What do YOU do when local keyword research says zero?

Greetings from inside the SEOigloo!

Here is a challenge we’ve run into enough times over the years to make it seem worth a blog post to me. I would really appreciate hearing from all readers on this subject.

What do you do when you’re building a website or doing SEO for a local business and your keyword research shows zero searches for what the service does or sells?

For example, supposing you have a client who sells sewing machines in Houston, Texas. While, obviously, you can find generic searches for sewing machines, brands of sewing machines, etc., keyword research indicates that 0 searches like this are taking place locally. If keyword tools like WordTracker, Google Daily Estimates, etc show you that 0 people are searching for things like ‘sewing machine houston texas’, what are you to make of that?

You know your client sells their machines. You know they experience enough sales to stay in business in their community. Are you to believe that no one is using the Internet to look for what they’ve got? It would almost make you feel that having a website will be pointless, though this can’t be true.

In some cases, using a tool like SEOdigger (thanks, Dave!) can be useful in looking at your client’s competitors to see what they are ranking well for and that can at least help you brainstorm which phrases might be valuable, but what if the client has no local competitors, or the competitors’ sites are are completely unoptimized and therefore not ranking well for anything?

You just keep coming back to zero.

This is a dilemma we struggle with and, in such cases, it’s been our custom to optimize for whatever we discover to be popular broad searches (sewing machines) plus the location, but at the end of doing this, it feels like you’ve not developed a very strong strategy. It’s like you’ve said to yourself, “well, I’ll do this just in case anyone ever searches for a sewing machine in Houston.”

There’s got to be a better way to go about this, but I’ll be darned if I can think of what it is when no amount of deep digging is bringing you any data.

Have you faced this with projects? What do you do?

18 Responses to “What do YOU do when local keyword research says zero?”

  1. on 18 Mar 2008 at 3:19 pm phaithful

    The issue is that keyword tools such as WordTracker and KeywordDiscovery have different sources for their data. Many of which are small ISPs which may not include users who are in the local that you’re looking to target for your local business.

    My suggestion is to search for locals that may have similar demographic characteristics and see if there are searches for that particular local and then search and replace with ‘Houston, Texas’.

    For example, I use ZipSkinny.com to figure out the demographic information for a certain local. In this case, Houston, Texas. Then I use the ZipSkinny compare tool to compare it to a number of other locations. Possibly, Phoenix, AZ; Little Rock, AK; Henderson, NV; etc….

    You can compare up to 100 locations. Depending on what demographic information you find most important, I’d start there. In this case Phoenix and Houston have very similar Population numbers.

    So I’ll go to KeywordDiscover or WordTracker and do similar searches for my niche with Phoenix as my location instead of Houston. Hopefully that brings up something useful.

    I repeat the process for other locals until I’m satisfied that the data set returned is of significant volume. Then the search and replace begins.

    Although this is not a perfect method, it is one that I’ve had much success with.

  2. on 18 Mar 2008 at 5:09 pm Pittsburgh Bartending School

    That is a tough one, Miriam. Phaithful has an interesting approach.

    Wordze.com has the largest data base of keyword phrases I’ve seen. That might give insights.

    Often I’ll go with the generic business terms and add the appropriate local phrases to it in both applications…one with the location first and the business terms 2nd and vice versa.

    here is some hard data. I’ve counted greater than 2,000 geo/business term phrases for a business in google alone, yet the single greatest # of visits for a specific term might only be 40 in a month. So the 2,000 searches equals a lot of keyword expansion.

    Now that business has high rankings for appropriate business terms and 2 state names and a city name. Its really optimizing on a regional basis. Assume slightly less than roughly a third of the terms cover each major regional term–so that’s about 600 per geo area. A lot of that will be irrelevant because its too far away…but what the heck…with #1 rankings I’ll catch folks for whom it is relevant.

    The key is widespread keyword diversification in 2 ways; one is a variety of terms relating to the service and the 2nd is geographic terms. I like adding town names. So If I were looking at a county, I’d add content in the site somehow that references every or most towns, and regional terms that might be applicable.

  3. on 18 Mar 2008 at 5:38 pm admin

    Wow, that is a totally amazing insight. I have never heard of anyone else using that technique. What a cool tip!

    May I ask you another question? Have you had to tackle this challenge for a business that, say, offers something that’s going to be more specifically popular in given areas of the country?

    Let me give another hypothetical example.

    Say you’ve got a client in a hot part of the country who sells swimming pools. Say they are in a somewhat sparsely populated region – maybe along the desert in Utah. Despite the fact that the client is doing well offline, selling their swimming pools, the typical kw tools are returning zero data, perhaps because the ISPs are not the ones being pulled from by the tools.

    In such a case population comparisons might be a poor metric. A similarly populated region in, say, North Dakota, would probably have a much leaner demand for a hot-weather item like a swimming pool.

    So, what would you look at? Similar population + average temperature?

    Sewing machines are likely popular everywhere, but items like swimming pools, snow gear, air conditioners, pellet stoves and other things tied to weather seem to present special challenges.

    I’d love to hear your opinion on this and am totally thrilled that you’ve shared this interesting strategy!

  4. on 18 Mar 2008 at 6:21 pm admin

    Hi Dave,
    Oh, I see. Can I repeat back what I think you’re saying.

    Say you had 100 basic terms and 10 regional areas you wanted to optimize for. This would boil down to 10 terms per geo area that would need to be split up over the architecture of the site. Is that right?

    I appreciate your hands-on experience in this, Dave.

  5. on 18 Mar 2008 at 8:21 pm Max

    I like Phaithful’s answer and will definitely be testing that out. But I guess I have one experiment to add. Pretty quick and dirty.

    I do the website for my Martial Arts school in St Paul, MN. It’s called Kuk Sool Won of St Paul. So not too many people search for “martial arts st paul”, but I wanted to figure out exactly how much. I also wanted to figure out if maybe people were typing in “karate st paul mn” or “hapkido st paul” or something else. All which came up with zero searches a day in Google and other keyword research stuff.

    So I convinced my instructor who is also the owner to give me some money to run a Google Adwords campaign. So I just ran one on all the keywords I thought people would be typing in for local just to measure how many impressions were on each keyword. Most popped up zero every week but some were surprising and that’s how I could see which local phrases had at least a little relevant traffic and which had zero. I just ran a small PPC campaign.

  6. on 18 Mar 2008 at 9:32 pm phaithful

    Interesting dilemma to face when you’re dealing with a niche within small markets.

    Well I’ll make the assumption that the market (city / zip) isn’t tiny but it’s relatively small in comparison to say Houston, TX. Because if it was that small the pool vendor would probably be 1 of say 3 other competitors. In which case optimization is somewhat moot.

    Also I’d assume that in a more sparse area, the vendor would service 2-3 different cities or counties.

    So I would probably do as you said, look at Population and find a local that has relatively the same pool market. If that’s not possible, I’d take a look at the pool vendor’s existing customers and see if there are any other commonalities.

    Are the pool owners mostly families?
    Are the pool owners mostly wealth? Possibly over $100k in salary a year?
    Do they share similar ethnic backgrounds?
    Are they mostly male or female?

    You can probably even take the above demographic information about the general population and then find a larger city that has a good pool market and see if there are any matches.

    Chances are, people with the same income level, who have the same ethnic backgrounds, and same household culture think in similar ways. I would probably target states that you think would also have similar cultural alignments if possible (e.g. Red vs Blue states, etc.)

    ZipSkinny is only 1 point of reference for demographic information. It’s a great place for free info, but if you’re doing this at a much larger scale and will need to access local information from many places, over many times… I’d suggest investing in data services from Axiom, Donnelley Marketing, or Experian (these are the largest and most expensive players in the demographic information gathering field… ment for large businesses).

  7. on 18 Mar 2008 at 9:35 pm admin

    Welcome to the SEOigloo Max,

    That was a good way of going about it, and good that your client was willing to fund that test!

    What I’d like to know, then, is whether the data you found from the PPC campaign was in opposition to what your initial keyword research indicated.

    In other words, was the Google Daily Estimate reading zero for “hapkido st paul” whereas the PPC campaign showed that their were, in fact, impressions happening for that, as an example?

    It seems like that’s what you found out, but I wanted to make double sure…that PPC yields different numbers than what KW tools indicate.

    I really appreciate you taking the time to share that hands-on experience!

  8. on 18 Mar 2008 at 9:52 pm admin

    Thank you for returning and further expanding on that. Yes, this makes sense. I have to dig into zipskinny.com. It sounds a bit better than what I’ve been using. The paid services (axiom etc.) are likely major helps to big business…maybe overkill for the small business realm unless the budget is large. Free is always good!

    Really, what this comes down to is understanding the user base, but I guess the SEO in me loves to see numbers and feels shaky when they don’t readily appear. Your suggestions are really good and useful as confidence boosters when those magic numbers don’t automatically appear! Thank you so much for the in-depth comment.

    I thought I’d add, I was talking with Mike Blumenthal about this today and he had another very realistic bit of advice to share that I thought worth adding here.

    Sometimes, the key benefit of a website to a business owner is not going to be huge volumes of search engine traffic. Sometimes, particularly in rural/small-town areas, for some types of products and services, most locals already know a business exists. They see it every time they go into town.

    Because of this, setting correct expectations for the client is important and better goals in a case like this will be along the lines of:

    1) Ensuring that the business is being represented in a professional manner online.

    2) Taking control of local data wherever possible.

    3) Providing an online destination for offline marketing.

    4) Cutting operating costs. For example, clients being able to print company materials off the web for themselves rather than the business owner printing them. (I’ve seen this cut costs hugely for people doing newsletters, actually)

    5)Ensuring that new people who move into the area are able to find out about the services that exist in their new location.

    All of these factors, and others, can be of great value to a business owner even when major Internet traffic does not exist for terms+location.

    It’s good to remember that!

  9. on 19 Mar 2008 at 10:52 am Don

    @Max –
    I’ve tried the local-PPC strategy, as well. What I’ve found is that after a couple months of running an AdWords campaign, Google’s Keyword Tool suddenly has stats for a lot of new, focused phrases. It appears that the PPC campaign triggers the AdWords bot/algo to begin gathering (or at least displaying) web-wide KW data for that niche.

    Now, several months after starting an AdWords campaign for which I had to do a lot of guessing, I not only have a lot of data from the campaign itself, I also have loads of data via the keywords tool.

  10. on 19 Mar 2008 at 12:30 pm Pittsburgh Bartending School

    I think all the ideas are great.

    My main point is that for a local/regional buisness there may not be a measurable amount of traffic for most keyword tools. Wordze, of course uses an enormous database, and it appears that seodigger also uses an enormous data base and presents the information differently.

    Keyword expansion is the key to drawing lots of traffic regardless how tiny the traffic for the main phrase might be.

    I like all the suggestions above, but my experience dictates that in a variety of industries there are a lot of applications of the secondary keywords for a generic business term, and within that geography there are many searches of the generic business term and logical geographic areas.

    In my example above, if I broke it down to one state, the single most active business term with the state name might deliver 40 visits, but I’m getting about 600 total visits with the state name and a variety of business terms. A couple of relatively highly trafficked terms, might include the main business term and the state initials, and the logical variations of state name first, main business term 2nd; state initial first main business terms 2nd and then searches with the main business term first and the state name or initial 2nd. But even after those big 4 variations, I still might have all of 100- 120 searches out of a total of 600.

    The bulk of traffic is coming from variations on business terms.

    Another way to expand traffic for a region or a county might be to add town names within the county onto the content of the site. I don’t know the towns in California counties, but in Maryland/Montgomery County, I’d add content that includes Bethesda, Silver Spring, Germantown, Rockville, Olney, Wheaton, etc.

    That adds traffic with some searches representing searches for the business term and town names. Cripes, I’d add zip codes …some people search for a service and add their local zip code. I’d add regional phrases that are used, for Chicago, I might add “the loop” to content, and for my area I’d add Southern Maryland specifically to content.

    I really like the ideas above for ascertaining the usage of a business term and the geography….but I also like using generic terms for a business service from any basic keyword tool, ensuring that you are adding all the secondary terms and then just adding all the geography you can into the content to try and show up for any type of search in the area.

    Frankly, Will’s example of weak links working for a local business also gives insights into the wide variety of 2ndary terms that might be applied.

  11. on 19 Mar 2008 at 2:26 pm admin

    Thank you so much for adding your PPC experience to this. It sounds like it has yielded some very good data after a couple of months’ observation. Cool!

    Now, the trick with your extremely good route will, of course, be getting clients to understand the importance of expanding their website so that multiple terms can be covered. So often, small businesses are looking for small sites – 5 pages or so. That’s not going to cut it if you’re looking to cover dozens, scores or hundreds of phrases.

    This is why I love to advocate the addition of a blog to a business site because of the ability it gives the siteowner to continuously expand the keywords they are covering, widening that ring around the bullseye ever farther.

    Great stuff, Dave!

  12. on 19 Mar 2008 at 3:21 pm SMG Directory Marketing

    One thing that I’ve done working with local businesses, and something that was briefly touched on prior is running a paid search campaign in that area to see what people are searching for. We generally run an IP targeted campaign, which will give you an overall idea of what people are looking for, and a geo-modified campaign, which will provide examples of which phrases are being used with geo-modifiers. This has worked out extremely well for us.

  13. on 19 Mar 2008 at 3:35 pm Pittsburgh Bartending School

    I went into the analytics package for a strongly ranked local business site for the last 30 days:

    Over 5300 searches

    the site is highly ranked in engines for the generic business terms and of course the local terms.

    Most searched phrase, over 400 searches

    most searched local/regional phrase 39.

    I scanned hard over search terms that had hit the site at least twice.

    Of that I counted 18 different variations of applicable business terms with an appropriate geo phrase. This is a site that works 2 state names and a city name and it is a large market. Business wise it can draw customers (not many) from 2 nearby metro markets. I had previously checked the metro regions population totals and between 3 smsa markets there is a population of about 7 million + so it appeals to a pretty big market.

    Then I did a re review to look more closely at the impact of organic search traffic that was very local.

    Of the search phrases that had a local context (the city name or an appropriate regional/county/town name that included a business service phrase there were 441 searches (i.e. I excluded the phrases with state names since some of those people were looking from too far away.

    Then I scanned the single visit phrases. I stopped counting at 70 visits for a phrase that represented a nearby town/county name/zip code with the business service.

    So for the most recent 1 month period I’m looking at well over 500 search phrases that represent a purely local context, at least.

    We get more search phrases by far for the state names than the city names and I’d guess that at least 30% of those searches are within the geographical reach of this business. (It could be close to 50%)

    Anyways this doesn’t answer your question but it sort of speaks to the long tail of adding content to capture a lot of phrases.

    I think all the methods above to pull search terms are appropriate. My experience is that I’d use a search for the generic business terms without the local geography phrase and just start adding content than optimize for the most important combinations, based on guess work.

    If your client will be willing to run an ad campaign I’d do that: even without hits your getting estimates on impressions and visits and that gives you insights on how and where to optimize.


  14. on 19 Mar 2008 at 9:08 pm admin

    Hi SMG Directory Marketing,
    I have great respect for you PPC folks. That is not our area, and the whole process of targeting ad campaigns is a fascinating one. It sounds like, from your comments, and others, this is a very rewarding way to do research.

    Thank you for taking the time to share what you know. Hope to see you here again.

  15. on 19 Mar 2008 at 9:15 pm admin

    “We get more search phrases by far for the state names than the city names”

    See, and that’s not something you’d intuitively expect, Dave, though I suppose it would depend on the nature of the search. I can imagine searching for ‘Audubon Society CA’, but not ‘pizza California’. I guess the type of business will influence, naturally, to some degree, whether state or city is more prominent, but your data shows that state names may yield much great traffic in some instances.

    That’s really neat.

    You blow me away, Dave.

  16. on 19 Mar 2008 at 10:56 pm Max

    When I ran the PPC campaign for my Martial Arts school some of the keywords I tried did get zero but some got more. For example, my Martial Arts is a comprehensive one so I could say we are a sort of mix of hapkido, karate, tae kwon do, and kickboxing. So I could run adds about all the style, but if I was just a tae kwon do it would not make sense and my bounce rate probably would have been a lot higher. So I tried a bunch of different Martial Arts style titles with our city after it and some did get zero but some also got more. Some reaching 50 to 100 impressions in a week. Those weren’t the only keywords I was testing but you get the idea. It was pretty cheap and quick to figure out what was happening in the local market for Martial Arts.

  17. on 20 Mar 2008 at 3:39 pm admin

    Thanks for the clarification, Max. Sounds like it’s a good thing your school offers diverse classes!

    Hope to see you here again.

  18. […] What do YOU do when local keyword research says zero? 1Cat.biz Tags: keyword research, keyword tools, wordze […]

Trackback this Post | Feed on comments to this Post

Leave a Reply