Ah, keyword research: few people’s favorite part of the job but one of the most critical. And if you’re doing it, you might as well be doing it right, so learn not to make the following 13 (common) keyword research mistakes:

#1: Caring more about position than SERP decorations

So much has been made of the click-through rates for positions 1 through 10 in the results. Various studies have put the boost in click-through rate at 40% or more when Google+ rel=author markup gets the author’s photo to show in the SERPs next to their result. Similar boosts are seen in click-through rates for getting different kinds of rich snippets (reviews, ratings, price, availability, etc.).

These “extras” that Google chooses to decorate a given result with scream “LOOK AT ME, LOOK AT ME” to the searcher, and like the sheep that we all are, we do notice those results more, and tend to click on them.

rel=author is fairly easy to implement, and so is schema.org.  So why stress out about getting to position #1, when you can get all that traffic from position #3 with a little Las Vegas neon, in the form of authorship or rich snippets? Google also makes it easy to learn about both.

Oh, and it’s certainly easy to test both of these with Google’s structured data tester.

#2: Missing “barnacle SEO opportunities”

Will Scott seems to have invented this term, and it’s a good one.

Google will often put organic listings from companies like Yelp, Angie’s List, etc., in the page 1 results for queries with local intent–often above the local 7-pack. And YOU, Squire, are NEVER going to outrank Yelp, no matter what you do.  But…you might be able to convince Google to pick your client’s page from Yelp as the one to show for that term. How might you do that? Why, build links to your client’s Yelp page, of course, to make it the strongest Yelp page in that business category.

#3 Failing to check plural as well as singular forms of a term

True, Google’s getting pretty good at recognizing synonyms, misspellings, plural vs. singular. But many times, the difference in results for what seem like virtually identical terms is astounding – as is the search volume. Here’s an example: travel agent vs. travel agents vs. travel agency.

Search volumes are radically different:

Search Volume Travel Agent

As are the results (with my location set to Portland, OR):

For “travel agent“:

travel agent

For “travel agents“:

travel agents

For “travel agency“:

travel agency

#4 Being too much of an insider, and using “insider” terminology

Make sure you’re using terms your customer understands.  A great example: let’s say your client is an HVAC installer.  Your client may be mildly obsessed with their rankings for “hvac installer”, “hvac contractor”, etc., because “hvac” is THE term for them that covers what they do, right? Well, while consumers probably all KNOW what “hvac” means, when they’re looking for someone like this, they’re probably either hot and wishing they had working A/C, or they’re cold and wishing they had a new furnace.

So, guess what…the search volumes for “heating contractor” plus “air conditioning contractor” total more than “hvac contractor” (and of those 880 searches/month for “hvac contractor”, I wonder what percentage of them are installers obsessing over their own rankings.) :-)

#5 Using broad match instead of exact match in AdWords tool

Most of us use the Google AdWords keyword research tool, right? That tool is REALLY designed to help you pick the right targets for your ads, which is why it defaults to Broad Match instead of Exact Match. Broad Match essentially shows you the search volumes for which Google would potentially show your ad. This is NOT the same as the number of searches for that term–it includes searches for similar terms, where Google believes the searchers’ intent is close enough to your target that it’s worth showing your ad.

Exact Match shows you a rough average of the number of searches per month for that term over the last 12 months (more details are here).

For your keyword research, you’ll always want to set this to Exact Match.

Google match type

#6 Seeing a competitor tune for/spend PPC $ on a keyword and assuming it’s a hot one

Lots of people get excited by tools like KeywordSpy and SpyFu, where you can see what keywords your competitors are bidding on. I do, too, and I use them all the time – they’re a great source of ideas for keywords.

But: do NOT presume that the people doing the keyword research for your competitors have brains bigger than dried peas. Sometimes SEM consultants will “cast a giant net” by targeting enormous volumes of keywords. Sometimes this is a good strategy and can bring in some paid traffic cheaply in a onesy-twosy fashion. And sometimes, this is a great strategy to impress a client: “we’re managing 12,500 keywords for you in Google each month, and we’ll keep doing this for you each month for the lowly sum of $xyz, unless of course you’d like to do it yourself”…ahem.

Take the keyword ideas you get from looking at competitors, and throw them all into the AdWords keyword tool, and let THAT tell you which ones are worth tuning for.

#7 Forgetting to add pws=0 to your queries when you’re looking at the SERPs

This is the magic trick to depersonalize Google search results. Kinda.

Google will favor sites you’ve visited before, even if you aren’t logged in to Google. Simply append &pws=0 to the end of the URL after you do a Google search. Note that it depersonalizes, but doesn’t delocalize!

weather serp

Note also that if you’re signed in to Google, it might still affect the results, e.g. you might see that you’d +1’d a result.

#8 Not recognizing Local Universal results in the organics

The Venice update brought many more organic-looking local results into the mix–these are now referred to as “local universal” results

Google began including organic results from the searcher’s city that otherwise wouldn’t have the strength to rank nearly that highly, and it appears to be based on Google’s estimation of whether the searcher had “local intent”.  If you are looking at Google SERPs and see what looks like a great opportunity for your client to knock out a competitor who’s weak and holding a nice spot on page 1, check to see if that competitor is in the region you, the searcher, is in.

You can click on Search Tools, and then set your location to “USA” to eliminate the local universal results. Keep in mind, however, that if your client is in a major metropolitan area, you do want to know if local universal results are showing, as it still may be an easy “win” for them for all the searchers in their own city.

LU eliminate

#9 Presuming a high-volume term actually gets its results clicked on

OMG did you see the search volume for “tahiti”? It’s 60,000 searches a month, and the average couple spends about $10K to go there. Marketing department now needs fresh underwear.

But, is the organics their final destination? Or are they looking for cool images for their blog background, or for a school project? They may search from the title bar of their browser, then when the Google page comes up, click the Images tab, or the Maps tab, etc.

As well, many searchers will start with a relatively simple term (such as “tahiti”), see nothing that fits in the broad range of kinds of results that come up, and change their search to something a bit more specific (such as “tahiti hotels”). Even that might still not show anything that gets them excited, so perhaps they refine it further (such as “tahiti hotel reviews”). I’ll suggest that the broader the topic, the higher the percentage of searchers click on NONE of the results.

#10 Not recognizing alternate intents/meanings of your term

Our lovely, cryptic, rule-slaying language has endless examples of words that are spelled the same but have different meanings. As an example, I actually have two clients, both of which would say that “frames” are a good target keyword for them–but they’re not competitors. One sells picture frames, the other sells designer glasses.

This is a great reason to actually run the searches yourself: when a term has another relatively common meaning, Google will generally show some results for each of the meanings.  So what do you do with the search volumes for this kind of term? Well, like a lot of internet marketing, you have to wing it: look at the number of results for each of the meanings on page 1, and use that as a gauge as to what percent of the searchers are looking for YOUR meaning.

Keep in mind: if nearly all the results on page 1 are NOT matching your meaning, a real user is probably going to skim the results, see that, and then refine their search, so don’t waste your time trying to be the one exception on that page.

#11 Not recognizing PMDs (Partial Match Domains) in the results

Despite all the talk about reducing its influence, and the EMD Update from last October, Google continues its absurd fetish for exact match and partial match domains, giving them (in my mind, anyway) silly amounts of boost in rankings.

EMD and PMD boosts are meant to help searchers find brands, of course, and it does that WAY too well still.  So if you’ve got your trusty SERP Overlay turned on, and you’re running queries, looking for weak results on page 1 that your client can “bump out” with a little tuning, and you spot what appears to be a truly golden opportunity, take a close look at the domain name and see whether your search term is at least partially included.

#12 Insufficient messing around

Before you start the hard-core (yeah, I really mean “boring”) part of keyword research, dedicate some time to exploring your client’s site, their competitors’ sites, thinking about what their prospective clients are really looking for in their quest to choose the company to do business with.  As you learn about their products/services, you’ll have questions about what makes theirs different, what a certain term means, etc.

DO NOT ASK YOUR CLIENT…at least, not until you’ve asked those questions with Google searches. And if you asked such a question, and didn’t write it down somewhere as a search term candidate, well, shame on you.  :-)

Even 15-20 minutes of messing around – searching like you were a prospective customer for your client – will give you a ton of out-of-the-box ideas for search terms, and possibly help you spot opportunities that your client (and their competitors) have missed.

#13 Checking your rankings with a tool

Ok, one last mistake…and this one is really more for AFTER the initial keyword research is done.

Don’t rely on ranking tools. What I mean by this is: too many people look ONLY at their ranking reports, without actually going to do their top searches themselves on a regular basis. I had a client recently who wasn’t worried about their ranking for one of his most important terms, as his ranking report showed him steadily around position 4 or 5 on page 1. The problem was, the term had his city name in it, and he’d previously had an organic listing as well as a Google Places listing. The tool still saw the Google Places listing on page 1, and reported that…and he was unaware he’d lost his organic placement.

But there are more reasons to do the search yourself: you spot things like rich snippets all of a sudden starting to show up for that search; new competitors; existing competitors disappearing due to penalties; a switch in the kinds of results (local vs. organic vs. image vs. shopping etc.); a change in page titles or meta descriptions from competitors that make your result look less appealing.

An example of the later: your competitor starts offering free shipping, and puts that in their page title. Maybe you offer that too, but when a consumer does the search, that competitor screaming FREE SHIPPING in his SERP headline gets the clicks.

Conclusion

Doing keyword research is to search marketing as pouring concrete foundations is to building a house.  It’s messy, hard work, and nobody wants to do it, but the success of the entire project depends on it.

No matter how hard you work on content, your PPC campaigns, your social media, your link-building, if you’re doing it for terms that you’ll never be able to drive traffic for, you’re not going to succeed. And with a little luck, and a little creativity, you might find one or two gems that make you look like a hero without really breaking a sweat.

2 Comments

  1. Richard H. April 29th, 2013

    I think the biggest mistake people are making in PPC is still using “broad match” and not completely understanding it.
    It might seem like a good idea at first, but often times you can end up spending way too much money for clicks that don’t convert to leads.

  2. Spook SEO January 9th, 2014

    Okay I would love to delete this post for making me look like an idiot :( but seriously you are such a hero thank you for saving me from drowning in the river of mistakes. For quite a while now been doing keyword research based on simple guideline and mainly pure instinct. I came across this post and I was stunned, information still sinking in, digesting everything as much as I could. This is awesome. Out of all the 13 mistakes 11 is what I always do!

Leave a Comment


Michael Cottam is an SEO consultant and native of Portland, OR, with addictions to travel, motorcycles, and flying. He serves on the board of SEMpdx, is an Associate at Moz, and is the founder of the travel planning website Visual Itineraries.