You depend on your clients for success, and they depend on you for results, but sometimes, simple misunderstandings or inefficiencies can get in the way of an otherwise perfectly functional relationship. Of particular note are “best practices” in SEO that are no longer best practices, or elements of SEO that are poorly understood. A client may demand a certain approach even if it may end up hurting their campaign, or they may blame you if they aren’t getting results in the exact form they wanted.
Few SEO areas are as hotly debated or as dynamically evolving as keywords—in research, targeting, and optimization. Keyword-based strategies still exist, and are still effective, but they’re almost unrecognizable from their older counterparts, and as a result, many clients enter into an agency agreement with unfounded expectations or preconceived notions about how keywords are supposed to function.
It’s your job as an agency to stay up-to-date on the latest best practices for keyword-based optimization, and make sure your clients understand these standards perfectly well. Otherwise, you risk damage to the relationship when they believe your side of the bargain to be unfulfilled.
First, I want to address some of the old ways that keywords permeated SEO as a sharp contrast to the way they’re used today. You may already be familiar with these standards, and you may believe some of them to still be how the search world works. If you or your clients currently still accept or operate by this old model, you’re in desperate need of an update.
Google’s search algorithm once relied on exact keywords matches to produce results. If a user searched for “chicken tacos,” for example, it would scout for all the pages on the web that featured the phrase “chicken tacos,” and then sort them based on how many times that phrase was used, along with the authority ranking of the site itself.
This led to a series of practices designed to exploit this keyword basis for search; companies would research the most competitive keywords, then stuff them into page titles, body copy, and anchor text for links in order to maximize their relevance for those terms. Analysts would then monitor ranking progress for those specific keywords, gauging campaign effectiveness on this upward trajectory.
If you’re still using these strategies today, the same way you would have a decade ago, there’s something wrong.
There are a number of interrelated factors for why keyword-based optimization is so different today than it was several years ago. Make sure your clients are aware of these factors, even if it’s only at a cursory glance, to help them understand the “modern” era of keyword optimization.
Hopefully, your clients already know what Google Panda is. At its core, Panda is a quality update, focused on content, designed to penalize sites with low-quality content in any capacity, and reward sites with higher-quality content. One of the biggest offenses Panda targeted was “keyword stuffing,” the act of deliberately placing targeted keyword phrases in the headlines and bodies of articles throughout your site in an effort to achieve higher ranks. Panda introduced natural language detection and quality evaluation into Google’s algorithm, weeding out any articles or sites that were thought to be using too many keywords in their content. Today, if you try too hard to squeeze unnatural-sounding keywords into your content, you’re going to trigger this algorithm—plus, you’ll turn your readers away.
Google Penguin, equally recognizable, offered a similar quality update for links, rather than content. The old-school keyword practice here was to build links with anchor text that included your target keywords, or at a minimum choose offsite articles with headlines that included the keyword in question. Thanks to Penguin, the “quality” of links is more easily detectable by Google, so if you try to stuff keywords here, you’ll also be penalized.
The bottom line here is that domains that try to exploit keyword stuffing are going to be penalized, no matter what.
Google Hummingbird came out back in 2013, and completely overhauled the way Google evaluates user queries. Rather than taking a look only at keyword phrases, Google introduced a semantic focus to the search engine, making it capable of evaluating and meeting user intent. Instead of mapping instances of keyword phrases to exact matches on the web, Google now dissects the intention behind a user’s query and attempts to grab results that meet that intention. In some cases, this results in radically different SERPs, but Google still relies on keyword detection to understand the subject matter of various sites and web pages. The power of keywords has been weakened, but not obliterated.
RankBrain is a machine-learning modifier that was added to Hummingbird just last year, and it’s a sign that Google’s semantic capabilities are only going to grow more sophisticated. RankBrain’s purpose is to better understand complex and ambiguous long-tail user queries, essentially boiling them down to a more manageable level for Hummingbird to take over.
There’s a great example of RankBrain’s effects floating around:
(Image Source: SearchEngineLand)
(Image Source: SearchEngineLand)
Notice how the second query is asking the same thing, but in a more complicated, convoluted way. RankBrain’s job is to take the second query and figure out that it’s just a long way of asking the first query.
Keyword optimization gets more complicated with the rise of the Knowledge Graph and local SEO, two very different concepts that share an undermining effect toward keywords.
The Knowledge Graph is the system of indexed information that Google uses to provide you answers to specifically or concisely answerable questions. As an example, when you type in “why do baking soda and vinegar react,” you’ll see a short explanation pop up above all the organic search results:
There are many forms and types of queries that allow the Knowledge Graph to kick in. It’s a great resource for users, as it saves them the step of sorting through organic search results, but it has the indirect tendency to divert organic traffic, and it increases Google’s capacity to provide direct answers to queries, rather than mapping queries to existing locations on the web.
Local search has the same “diversion” and “query answering” effects, but for a slightly different reason. Here, local searches operate using a separate algorithm, which kicks in when user location data is available (or when a user utilizes geographic-specific keywords). Older local SEO techniques, like stuffing unnatural phrases like “best plumber dallas tx” into content, are no longer valid here.
Personalization is also affecting the significance of keywords in the modern era, thanks to Google accounts, browser histories, and personal digital assistants, all of which can feed or use data on your history and geographic location to alter your personalized search results. Two users who search for an identical query—let’s call the “chicken tacos” reference back to the forefront—might get totally different results. One might get chicken tacos recipes, based on his/her strong disposition for recipes in the past, while another gets chicken taco restaurants. This makes it harder to predict what your users are actually searching for, and more difficult to guarantee any kind of visibility from a ranking increase. After all, your ranking increase will only be for a part of the audience doing the search in the first place.
After reading all this, you might start to wonder whether keywords are important at all these days, and if not, what the alternative might be.
The answer isn’t exactly straightforward. In response to Hummingbird, some optimizers have suggested that a suitably alternative for keywords is “topics,” which gives you more freedom when it comes to phrasing. The goal here is to predict types of user queries and write topics that address those queries, such as answering common user questions or proactively addressing user concerns. You’d do topic research, much like keyword research, tracking down popular topics and ones that haven’t been suitably covered by competitors, then produce high-quality content that naturally contains contextual clues that help Google categorize it and call it up for the appropriate queries.
Topic-based SEO is highly effective, and a suitable alternative to keyword-optimization in some ways. However, keywords still have a power of their own, giving you shorter, more precise phrases to work with, more trackable results, and generally higher potential volume. Even though Google doesn’t map keywords from queries to pages like it used to, it still uses keyword phrases to help it understand site pages, so they’re still a valuable strategic focus in an SEO campaign.
One of the most important parts of a keyword optimization strategy is the research. The entire point of keyword optimization is to choose the most valuable keywords to optimize for, so getting the right information (and therefore, the best list of possible targets) is essential if you want your clients to see progress. Modern keyword research is a bit trickier than it used to be, but with the right tools, the right approach, and enough communication with your clients, you’ll do fine.
(Image Source: AHrefs)
First, you need to know the difference between basic keywords (sometimes called head keywords) and long-tail keywords. There’s no exact cutoff here, but long-tail keywords are essentially the same as basic keywords, but… well, they’re longer. These are long phrases, sometimes colloquial, like “where’s the best place for chicken tacos” instead of the basic “chicken tacos.”
Generally, the longer the query becomes, the lower the volume and competition become. This makes them easier to rank for but also makes them yield a lower potential traffic rate with a high rank. Compared to head keywords, they offer fast-paced gains, but a lower long-term payoff (assuming you invest sufficiently in the basic keywords). They’re also great material for topic-based optimization.
A good, balanced strategy should have both basic keyword and long-tail keyword topics as part of your research, though depending on your approach, you may qualify your long-tail research separately, or as part of your topic research.
When you first get started generating keyword ideas, you’re going to rely on your own brainstorming (and don’t worry, we’ll dig deeper in a minute). For this, you’ll definitely want to consult with your client; they know their industry, their business, and their customers far better than you do. Together, come up with a big list of various keyword terms you think your client’s customers might search for, and try to target specific products or services if you can. This will get you started in the right direction as far as relevance is concerned.
Start compiling your keywords in a spreadsheet; we’ll be expanding on this shortly.
(Image Source: AHrefs)
After you’ve got an initial list scrapped together, you’ll have a working foundation for some heavy expansion. For this, you’ll probably need to rely on some external tools to help you get the job done. It’s almost impossible to pull all this data in by yourself.
Google’s Keyword Planner is a tool within Google AdWords designed to help advertisers plan their campaigns, but the information it offers can be used for an organic campaign as well. Plug in all the keywords you came up with (and any subsequent keywords you find with the tools below), and it will tell you the average volume and competition rating for each; this will be vital in narrowing down your list of targets.
(Image Source: Software Insider)
AuthorityLabs is a major name in the industry, because it helps you come up with new keyword ideas, measure things like search volume and competition for each, and even track your ranks as you implement them as part of your strategy. There are also a number of filters to play with to see how keyword results play out in different scenarios.
(Image Source: AuthorityLabs)
SEMRush takes things a bit further by offering a number of research tools. If you enter a specific keyword, SEMRush will help you break it down in terms of its fluctuations, current competition, and volume. You can also find related terms here, and chart differences in desktop and mobile devices.
(Image Source: SEMRush)
Google Trends is a better tool for topic research and long-tail keyword results, but it’s still useful to see how your user demographic trends change over time. Start general here, and work your way toward more niche topics; you’ll learn much about user behavior and search patterns to inform your growing keyword list.
(Image Source: Google Trends)
Plug in a competitor’s URL here, and Spyfu will tell you some of their most profitable keywords and search positions. This is useful for finding key opportunities to outrank your closest industry competitors.
There used to be several keyword idea generation tools leveraging the power of Google’s auto-suggest API. Google’s auto-suggest comes up with popular related keywords, saving you the trouble of trying to think them up on your own. However, now that Google has privatized that API, these tools don’t have quite the power they once had. Ubersuggest is among the best here.
(Image Source: Content Marketing Institute)
Ultimately, you’ll want to work with your client to narrow down your list to the top potential candidates, zeroing in on a dozen or two strong keywords with the highest potential return. You’ll want to look at three factors here:
At this point, you’ll have a solid list of target keywords with which to begin work.
Modern keyword optimization is all about balance, in more areas than one. Let your clients know that there’s no one right way to optimize for keywords, nor is blunt force ever a good strategic approach in the realm of keywords.
First, you’ll need to split your attention between optimizing for keywords and optimizing for topics. As we’ve seen, both are important if you want to host a successful strategy. However, it’s not always as simple as splitting your efforts down the middle, fifty-fifty. Instead, you’ll need to actively monitor the ebb and flow of your work, and make adjustments accordingly. Are there a lot of potential news topics to cover? Start optimizing for those topics. Is your client on the verge of a page-one ranking for a specific term? Start putting more effort into that term. And of course, if you find that your progress is slowing or that you aren’t getting the results your client needs, you can make adjustments to your strategic lineup.
You’ll want to include keywords in your blog posts, and meta data, and really, throughout your site. But thanks to Panda and Hummingbird, if you include too many, you’ll end up getting your client’s site penalized. What’s the solution? The old method was one of percentage, making sure your targeted keyword phrases don’t appear more than 2-3 percent of the time. However, a better solution is to avoid stuffing keywords at all; the less you think about it, the more naturally you’ll write, and the less you’ll have to worry about a penalty.
For starters, only choose keywords that you can work into your content naturally, and then, work them into content titles only when they’re appropriate. From there, they’ll probably appear naturally as you complete the content work. For some keywords, this is easier said than done, but your first job is choosing the right keywords to begin with.
Reporting is a big deal for agency-client relationships, and keyword ranks tend to be a sensitive issue. You’ll find your clients want high ranks, as fast as possible, and may grow irritated if they aren’t getting the ranks they want (or overly complacent if they are).
First, set the expectation that ranks aren’t everything. Yes, you have target keywords and your goal is to rank for them, but you’ll be rising in rank for dozens of long-tail keyword phrases you didn’t even know you were optimizing for (thanks to your brilliant content marketing strategy). Plus, keyword rankings can only tell you so much—what’s really important is your inbound traffic.
Second, set the expectation that ranks are volatile, and aren’t entirely predictable. Your rank may change from day to day, and may appear differently for two different people in the same room. There’s a degree of relativity to be expected in the modern realm of keyword-based optimization, so try not to let your efforts be judged too precisely.
When it comes down to it, the vast majority of issues with keyword-based optimization can be avoided with a bit of proactive communication with your client. Here are some of the most important points to touch on, early in your relationship.
If you make these points clear, and you follow the keyword strategies I’ve outlined above, you should have no trouble keeping your client happy and up-to-date with the latest best practices in keyword optimization.
It’s hard to say exactly what the future holds for keywords (or SEO in general). But I suspect that the world of keywords and topics is only going to get stranger. Technologies like Hummingbird, RankBrain, the Knowledge Graph, and digital assistants are evolving at a remarkable pace, and all of them are, in some way, making it harder to get your site ranked for a specific keyword term. Overall, keyword focus is only a part of SEO—building your authority, earning links, providing great content, and offering the best user experience are other fundamental pillars you need to worry about. So instead of trying to perfect the keyword side of things, hedge your bets, and try to develop the best overall strategy you can with your clients.