In this part of SEP's Expert Week - International SEO, well a look at how search engines go about the magical mystery that is geo-relevant results the technical view at least.
So let's get geeky wid it and look under the hood at the core elements behind geo-targeted search results (and how you can get the most from it).
Local SEO it is as much a vertical search element as the traditional ones were used to such as Image Search, Blog Search and Video -- and a great example of effective geo-targeted search. From interweaving of local urls to addition of local results right in the SERPs, it is everywhere and by understanding its nuts and bolts, we'll better understand the technical side of geo-targeting.
How search engines geo-target
So, the main question is; how do search engines go about geo-targeting in the first place? While methods vary, the main points that are common to most approaches in modern information retrieval include;
Triggers - triggers for localized results can include;
- Query characteristics; the language used, location information in the query, revisions to the query etc. The search engine would look to see if there are geo-graphic elements to a given query to establish if it should start looking for documents relating to the geographic area.
- Interface characteristics; what are the language settings of the browser? Mobile or desktop? IP address? They can check the application the user is operating and even the type to establish potential signals. This phase is also used to detect non-query related signals such as the location of a given user.
- SERP characteristics; are there heavy geo-graphic tailored results to the query? By looking at top documents for a given query or using historical query analysis, search engines can also find potential geo-graphic bias which can affect the results returned.
In some situations search engines will even seed a small set of the top results with related terms or locations to a given query. Through a process of query analysis they can see what query revisions and geo modifiers occur in a given query space to possibly return a greater number of local results into the mix.
Explicit signals
Often times a query can be one that is obvious to being a geo targeted search. But even those can contain ambiguity. One such mechanism can look at a query that has inherent ambiguity and try to offer possible results such as using a term such as; ' PIZZA RESTAURANT ARLINGTON '
The search engine might be unsure as to which location Arlington, Va. or Arlington, Tex. By looking at secondary signals (such as user IP) they may seed the results with more bias to one or the other. With a query characteristic that is less obvious such as searching the term CRIB MOBILE a search engine might not be sure if it is informational or local retail related and once more, seed some local results.
Implicit identification
Another area that can be used is what is known as an implicit location. Lets say that the user searches for restaurant around Space Needle " there is an implied location not a direct one. This is one area where looking at the user location can come into play, but if theyre planning a vacation, another false positive is created. At which point cross referencing data from top results in combination with other factors can lead to the best possible results.
The inverse can also be true when a lower level query may not be expanded in an implicit fashion. When searching for History of Quebec a search engine would not assume that replacing Quebec with Canada is unnecessary since Quebec is the lower level target.
For more check out; Indexing implicit locations for geographical information retrieval (pdf)
Event anomalies
There are also a few instances where signal decay for a given query type can assist in knowing what the user is after;
Snowstorm in North America " could be identified by looking at other query events (with more specific geo locations) or the users IP to seed local results
U2 Concert Canada " could also be identified better by understanding other queries and their temporal nature, (as could the user IP).
Essentially many queries have an eb and flow in their temporal signature which can be used to further understand the intent of the user. The nature of news/events can be used to refine explicit/implicit local searches.
At all times, with any of these systems, they would collect a variety (historical) query data and constantly refine various query spaces to establish the likelihood of a given term requiring more/less local results. This can also be tied into personalization and your search history can come into play.
. Lets see how they judge the target pages shall we?
A geeks guide to Local SEO
The next link in the chain is understanding how search engines go about establishing potential location information for a given page/site. This is the area most SEOs want to know about when it comes to geotargeting.
Web page characteristics; for classification of a page can include -
- URL for domain extension " using the TLD of the actual domain the page is on.
- Language and nuances " a mix of best results and native language results.
- IP of web server " the location of the server hosting the page.
- Location information of links pointing to the page (by IP)
- The text near the (inbound) hyperlinks, such as location information of the business.
- Textual information on the page " signals that denote a geographic reference.
- Other web pages on the same site (eg Contact pages) that the page is on.
- Registrar examiner " for business location information;
- Links pointing out from the page
- Business listing (local directory or maps)
- Postal codes
The reason that a combination is used is that alone, they often dont tell the whole story. Lets look at the IP of the page/site in question; it can be entirely wrong. Not too many businesses actually host their site in the same area as their brick and mortar location (or even the same city many times). This can create a false positive. Thus no single element is going to be a required part of any geo targeted local SEO campaign.
Also, each search engine is going to treat things differently, these are simply some of the more common elements used.
Building your local SEO campaign
And so I hear ye grumbling, WTF eva brother Dave, gimme the goods and give the babble a break!-- Ok, ok I know that most of you really just want to know how this all plays out (and some of U sneaky types skipped to this part as well) so lets look at what it all means.
For starters, dont panic. Many times when folks approach me to ask about local SEO targeting they are concerned about the finite details. Dont sweat the small stuff as the best approach is a balanced one. Of the many factors weve looked at (sorry U skipped ahead now aintcha?) search engines dont rely on any single aspect. You want to try and satisfy as many as possible that make sense.
Each market and query space will have leaner or greater requirements. Understanding the existing results (competitive analysis) will be an important element to any SEO campaign.
Heres a good framework to work from;
Start up
- Identify primary target market (country, region)
- Establish secondary targets (neighbour countries, regions, secondary markets)
- Get to know the demographic (searching habits, intent)
Language
- Primary language " what is the main language in the region? (chart)
- Secondary languages " what other acceptable languages are known to the region? (chart)
- Language nuances " are there dialects? Are there any spelling/grammatical subtleties?
Website/page
- Is the domain/page URL targeted?
- Registrar data " is the business location information targeted?
- IP of web server " is it hosted locally?
- On page triggers (language, keywords, meta data, images etc..) are they present?
- Other web pages " is there location information on the site (eg Contact pages)
Links
- Location information of links " are there a high percentage of regional links pointing to the page?
- The text near the hyperlinks " are the inbound links on relevant pages?
- Links pointing out " are there regional links pointing out from the page?
It is important to remember that one doesnt need to necessarily nail down ALL the requirements in order to rank in a given region for a desired qery (term). If you look at some of the larger companies out there such as Dell, while they have country specific TLDs and contact info, the site are generally all located on servers in the US.
You simply MUST research the top ranking sites in the query space youre after. They WILL be unique....
/end journey
More stuff
Blog reading;
If youre looking to get even more into local SEO aspects, try some of the following posts;
Local search ranking factors and A Framework for Thinking About Local Search Campaigns - from David Mihm
Getting to Know Local SEO " Search Engine Watch
Using Images For Local SEO " Search Engine Land
Making Geotargeted Content Findable For the Right Searchers - Nine by Blue
Guide to geotargeting - Blog storm
Building links for geo-targeting - Search Engine Journal
On-Site Geo-Targeting and Local Search Optimization - Search Engine Journal
Local SEO predictions - Local SEO guide
SEO tutorial on geo targeting and language targeting " SEO Guru
Location Still Matters on the Web: Types of Location Information - SEO by the Sea
Google Mapping Out Its Search Results " Search Engine Journal
Search Patents;
Here are a variety of search related patents from the big 3 that are interesting reads for those looking to dig even deeper.
Google patents
Shared Geo-Located Objects
Ranking and Clustering of Geo-Located Objects
Ordering of search results based on language and/or country of the search results.
System and method for providing preferred country biasing of search results
System and method for providing preferred language ordering of search results
Indexing documents according to geographical relevance.
Scoring local search results based on location prominence
Determining unambiguous geographic references
Automatic expanded language search
Yahoo Patents
Techniques for representing location information
Contextual Mobile search based on social network vitality information
Real-time search term popularity determination, by search origin geographic location
Microsoft patents
Generalized location identification
Search query dominant location detection
Detecting a user's location, local intent and travel intent from search queries
David Harry is an IR analyst, SEO consultant and Seriously obsessed search geek that can be found on Twitter or his SEO Blog for the latest in search geeky goodiness!