Many businesses seek to target local searchers for their products and services, and careful review of a site’s analytics is one of the key weapons in a local business’s competitive arsenal. Here are a few tips on what metrics to look at and how to use some of them in optimizing your site to rank better in local search results.
I recently had the pleasure of speaking at the SMX Search Analytics conference, and one of the topics I spoke upon was how to use Google Analytics in optimizing local sites. Some of the examples in this article come from that presentation, and I’m adding a few other details to the mix.
First, to set the stage, many businesses have a moderately good idea as to the sort of keyword phrases for which they would like to rank. A “family law” attorney in Miami, might believe they should appear high in the results for all “family law” queries, but if they ranked at the top of all searches for that phrase tons of people elsewhere in the country would see their listing, which would not be all that relevant to those people’s needs, and the attorney’s site would get a lot more traffic and needless phone calls than they should have to deal with. Searchers trying to zero in upon businesses local to their area might add a geographic qualifier to the end of the term, and search for something like “family law miami.”
In actuality, a business ought to first do some keyword research to figure out what keywords and phrases consumers are most likely to use when seeking their type of company. In this attorney’s case, they might quickly find that while “family law” is a formal term more preferred by their profession, more of their potential customers are likely searching for the term “divorce”. And, in most cases, consumers are searching for “lawyers” when trying to find listings of this type of business, rather than “attorneys”.
So, if thorough research were done, we’d find that the more advantageous terms to target for their area would be both “divorce lawyer” and “divorce attorney”, coupled with the “miami” geographic qualifier.
Even more subtly, should the phrase have the city name before or after the business type keyword? “divorce lawyer miami” or “miami divorce lawyer”? Again, diligent keyword research would tell us (both phrases, “divorce lawyer miami” and “miami divorce lawyer” are most frequently occurring). In most cases, the slightly higher-frequency phrase will have the geographic qualifier first.
Once you have identified your ideal keyword targeting phrases and have properly optimized your site for those terms, then you need to do a bit of review with your analytics results to double check how things are actually functioning.
Your assumption is that now you’ve targeted your local search keyword phrase, you should be getting mostly searchers from your local area coming to your site (some types of businesses are excepted, such as travel-oriented companies where many people from outside the area could be seeking it out, such as for hotels, rental cars, cruises, etc.).
Top analytics packages provide reports on the geographic locations of your site visitors via geolocation technology, so you’d first want to look at those reports to see if you’re getting traffic from visitors in your local area. For example, if you were a Texas-based business, you’d expect to get a lot of traffic from visitors within that state. Here’s the sort of reports you might see within your analytics package (this example from Google Analytics):
There are some instances when you may get substantial traffic from really distant locations when perhaps you should not. For instance, if your business name contains the name of some other geographic area, it could be coming up as matching user queries for that area. For example, the “Denver Southwest LP” company has offices physically located in Houston, Texas. And “Houston’s Restaurant” has a physical location in Dallas, Texas. The “Adele Dallas Orr Fashion Boutique” is located in Chicago, Il. There are simply tons of instances where businesses are sharing names with distant cities.
There are also cases where a business’s city name is shared elsewhere as well. “Houston, Mississippi,” “Dallas, Florida,” and “Denver, Pennsylvania” are just a few cases in point.
You may not even realize that your site could be getting a substantial bit of unqualified traffic from people way outside of your area! Analytics can tell you if you’re getting an unusual amount of traffic from places outside the area where you offer service, and even tell you which landing pages are receiving the most traffic from these people.
If you’re not getting traffic from the people in your local area, and/or if you’re getting suspicious amounts of traffic from another city because you share a name with it or the city name is shared by another city elsewhere, then this indicates you may need to adjust something.
If you were a pizza restaurant in “Houston, Mississippi”, you might feel a bit of despair when you quickly realize you might never expect to rank for searches for “pizza in houston.” However, internet users rapidly clue into how the search engines function, and they respond by self-adjusting their queries to get what they’re seeking. Natives of Houston, Mississippi likely very quickly learn to append the state abbreviation to their queries, such as “houston ms pizza”.
So, if you found yourself getting a lot of unqualified traffic from people very distant from you, you might review to see if the SEO design of the frequent landing pages of those visitors was sound, and you might find you need to make phrases in the titles and header tags, and perhaps other page signals a bit more specific by adding additional qualifiers. You might also need to dig around and check to see if any major online directories have incorrectly indexed your business in the wrong city, or accidentally applied your website URL to a listing of a business with a similar-sounding name in another city. Such data can get fed into the search engines, and it’s not unheard of for Google to apply URLs to the wrong business.
If you are unlucky enough to have a business name which is generating unqualified traffic due to these sorts of issues, it does not mean you need to totally freak out about it. You should expect you’ll probably continue to receive unqualified traffic from the related locality well into the future. Being overly aggressive about trying to eradicate all unqualified referrals could harm your overall SEO health, so don’t go overboard. Reviewing your analytics, though, may help you to realize if and when this is happening, and help you to tweak things to reduce the incidence. This is mainly additive, in the sense that it may lead you to add more qualifying terminology into your pages such as names of neighborhoods, zip codes or regions.
(In an effort to further refine the local search experience, Google has recently begun to geotarget users for some non-locally-specified queries with some local business results, reducing user dependency upon adding geographic modifiers to keyword phrases, and reducing the incidences of local businesses getting unqualified traffic. My expectation is that if the functionality is not later revoked, this will actually change searchers’ behavior over time so that users will become lazier about adding in geographic modifiers as they find they do not need to in order to get back the information that’s precisely what they’re seeking.)
Another area to look at closely in analytics will be your keyword reports. These are derived by your analytics system from the “referral URLs” of popular search engines. Your analytics system parses out searchers’ keywords from the query strings in referral URLs of major search engines which send traffic to your site, and then they compile the numbers of referrals brought to you for each keyword and keyword phrase over time.
Once again, you’d expect that a lot of the keyword phrases which bring traffic to your local biz site would include those local qualifier terms—primarily your city name in most cases. Here’s an example from a friend’s site about Texas history (he lets me experiment with his site in return for some optimization advice):
Keyword reports may help establish whether the terms we’ve set up following initial keyword research are more popular or not in actual practice. If you’ve experimented over time with targeting both “miami attorneys” and “miami lawyers” (and achieved equivalent rankings with both), yet you find that the “attorneys” term actually brings in much more traffic, you may want to adjust your strategy to exploit what works best. Keyword research is a starting point, while keyword research coupled with analytic results is an ongoing refinement process.
The keyword reports not only help to establish if you’re receiving traffic from the phrases you’re targeting, but they’re highly valuable for researching to see how users arrive at your site, and also for discovering additional phrases that you might not already be specifically targeting. If you dig down into such reports and find some phrases or references to types of content which you had not targeted, consider building pages and content specifically to be relevant to those terms and you could key into even more qualified traffic.
I recommend that you look at your keyword reports and search on popular phrases yourself. In this way, you can see things exactly the way that your site visitors will experience them, and you may discover areas where you can improve and optimize further. What if a user is arriving on your site on a page that’s delivering an error message, or what if it’s not the most-ideal content for the query? It serves little purpose if you initially bring in a visitor only to have them immediately bail out when they feel you’re not delivering what they’re seeking.
Just to underscore this point, Avenish Kaushik, Evangelist for Google Analytics, recently blogged about how Bounce Rate can be effectively used to improve site performance.
I don’t know how many times I’ve delved into a client’s analytics, only to discover that users are arriving on the site and are then unable to locate the info they’re specifically seeking. For example, on my friend’s Texas history site, one of the top referral keyword phrases has been “alamo historical marker:”
Yet, when I clicked through, I found that Google had decided that the page for the “City of Alamo, Texas” was most-relevant on the site for this term, rather than the pages dedicated to the famous Alamo building in San Antonio. There are a number of ways of addressing this issue, but we decided to deal with it by providing a little helpful navigation aid on the upper right corner of the page to assist such visitors with rapidly locating precisely what they wanted:
Once this was added, the page views per visit for those arriving via the query “alamo historical marker” was improved considerably:
Noticing things such as this and making ongoing quality improvements can really help to improve traffic, conversions, and, yes, even rankings over time.
Since most top analytics systems can inform you how many referrals come from each search engine, it somewhat reduces the demand for those engines to inform businesses how much traffic they pass via the organic keyword search and local search listings. However, external analytics cannot tell you what the percentage of total searches is for your desired keywords for which you’re receiving clickthroughs (i.e. “clickthrough rate”). You’d need to know the total number of searches going on by each phrase in addition to how many referral visits you’re getting to achieve this, and you only know the total referrals portion of it. In the case of Google, it also doesn’t allow you to know how many of the referrals are coming as a result of your listing appearing in the local one-box versus in the regular keyword results listings.
Google’s local business center has begun displaying how many impressions versus “views” your listings are receiving (I have slight heartburn over this terminology—I think they should use “impressions” and “clickthroughs”.) But, unless you capture this number on the first day of each month, subtract out referral visits from the “maps.google.com” subdomain, and then compare against your overall visits referred by Google search results, you still cannot deduce your general CTR on your Google local one-box versus the regular keyword search results.
To my knowledge, none of those software packages which monitor search result rankings are really able to differentiate between the local one-box versus regular search results listings on pages, either, making it very difficult for people to track their rankings in the one-box results over time unless they do so manually. This is disappointing, because it would probably be very easy for Google to provide these metrics.
Martijn Beijk wrote some really great instructions (published on Mike Blumenthals’ blog: Tracking Local Search Traffic With Analytics) on how to set up tracking URLs to use for your listings in the Google Local Business Center, using some customization in Google Analytics to help differentiate the clicks between the 1-box/3-pack/10-packs and regular search results.
Local businesses also receive phone calls from their local listings, and there are some businesses which make use of various call-tracking methods to attempt to gauge which of their promotional channels are more effective at sending them customers.
Purely from the standpoint of search engine optimization, I’m not a terribly big fan of using tracking URLs or special tracking phone numbers. I believe that there is a risk that such tactics can negatively impact potential success of optimization work. One “happy medium” solution is to set up tracking URLs and phone numbers for a briefly limited period of time, such as one to two months, and then go back to using the primary URLs and phone numbers for the long term. While this might disappoint data-junkies, it would get some actionable data while staying out of the way of long-term optimization potential.
These are just a small handful of ways in which your web analytics may help you in understanding your visitor traffic and in optimizing your site for local search. I encourage businesses to get comfortable with their analytics packages rather than merely implement them and ignore them, and to start intelligently exploring what the data means, and how it can be used to help improve their site performance.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.