BLOG

Guide to Keyword Research

Published March 22, 2018
Language matters in marketing. For copywriters the right words make for effective ads and memorable taglines. For Search Engine Optimization (SEO) and Search Engine Marketing (SEM) specialists finding the right words is critical to successful campaigns. That’s why keyword research is a cornerstone skill set. While there’s no shortage of articles on this topic, they often focus on the primary goal of securing top rankings and driving tons of traffic to your website.
 
So, allow us a slightly different take on the subject: it's not just about rankings and traffic.
 
Yes, leads and sales require traffic, which comes from visibility. But ranking for the wrong keywords is not helpful. It’s also not useful when a site ranks for the right keyword but with the wrong page.
Keyword research, when done right, can do so much more than support rankings. It helps us better understand users, and it can help us better visualize conversion paths and user journeys. It can also help inform persona development by tying demographic and psychographic information about audience segments to specific language. In this post, we’ll cover the broader scope of the concerns, needs, and processes that all go into keyword research in modern digital marketing.

Keyword Sources

Ok, so let’s start with the basics. Where do we find keywords? There’s also no shortage of articles compiling listing various keyword tools. It’s such a tried and true evergreen topic, one could probably compile the articles themselves: “The Best Articles on the Best Keyword Research Tools”.

So while this is not an exhaustive list a few of our favorite tools include:
SEMRush
Google Keyword Planner
Moz Keyword Explorer
Ahrefs
Soovle
Ubersuggest
SerpStat
Keyword.io

Of course, if you are using Google Analytics, Search Console or AdWords to analyze keywords you will find useful information, right from your own data. Before you go looking for “new” keywords, be sure you understand where your current search engine traffic is coming from.

 

 

SEMRush Guide
The SEMRush Dashboard showing data for the keyword "keyword research"

 

 Understanding keyword metrics

Many of the keyword insight providers above will append various metrics to their keyword data. Part of keyword research is not just understanding what words search engine users enter in their queries, but some of the numbers surrounding them. This helps us assess value and competition for keywords, which helps us plan strategies and campaigns around them once the research is done.

Search volume
This figure indicates the average number of searches per month for a keyword. The exact number of searches per month will of course vary, so many data providers use 12 month rolling averages. But always be mindful of seasonality. A keyword related to winter holidays might have 480 searches per month on average, but that keyword will likely see fewer than 480 searches in the summer months. Other keyword phrases, like “oil change near me”, will typically have less variance between months. Overall “search volume” is the basic “how many people are searching for this?” metric.

Cost Per Click (CPC)
This figure indicates the average amount bid in paid search campaigns for that keyword. It’s a “how much does it cost?” metric. In SEM campaigns, this is going to inform budgets and bidding strategies, but even in organic SEO research, it’s a useful metric. All things being equal, companies will often be willing to bid more on queries where the intent suggests a greater likelihood of purchase or conversion. Low CPCs suggest that either the query is not easily monetized (“how to tie your shoes”) or that the readiness to convert is still low and overall conversion rates are low (“low price airfare”).

Rank/Position
If you are looking at organic rankings, this is typically the position number for an organic listing. It can also be ad placement in paid advertising. Although the makeup of SERPs have changed quite a bit over time an organic ranking in spots 1-10 represents Page 1 visibility on Google, Bing, or other search engines. These are typically the most clicked results. Positions 11-20, or Page 2 rankings, often have click-through-rates at or below 1%. Rankings beyond Page 2 do not tend to generate meaningful clicks on a consistent basis. However, it can be useful to track a site’s deeper rankings in Google results. This research can reveal places where a page, if improved, could potentially move to Page 2 or even Page 1 for a keyword. A site ranking on Page 3 for a phrase might not be generating traffic now, but it’s in a better position to achieve visibility than pages that don’t appear until Page 7.

Impression Share / Share of Voice
This is a competition based visibility metric. Both paid and organic search tend to function as a zero-sum game: clicks to one site (typically) cannot also go to another site. But the first step toward getting clicked on is being visible in the first place. Impression share or share of voice metrics indicate the percent or share of searches for a given keyword (or broader keyword space) for which a website is visible. This is an important metric in growing or shrinking markets. A site could technically see more organic search traffic in a growing market but less impression share. Conversely, another site could be in a shrinking market and see less search traffic over time, even while their impression share grows. Competitive share metrics help analysts better gauge performance, success and campaign trajectories.

Share of organic traffic
This metric, available from some but not all, keyword data providers estimates the percentage or share of organic traffic for a website that any one given keyword drives. This will be a function of a site’s total rankings, their ranking levels, and the estimated click-through rates for any given ranking position in a search engine. For example, a #1 ranking for a keyword with monthly search volume of 100 might drive more organic traffic (probably 25 to 40 clicks) than a number #10 ranking for a keyword with a search volume of 1,000 (probably 10 to 20 clicks).

Understanding Tails and Funnels

One of the fundamental concepts you must appreciate to research keywords is the demand tail. The demand tail is a visualization of the relationship between 2 different numbers; the total search volume of a keyword and the number of keywords out there that reach that level of monthly search volume. There are 3 sections of the demand tail, each with their own characteristics. What makes one keyword belong to one segment or another will vary by topical space and by query type, but there are some general characteristics that align to most spaces.

Short-tail keywords: There are only a few of these within any given topical space. Short-tail keywords are the phrases that have the highest search volume. The search volume for these keywords can be anywhere from a hundred to thousands or tens of thousands of searches per month. The notion of high volume is relative to each space. The short-tail searches in a niche space will have lower overall search frequency because there is a smaller set of individuals looking for these terms. Short-tail keywords also tend to have the highest competition, both in organic search results and paid campaigns. As you might expect with a designation of “short” the number of words in these searches themselves tend to be few, for example “running shoes” or “men’s running shoes”.

Mid-tail keywords: These keywords have fewer searches per month than short-tail keywords, but the monthly volumes are still significant for the space. Search volume for mid-tail can be anywhere from dozens to hundreds per month, even into the low thousands. But there are also more of these keywords within the topical space. These queries also typically have more words in them than in short-tail queries and as a result, they tend to be a little more specific: “Nike running shoes” or “Nike men’s running shoes”.

Long-tail keywords: These keywords have the lowest number of searches: probably less than 100 searches per month and maybe as few as a just a dozen or two. However, the number of unique keywords that have these lower search volumes is the highest. The number of words within the search is also the highest, and these searches tend to be longest and most specific for a space: “size 10 Nike men’s running shoes” or “size 10 men’s Nike air Jordans”.

 

Long Tail Keywords

 

These demand tail segments will very loosely align with stages of a user journey or conversion funnel, and with the depths of your site architecture. Homepages and major category pages, which are just one click away from the homepage, tend to be upper-funnel type of content geared toward short-tail searches. As the keywords move into the mid-tail, the pages that rank for them tend to be a little deeper into the site architecture, like specific subcategory pages. As the keywords move into the long-tail, the landing pages that rank for them tend to be even a little deeper into the site. These are most often secondary subcategories, product level pages or individual blog posts.

The important takeaway with demand tails and conversion funnels is that you want to have the right page for the right intent. Many websites have secured a “good” Page 1 ranking for a valuable keyword term, only to realize it is to the wrong page. For example, when a blog post outranks a sales/service page. If a user is conducting a long-tail search, the homepage is likely too broad to serve as an ideal entry point into the website.

Understanding Query Types

Beyond knowing specific keywords strings and their associated metrics, keyword research requires basic taxonomy skills: classifying information into categories based on a set of characteristics. One of the best documents anyone can read to truly understand search is the Google Search Quality Evaluator Guidelines. In that document they distinguish between three of the broadest types of search engine queries. Simply put, nearly every keyword is one of three very broad types: Go, Read, or Do.

Navigational (Go): These keywords indicate a user intent to go somewhere specific online. “Facebook login” is one of the most highly searched terms in this category. Similar examples might include “espn.com” or “Verizon billing portal”. Certainly these users could use bookmarks or enter the URL directly into their browser (if they know the URL). But for convenience, they are using a search engine to send them in the right direction.

Informational (Read): These keywords indicate the user intent to read or learn something. Almost any query that includes the traditional question words (what, who, when, where, why or how) is going to be informational. But queries do not have to include question words to qualify as informational. “Nike men’s shoes size chart” would be an example of an informational query, as would “Nike running shoe reviews”. These users haven’t definitively demonstrated intent to take a specific action, but they are not entering in navigational queries either. Most keywords are going to be informational keywords in any given topical space. Success in the informational query space is crucial for attracting qualified visitors and moving them towards the specific actions site owners depend on: adding items to carts, newsletter signups, free trial downloads, etc.

Transactional (Do): These keywords indicate a user intent to take a specific action. These could be commercial intents (buy, purchase, send, ship, etc.), but they don’t always have to involve commerce. Because they are behavior/action oriented, they do tend to represent users who are closer toward the end of a specific user journey or the bottom of the conversion funnel than users entering in more informational queries.

For any given website, or the topical space it occupies, it’s important to have a general understanding of what kinds of keywords appear from each of these three broad types, and what that means for you and your users.

 

Search Word Cloud

 

Branded vs Non-Branded Queries

When people use search engines to look for a company by name, or by variants of a name, that’s branded search. These queries can be thought of in some ways as a navigational search. Within reporting and analytics, even though it will be attributed to organic search, branded search can often be a “hidden” kind of Direct traffic. Because the name is built into the search, all branded queries come from users already familiar with the brand.

Non-branded queries, however, are ones that do not include brand references within the search phrase. Searches with no specific brand intent can be informational or transactional or navigational.

The specific ratios and balance between branded and non-branded search queries will vary from site to site. Some organizations seemingly drive a lot of organic search traffic to their website, but almost entirely for their brand name. If you look at non-branded commercial intent, however, their visibility may be much different. Conversely, some organizations rank consistently well for non-branded queries, but the number of searchers looking for them by name is quite low. The “right” ratio depends on a lot of variables: too many to identify in a single blog post. But no organization should be relying on more than 80% of their traffic coming from either branded or non-branded searches. Cross that 80% threshold and the site might be over-leveraged on that particular query type.

Beyond Exact Match: Understanding Intent and Entities

We probably shouldn’t have to say this, but you might be surprised at how keywords continue to be misused. So once again, for those in the back; exact match keyword stuffing is dead. Stop it! Yes you should use high volume, highly relevant, keywords in your title tags. But leave the exact match keyword spamming where it belongs: in the past.

Keyword research is really about cataloging the range of language used within queries relevant to the vertical, not tacking them onto a URL thoughtlessly. That means breaking the queries down into their basic parts of speech: nouns, verbs, and qualifiers (adjectives, adverbs, numerals). Every keyword you can gather data on, now and forever, will still be built from these very basic building blocks: things, actions, qualifiers. That’s it.

So, when conducting keyword research, you want to have a grasp on the numbers, the value and the opportunities. But, more importantly, you want to have a comprehensive survey of the search space you are investigating. The goal is largely to identify and list the most common nouns used in the topical space (shoes, sneakers, cleats), the most common verbs (review, compare, shop, buy, save, ship) and the most common adjectives (red, size 10, men’s, cheap, rare, best).

These descriptors (for lack of a better word) should inform future content pieces and strategies, but don’t get mired down in them either. Learn about synonymy but understand it isn’t a magic bullet. Google is getting better and better tying different descriptors to the same fundamental people, places and things, entities. This is why you can Google both “kitty cancer” and “feline cancer” and many of the top results will include the word “kitty” or “feline” in the title tag or even within the body copy. Google knows both terms really mean “cat” and that cats often get “leukemia”. The entities “cat” and “leukemia” are so well understood by Google that even seemingly different keywords get the same results.

Once you have a thorough catalog of the types of things, actions and qualifiers that people are entering into search engines, you can correlate them with relevant pages on a website. You can also begin to create new content to target new phrases. But never lose your focus on the intent behind the keywords. Search phrases are a wealth of insights into a user’s mindset and where they are in their research or shopping process. Content optimization and/or creation is what comes after the keyword research, but it all starts with a spreadsheet or two, some tools and ideally, a lot of coffee.

Brew a pot and good luck!

ALSO IN THIS BLOG

Remember how your mom told you not to stand too close to the television because it might hurt your eyes?
The same rules can apply to data. If you’re too close, you may miss the patterns and trends that are crucial to understanding your website’s performance. You can’t judge a site’s performance looking at data in the bubble of a single day, you must consider any day’s traffic compared to the days before and after.
Google Analytics makes it fairly easy to analyze trends over long periods of time. But it also allows you to stand right in front of that TV, to look at more granular levels of time, right down to the hour.
There’s a better way to get that close to the data, without burning your retinas. We’ll cover how to analyze traffic effectively in today’s post.

 

 

When the client first came to you, you talked up the value of Google Analytics. You emphasized the importance of seeing where your traffic was coming from. You went on and on about how Google Analytics can show traffic sources to pinpoint whether people came from search, social media or a specific site referral, and how valuable this data was. You sold them on it, so much so that your client looked forward to receiving that first report, the magical day when they would finally understand where visitors were coming from.
But then the report came, and it looked like this:

 

 

It showed that 10% of your client’s traffic came from “(direct)/(none)”. What does this label mean? How do you explain Direct traffic to your client? Better yet, how do you explain “none”?
Let’s take a closer look at understanding Direct traffic in Google Analytics and how we can address it with clients.
One of the most exciting and important aspects of digital marketing is the ability to understand exactly how your customers are finding you. It informs every single part of integrated campaigns and helps determine which efforts are working and which ones need to be revisited. Google Analytics allows you to zero in on the performances of different marketing channels to evaluate everything from brand awareness to social media messaging. To get the most insight from that data, it’s crucial to understand exactly how Google sorts your traffic.
Channels in Google Analytics are high-level categories indicating how people found your site. While the Source/Medium report shows you in more detail where people came from, Channels are broader, more “user-friendly” names lumping visits together in buckets useful for high-level reporting categories.
For instance, Facebook Sessions often show up in multiple ways in the Source/Medium report. They may appear as facebook.com, m.facebook.com, and l.facebook.com, all of which are variations of the same source. The Channels report will include all of these in the Social bucket, so you can see less granular, aggregate numbers on social media performance.