search trends for the term “spring cleaning tips” over the last five years:
As you might expect, searches for “spring cleaning tips” spike every year in the spring/summer months. A closer look reveals that the peak search volume is typically the third or fourth week of March, meaning your spring cleaning guide should be published before or around then, rather than the first week of April, when search volume is decreasing.
Pro Tip: Be sure to check the SERP features for your chosen term. The SERP for the term “spring cleaning tips” includes both a featured snippet and a news carousel, as you can see here:
Trending topics can affect which SERP features appear at certain times, which also affects overall rankings. In this example, Reader’s Digest holds the featured snippet, but their actual position falls below the news carousel. With this in mind, it may be more effective to attempt to appear on Google News, rather than compete with the established entries.
Stay on top of trending topics for QDF. QDF stands for “query deserves freshness,” which is an algorithm that determines when a search term requires the most recent results. When news sources, blogs, and search terms all display an increase in coverage of a particular topic, Google will rank newer and more authoritative content higher for the duration of the increased interest in that topic.
Over to You
These factors often overlap and intersect to affect keyword rankings. Optimizing keyword research and SEO content for all of these factors simultaneously is immensely difficult, but with the right tools and guides, it can be much easier.
Which of these factors do you already consider while conducting keyword research? Care to share any case studies or insights with us?