I use no-nonsense and established methodologies that I have tested for YEARS to improve the bottom line of my clients
The concept behind paid search advertising is elegantly simple: you bid on keywords that describe your product or service and when potential customers search for something that triggers your keywords they see your ads.
But what keywords should you choose to bid on?
The success or failure of your Google Ads campaign seems to hinge on this point, so selecting keywords can feel a bit like picking lottery numbers. What’s more, if you’re picking your keywords off of “gut feeling” alone, then your odds of success are actually in the lottery-winning range.
Keyword research is the art of systematically and scientifically selecting terms which are most likely to drive customers to your site. This article will walk you through this process from start to finish. Whether you’re a first time Google Ads user or an experienced marketer, these concepts are the key to building strong campaigns.
The first question you need to ask yourself is “Who will be searching for my product/service?” The better you know your target audience, the better you’ll be able to predict how they will look for your product or service on Google.
Chances are that you’ve long ago identified your target market. Identifying your online searchers is a related, but slightly different challenge.
For example, let’s say that you offer consulting services for city planning. Your target market is government officials in small, growing towns in the Midwestern US, which do not have a full-time city planner.
But when the town needs to expand, whose job will it be to find a city planning consultant? A councilman? The chief of police? The mayor or maybe even his personal assistant?
Each of these groups will approach the task of finding your company in a slightly different way. As such, they make up distinct audiences that you will want to consider when choosing keywords, designing ads, and constructing landing pages.
Once you have a list of potential audiences, you need to know how they think, especially about your product/service.
From your existing bank of customers, select individuals from each of your target audiences, and interview them. Ask questions which will help you understand how and why they found you, such as the following:
Often, individuals and businesses search online because they have a problem (called a “pain point”). The rural mayor who is unsure of the best way to expand his city may choose put his pain point right in the search bar: “Need help understanding zoning laws.”
If you know how your customers express their problems, you are one step closer to being there when others have similar difficulties. And what better way to find out how your customers talk about their problems than to ask them?
This question invites customers to tell you what problem you solved for them, but it may also open up a discussion about features unique to your product/service which your audience values. These can also make great keywords. Sometimes aspects of your company which you find mundane will be surprisingly meaningful to your customers.
In addition to these questions, if you are already getting customers through online marketing you can ask even more direct questions, such as “What keywords did you use to find us?,” “Why did you click on our ad?,” and “Why did you convert/buy?”
As you interview customers from each of your audiences, you’ll start getting ideas about what they find important, and how they express it. This should give you have enough information to start building a keyword list…but how do you turn abstract ideas into keywords?
The answer to this is fairly simple, take the ideas from your customer interviews and start brainstorming!
I’m serious. It may not be the most scientific approach, but it’s a good way to generate a lot of initial ideas. We’ll get around to refining your list later, but you can’t refine what doesn’t exist.
Consider different ways of expressing your selling points, derivatives of terms your customers used, and audience- or location-specific modifiers that you could ad. For example, “Upper Amazon tree frogs” could be rewritten as “Maranon river tree frogs,” “Upper Amazon Herpetology,” or “Tree frogs of Peru.”
Try not to get too picky yet. You’re trying to create at this point, not evaluate. Excessive deliberation can kill a brainstorming session almost before it begins.
Before we get too deep into creating your keyword list, we need to talk about keyword length. Keywords can be long or short (e.g., “How to hike Mount Godwin-Austen on a starving student budget” vs. “K2”), but not all keyword lengths perform equally.
A large study published by Search Engine Watch illustrates this point. This study analyzed 1.5 million keywords that companies actively bid on and measured the activity that their ads received. I’ve compiled some of their data into a graph to make things easier to follow:
The curve seen in this graph is often compared to the shape of a Chinese dragon, with a short head and body, and a long tail.
As you can see, short “head” keywords (6-15 characters) get a lot of ad impressions (views), clicks and conversions.
This makes sense. After all, a short keyword will of necessity be pretty vague, and match a lot of people’s searches. If you bid on the term “Amazon Rainforest” (15 characters), you’ll get hits from conservationists, tourists, kids writing reports and who knows who else!
There are two troubles with “head” keywords though: 1) They’re very competitive, so clicks on these keywords are expensive and 2) they attract a lot of audiences you’re not targeting.
If you’re selling a book on tree frogs, most of the people who use search for “Amazon Rainforest” won’t be potential customers. They might click your ad and cost you a lot of money, but they may not be likely to make a purchase.
For this reason, it’s important to take into account not just how much search volume a keyword gets, but how effective it is at turning searchers into customers.
Here’s how keyword length and effectiveness relate, again drawn from the Search Engine Watch study:
Ad click-through rate (CTR) and conversion rate (CR) tend to be low for “head” keywords, but increase in the body and tail.
This trend (the longer the keyword the more effective it is) makes sense. The more specific “body”-length keyword “Herpetology of the upper Amazon” is likely to attract only those interested in reptiles and amphibians. An even more specific “long tail” keyword like “Upper Amazon tree frog Dendropsophus bifurcus” will draw an even more specific crowd. If somebody searches for this, they want what you’re selling!
But how many people actually search that term?
Clearly, there’s a balance between getting a lot of ad views and getting views that count. With that in mind, let’s look at the two graphs side-by-side:
The sweet spot, where search volume and converting power are high, tends to be in the 16-30 character range (the body and upper tail).
So, while you’re building your keyword list, look for keywords in the 16-30 character range. If you must vary from this rule of thumb, err on the side of the long tail. After all, Google Ads is pay-per-click advertising, so if nobody searches for your term, at least you’re not out any money.
Once you’ve got a working list of keywords, it’s a good time to stop and take a look at your competitors. This will help you to check whether you’re on track with your ideas, and can help you come up with new ideas as well.
First, take some of your keywords for a test drive. Open up your Google browser and type one of your terms in just as if you were your own potential customer. Press Enter. One of two things will happen:
If #1 happens, you may actually want to breathe a sigh of relief. It means that other companies find your keyword valuable enough to include it in their campaigns. As long as your competitor isn’t Wal-Mart, you can probably bid your keywords up to a competitive level and start getting some of the traffic that your competition enjoys.
In addition, if you’re just starting into Google Ads and your competitors are old pros, you may be able to learn a lot by observing what kinds of ads and landing pages they’ve built around the keywords. That being said, just because your competitors have more experience with Google Ads, doesn’t mean they’re doing it right…
On the other hand, if you put in a keyword and your competitors don’t show up among the results, it could mean a few things:
It may not be apparent which of these options is the case, so don’t throw out a keyword just because there’s no competition for it. Some of the later steps in our research process will point out any obvious flaws in your term.
Although you can learn a lot through the method outlined above, if you really want to get the inside scoop on your competition, consider obtaining a competitive analysis tool (SpyFu, iSpionage, etc.). Tools like these can tell you not just whether the competition bid on the same word you thought of, but all of the keywords your competition is bidding on!
Tools like SpyFu or iSpionage can be a treasure trove of quality keyword ideas, and can even form the backbone of a strategy to capitalize on the most productive keywords from your competitors (quite legally and ethically, too).
At this point, you’re probably getting a feel for how keywords work, but your list of potential terms and phrases may still be pretty short. After all, brainstorming sessions tend to generate a lot of ideas around a few trains of thought, but neglect others that didn’t make it into the creative “flow.”
It’s time to flesh out your list. There are a lot of resources you can use to do this, which means you can find a lot of potential terms pretty quickly. Some will be useful…and some won’t. So, whenever you see a potential word or phrase, run it through the following three filters:
If a new word meets these requirements, go ahead and add it to your list (we’ll get more picky in the next step).
The following are just a few of the readily-available resources to help you generate new keyword ideas. Experiment with a few and you’ll quickly start to find your own favorites.
Once you have an Google Ads account, you’ll get access to the Keyword Planner.
The Keyword Planner is kind of a one-stop-shop for campaign research. You can upload the keywords you are considering individually or in groups. The Keyword Planner will remix terms from your list into new keywords, suggest entirely new terms related to your current ones, give you feedback on search volume and even predict how your keywords are likely to perform.
The Search Console gives you access to an analytical report of how people are already interacting with your website.
Among the many features of this report is the option to see what queries people are using to find your site. Chances are if people searched with a term and clicked on your site without an ad, then the term would draw even more traffic with an ad.
Quora is a crowdsourcing website where people can have their questions answered by other Quora members (kind of like Yahoo Answers).
One nice feature about Quora, however, is that you can search by topic. Simply type a subject of interest into the search bar and Quora will display the top questions that people have asked related to your theme.
Results like these can be a nice window into what people interested in your product/service are thinking and asking about.
While Wikipedia articles may not be a great source for a college paper, Wikipedia articles are curated by a variety of people with unique perspectives on a single subject. As such, they can provide a fairly comprehensive look at the different facets of a topic.
This can be a great source of keyword ideas. For example, type “City planning” into the Wikipedia search bar, and you get redirected an article with 16 more potential keywords in the first paragraph!
When two things are correlated, it means that they change similarly over time. When one goes up, so does the other. Google Correlate allows you to input a search term and see which other search terms have experienced similar changes in popularity.
Often, keywords are related in this way because they express the same idea in different terms. Take “comic con,” for example. Feed that phrase to Google Correlate and it comes up with the following list of correlated terms:
Who knew that people were misspelling “comic con” as “comic com,” “coma con,” “com con” and “comicon?” Keywords like this could be interesting to try.
For example, Snickers bid on 25,000 common word misspellings in a truly brilliant paid search campaign that drove 500,000 impressions in just 3 days.
So, just because a word is misspelled, doesn’t mean that it can’t be a useful keyword.
Google Correlate also gives you an idea of how many searches keywords get over time. The graph above shows a steady increase in “comic con” searches, with a spike every July (Google Trends is another great resource for this kind of information).
When using this tool, it’s important to keep in mind that just because two things are correlated, doesn’t mean that one causes the other, or that the same people are initiating both searches. A lot of people search for “fly fishing” every summer as well as “comic con”, but that doesn’t mean they want to go to comic con, too.
You may never have heard of Google Autocomplete (AKA Google Suggest), but I bet you’ve used it before! When you start typing in the Google search bar, a Google Autocompleter sees what you’re typing and types in what they think you’re probably searching for!
Okay, so Google Autocomplete is actually an algorithm that tries to predict what you are searching for, but the end result is the same. For example, here’s what you get if you type in “paid search”:
Google Autocomplete is meant to save you time and to help you find what you want more easily on Google, but it can also help you think of new keywords. As part of its predictive model, Google’s algorithm takes into account typical searches that begin with a given term. By putting some of your keywords in, you can see other words that searchers may be using with your starting words.
When you started the last step I suggested you might be feeling like you didn’t have enough keywords. Now you might feel like you have too many!
You’re not alone. Most campaigns actually launch loaded with more keywords than they need, and this can set things up for a lot of wasted money.
To avoid this problem, you’ll want to sort that massive keyword list into the good, the bad, and the ugly. Let’s take a look at several criteria that indicate whether or not a keyword should be tried.or trashed.
There are a lot of reasons why people search online, but the intent behind their search typically falls into one of two broad categories: Commercial intent or information intent.
Commercial intent means that the searcher wants to buy something. For example, keywords like “buy,” “discount,” “shipping,” and “coupon” usually indicate that a person wants to buy something soon. As you might imagine, keywords like this are very useful for advertisers.
Other terms, like “comparison,” “varieties,” “prices,” “features” and “reviews” indicate that a person is considering a purchase, but is still scoping out their options. These don’t convert quite as well as the terms above, but they still show intent to buy.
Although the commercial intent keywords we’ve just discussed are fairly universal, many products, services and industries have their own specific commercial intent keywords. If a keyword or keyword phrase on your list includes one of those words, it’s definitely worth considering.
On the other side of the spectrum are searches with information intent. For example, if you search “How old was George Washington when he died?” then you’re searching for information. You don’t want to click on an ad and you don’t want to buy anything. You just want an answer to your question.
Obvious fact-finding search terms (“when did,” “how old is,” “who was,” etc.) are generally unwise to include in a PPC campaign, no matter how many searches they get. After all, why should spend your money to get visitors who have no intention of giving you a return on your investment?
Of course, keywords won’t always be this cut and dry. For example, if somebody searches “loyal dog breeds” then they may be searching for information to help them with a purchase in the far future. This kind of search is the bread and butter of content marketing and SEO…but not necessarily PPC.
As a general rule of thumb, it’s safest to put your money into keywords with some degree of commercial intent. If your keyword list is largely geared toward information you may want to rethink it.
Intent is important, but it isn’t everything. “Buy [insert your product here] now” would be a keyword with fantastic commercial intent…but what difference would it make if nobody searched for it?
In general, the more qualified impressions an ad gets, the greater chance it has to attract converting customers. If a keyword won’t actually produce a substantial number of ad views, it may just be a waste of space in your campaign.
So, you’ll want to think twice before bidding on “Pokemon Go.” Sorry. That ship has sailed.
Finally, you’ll want to make that you haven’t gotten so carried away with finding new keywords that you’ve lost touch with step 1 of the research process: Define your target audiences. The keywords in your final list should all be specific and relevant to at least one audience subgroup.
Imagine if you had chosen to bid on the word “buy.” Great commercial intent! Incredible search volume! But it’s so vague that you’d never make a cent off it. If some of your keywords are too short or nonspecific, consider combining them with others into a more precise term, or dropping them altogether (once again, that keyword length rule of thumb proves useful, eh?).
If you followed all the above steps—identifying your audiences, brainstorming words, examining the competition, fleshing out and refining your list—then pat yourself on the back! You’re researching keywords like a pro.
Take your list and put together your first Google Ads campaign. If you can build ads and landing pages with the same attention to audience that you gave your keywords, then you’re well on your way to PPC success!