SBS Mailbag: When Local Keyword Research is a Dead End

Filed in Local Search, MY BEST POSTS, SEO by Matt McGee on March 26, 2009 21 Comments
Local Keyword Research

This great question came in today from Mary Ann:

I am helping a small dance/fitness business with local SEO. Keyword research yields zero (0) search volumes (Keyword Discovery) for keyword terms once I add the location or any variation on the location/general area (Westport, Connecticut). Any suggestions as to the best keyword strategy in such cases?

My guess is that almost everyone involved in local SEO has run into this problem; I know I have. As much as I love Keyword Discovery, it has its limits (like every tool out there). In Mary Ann’s case, neither Keyword Discovery nor Wordtracker show results for possibly prime terms like “westport fitness club” and “westport dance studio.”

So, here’s a list of what I’d do next if I’m working with the client Mary Ann describes:

1.) Do non-local keyword research

The first thing I’d so is stay on Keyword Discovery and forget that this is a local client. Just do research on the industry/business terms. Find out if “dance club” is generally a stronger phrase than “dance studio.” Find out if “fitness club” is stronger than “fitness center.” Find out if “fitness dance studio” is a strong term. Look for appropriate misspellings and related terms. Do all the keyword research you would normally do, just without the location-related terms. Generate a list of relevant non-geographic terms and save it.

2.) Add geographic terms to that list

Once you have your list of industry-related terms, it’s pretty easy to just add the appropriate city, town, county, neighborhood, and other geographic terms as modifiers. If you’re dealing with a small geographic coverage area, this shouldn’t take long. If it’s a big coverage area, use this local keyword generator tool. In Mary Ann’s case, she’d input her non-geographic terms, a central zip code, and a radius, and the tool will spit out a list of keywords that mashes it all together. Mary Ann will probably need to prune that list because (like all tools) it’s not perfect.

But, at this point, we now have a big list of keywords that includes well-researched industry terms with appropriate geographic modifiers. The only problem is that we don’t know for sure how strong any of the individual keywords are.

3.) Try other keyword research tools

The Google AdWords keyword tool draws a blank when you type in “westport fitness club”, but it does show some numbers for “westport fitness.”

local keywords

Again, that’s not perfect, but it’s progress. Using related keywords like “westport gym” and changing to nearby city names like Norwalk or Fairfield should help create new keywords and/or confirm those already on the list. On a related note…

4.) Pretend you’re in a larger city.

Since Westport seems too small to generate a lot of good keyword data, I might pretend I’m in a larger city nearby and see if there’s good keyword data for that area. Hartford might be a good substitute for Westport, and if that doesn’t work, I might even try Boston. True, many cities have their own language, but there should still be some good insights to be had by substituting a nearby city’s name.

5.) Use Internet yellow pages for keyword ideas

You can go to, for example, and type the word “fitness” to see a list of related categories — note that you’ll have to hit the ENTER button twice, because the first time it demands a location. Just click again and you’ll get a page like this with several dozen category names in the left column. Each one that matches your client’s business is a potential keyword. And you can take these category names and redo the keyword research you did earlier.

(Note: It’s not a yellow pages site, but you can also browse categories with Google’s Search-based keyword tool.)

6.) Setup a PPC account

This is the best way, in my opinion, to do local keyword research, especially when you’ve hit a dead end with the traditional tools. Unfortunately, it’s not free like the options above. But setting up a PPC account, even one with a smaller budget, will give you exact keyword counts for all the terms you bid on. The danger is that, if you set your budget too low, your keywords won’t show enough to collect good keyword data.

What I would do is take the keyword list I’ve built via the previous items on this list and create several keyword groups and ads, run a campaign and watch over time to see which keywords get searched.

Your turn: How would you have answered Mary Ann’s question? What did I get wrong or right in my reply? Comments are yours…

(photo courtesy bennylin0724 via Creative Commons)

Comments (21)

Trackback URL | Comments RSS Feed

Sites That Link to this Post

  1. Mihmorandum | Links of Local Interest, Volume 5 | Local SEO | April 14, 2009
  1. Stever says:

    I use the “pretend you’re in a bigger city” approach. For most the broad type terms I assume most people search in similar ways. There may be special cases of a “unique language” for certain businesses in certain areas, but for the most part that’s pretty uncommon.

    Also when using the Adwords Keyword tool I always start with the one main keyword (sort by volume), “fitness” in your case, and see how the tail lengthens with other variations on that term. Things like fitness center, fitness gym, fitness clubs, fitness center, etc… Then it’s just a matter of adding a geo-modifier to the phrase.

    In some of those broader searches when you scroll further down you sometimes start to see the geo-modifiers popping up from larger centers like NYC, LA, San Fran, Houston, etc…

    A possible test of location specific unique uses of certain phrases (i.e. pop vs soda) would be to try Google Insights and look at the heatmap of USA. Maybe drill down to the state level too. You start getting more and more granular and there may not be enough data to show any trends there, but sometimes it might.

    Speaking of Soda vs. Pop (I say pop) check the differences in the maps between these two;

    Looks like the North Central part of the country says pop, while the east, west and south say soda.

    I went with the “pop bottle” and “soda bottle” terms because a word like pop fits into too much high volume entertainment searches (pop music, pop culture, pop stars, etc.) and would skew results too much

  2. John says:


    Great post (good mailbag question, too!). Local search keywords rarely have the statistical data behind them to make SEO keyword research “easy.”

    In this situation, I typically cycle between steps 1,2 and 3. #1 really is the most effective. If a general search term has relatively high volume, it can typically be assumed that search volume will translate to modified keywords for the microcosm of your city/state/region, etc.

    Thanks for the great post!

  3. Matt, this is an important article outlining one of the most confusing aspects of local seo and local sem for small business. I’ll be sure to point my readers and viewers your way in a future article. Cheers and keep up the great work!

    – Julian

  4. Its a good question, its a toughie, and I’ve seen the question asked several other times, specifically by Miriam Ellis, more than once. I’ve struggled with it.

    I’m not sure which would be first in my book, but I’d A) both focus on the industry term….and focus on the industry term w/ more than one large city to see trends.

    B. I’d run an adwords campaign as you suggested. I’d use exact search phrases and I’d include a lot of phrases with the relevant town and nearby towns.

    Assuming the search term and related terms don’t come up much for the relevant town I’d do the following:

    I’d optimize the site for a geographic area in which the business terms tend to show….ie state probably or county or popular local regional geographic area. I’d also optimize for the town. At the state level….assuming there are a reasonable number of searches for the phrase….you’ll capture traffic from a reasonable number of searchers in the town and nearby towns, etc.

    One saving grace is that more and more often Google is running a version of maps for generic industry phrases such as “dance studio”, “fitness club”. That can capture traffic for folks in the region who are searching for a dance studio, but aren’t searching with a town/county/regional phrase/state name. That brings to mind that for such a phrase one should run a regional ppc campaign aimed at the generic industry phrases along with the industry phrases w/ geo modifiers.

    After all is said and done it points out that the payback in search for certain geo focused phrases might not that be all that great…..there just aren’t enough searches in the region.

    But don’t despair….start turning to social media sites such as Facebook, Twitter, MySpace, Flickr, etc. and populating info about one’s business.

    It is a toughie, though, no doubt.

  5. Todd says:

    Good post. I find that saturating with both town related keywords and local targeting is a good (and constantly tunable) cocktail, but there really is no substitute to thinking like a searcher. With most of my customers, you ultimately end up finding success with the pretend-you-are-in a-larger-city dance.

  6. Martijn says:

    I have had interesting results in Holland with the following approach:

    trying a broader marketing approach to attrackt potential customers who are orienting themselves on sports in general. Their search motivation can also be health improvement or even weight reduction. Use keywords like: cycling, sports, weight reduction or even stress relieve, etc.

    Geo-target the relevant area with these broad keywords. Then use your advertising text to narrow down to the actual service offered in the area applicable:


    Title: Fitness in Westport (abbreviation for Connecticut).

    Note: this approach will be more expensive because the keywords used will not always be relevant to your website content.

    You might want to consider adding a page to your website addressing issues relating to the afore mentioned weight reduction & stress relieve related keywords thus increasing quality ratings of these keywords and reducing the CPC costs.

  7. I 2nd Martin’s idea. As a slight tangent I’ve had success by both adding content on the site and advertising for core reasons why people want to use a service. By example content about “losing weight, getting in shape,” etc have worked (oooff….that is exactly what he said) :D….Well I second it….and it works 😉

  8. Miriam says:

    Ah, Dave, isn’t that funny? We were just having this conversation last week. I immediately thought of this when I saw the question in Matt’s post!

    I get the feeling that most of us are kind of walking the same walk when it comes to this challenge. Martijn’s idea is quite a good one. Great post, Matt.


  9. Tommie says:

    I would also suggest that she look at markets that are parallel. Find a another market that is closely related from which she could gain customers from because it is directly or indirectly related.

    This would involve doing some market research to find out what her customer base likes to do online.

    Sometimes ranking for parallel markets can get astonishing results.

    There are plenty of freely available tools for finding them.

  10. Adam says:

    Yes – I’ve had this frustration a number of times. I’ve even told clients that I didn’t think SEO was good for them given the low search volume on their terms. Even in big cities, the volumes can be very low on all but the very popular searches.

    BUT – Google is expanding their local treatment into new search phrases. Maybe someone already said this, but it seems like there will be new hope for local SEO based on this change – I think. I can dig up a citation if anyone wants more info.

  11. Matt McGee says:

    Thanks Adam. I wrote about that Google change here:

    By the way, just because you don’t see much volume for local search queries in keyword research tools, I don’t think I’d ever recommend a local client not do SEO. As someone who runs our hyperlocal blogs, I can tell you that the amount of local search traffic is astonishing. And it never shows up on the keyword research tools.

  12. Stever says:

    @Adam – you cant rely on keyword suggestion tools to determine search volumes. They are not always accurate, they are just a baseline to start with. Like Matt just said the volumes for local are much much higher than you think. The long tail is very long in local.

  13. Jim says:

    Oh the dreaded “not enough data” local seo dilemma….how I loathe thee.

    I agree with all of the above suggestions as well. I have honestly had REALLY successful results just using my brain a bit and adding a geo qualifier to the popular search terms on a national level. I really like the comment about the “google insights” tool as well.

    Great post!

  14. Jo Dodds says:

    @Matt great point about local search traffic not showing up on keyword tools; I also take the view that with long tail keywords and truly local suppliers it actually doesn’t take much traffic for them to convert quite a few customers given the % of offline purchases that start online and specific searching using geo qualifiers. So only getting traffic in the 10s or 100s is actually very beneficial for them, with a higher potential conversion in comparison to internet marketers and the usual 1% conversion.

  15. Tyy Ward says:

    Hey Matt,

    Great post! I have been looking for information on local SEO for quite a while now and just found your post today.

    I have a question for you regarding doing searches on non-localized search terms and then adding the geo keywords to that term. Do you recommend only going after the really high volume keywords and then adding the geo keywords or would you go after the long tail (low hanging fruit) keywords also?

    Mortgage Rates = 9,085 daily searches
    VA Mortgage Rates = 166 daily searches

    Would it be worth going after VA Mortgage Rates + city and/or state? Or is that to little to spend time on? As you said in this post, there is no search volume when I add geo keywords and it has been very frustrating. Your post couldn’t have come at a better time for me. Even though I am finding it 7 or 8 months after you posted it. Thanks for the help.

  16. Andrew says:

    The problem with low volume keywords on any of these tools, even Google’s, is that they almost always inaccurate and unpredictable. The only way to really be sure is to run an Adwords campaign as you suggested. But I think we are neglecting the other avenues for traffic other than Google…I mean we have Bing, Yahoo, Facebook, Shopping search engines, etc.

  17. TJ Kazunga says:

    My 2 cents:

    1. Google’s data for local search is not very reliable so use your common sense. If you know that “fitness club” is a popular search term, then it is likely that “Westport fitness club” will yield some results, regardless of what the keyword tool says… at least in my experience.

    Also, remember that a large proportion (not sure of the exact % but remember reading 50%?) of searches are unique to that searcher so will not normally show up on a keyword tool.

    Most powerful tip. Optimise for the main keyword i.e. “Westport fitness club” but make sure that you use a lot of relevant modifying keywords, including secondary geo-terms (such as Hartford or even Boston). As you begin to build quality backlinks to the website, you’ll find that page will rank for multiple terms.

    2. I personally always go for the Adwords option first. Persuade your client to spend £200 ($300 – $400 USD – or just do it yourself and include the cost in your SEO set up fees) to test which keywords will convert and do that first – then optimise the website for those those proven keywords.

    Glad I found this blog.
    Cheers to all


  18. Don’t forget about your analytics information!
    I’ll have a look at what keywords are being searched to bring users to the website. Then I’ll go and have a look to see where the website is ranking for those keywords the SE’s.
    If the keywords are bringing in good traffic volume, but are NOT ranking in position 1, I’ll get them ranking in position 1. That SHOULD improve traffic volume for that keyword. I continue to do that without keywords in Analytics, and then use broader, generic versions of keywords.

  19. Greg Boyer says:

    You could use this equation to figure out local search counts.

    City Population (divide) US Population (times) global monthly keyword count (divide) by 2 = your local monthly keyword count.

  20. art says:

    I suggest also reviewing number of competing pages.

Leave a Reply

Your email address will not be published. Required fields are marked *