One of the most important goals for Google is being as relevant as possible to the searcher. One of the main aspects in this lies in the area in which now former Googler Marissa Mayer has been working in for the past few years: local search. Here Google is trying hard to personalize the results based on where you are located. After all, if you are living in London and you do a search on ‘Carpenter’, what use would it be for you if a carpenter from Manchester would show up in the number one results. Or even closer, one on the other side of town, while there is a perfectly good one, just as well optimized, right down the block?
For many SEOs and webmasters this is a battle they have been fighting for a few years with all the tools they can find. Some SEOs or webmasters however might not even realize their website is ranking high in one town, but is a lot less findable in another. This issue we call “Geo-Ranking”.
SEO tool Linkdex, which has been expanding a lot in the last year, now comes out with a new feature which helps web masters understand this exact issue. Linkdex launched this last week and has released a whitepaper pointing out exactly this problem and the solution.
Let’s take a look at this issue first before we talk about Linkdex’ solution. It is an issue which as said has become an increasingly bigger problem over the past few years with Google making their results more and more personal and local.
It’s important to note that we are talking about a lot more here than searching in different local Google directories, like the difference between the Dutch Google and the English Google or within the same language the difference between Google UK and Google Australia for example. What we are talking about here is not just that but also the difference between cities within the countries or even more specific: between postcodes.
In their whitepaper Linkdex shows the example of a search for ‘Carpenters’ in the UK. They looked at where the site “mybuilder.com” was ranking. Here we see that there are major differences between locations around the UK:
- ??????????????????????????Ranks number 5 in London
- Ranks number 12 in Manchester
- Ranks number 10 in Birmingham
- Ranks number 12 in Edinburgh
- Ranks number 14 in Torquay
and even postcodes show different rankings: number 16 in BA1 5ET and number 15 in SO23 7EA.
To act on this the first thing you need to know is these exact differences. If you know these differences in rankings you can figure out what you can do to make the difference in these areas. Matt Cutts actually commented on what you can do on these regional differences in a recent interview with Eric Enge:
“Imagine 4 gyms in the same small city all offering exactly the same advice. Even before you get to what search engines think, users aren’t going to understand what the difference is between these 4 places. As a user, after reading your content, why would I pick one over the other? For search engines, it’s the same challenge.”
The issue is under scribed by some industry peers amongst which our very own Sam Crocker, who was asked for the Linkdex whitepaper how he thinks Geo-Ranking insights will affect his business. Part of his answer:
“I feel Geo-Ranking intelligence will affect our business in varying ways depending upon client needs.The first and most obvious point is that rankings that vary by location make reporting on the value of SEO more difficult and this is magnified by personalized search results. Without the ability to look at highly location sensitive keywords from multiple locations rankings, as a proxy KPI, become even less valuable and comprehension for clients will be an increasingly difficult issue.”
Next to Sam Crocker the whitepaper shows the opinions of people like Nick Wilsdon, Andrew Girdwood (“It is very important that businesses understand that Geo-Ranking does happen because all businesses should know about Search”), David Freeman, David Harling (Razorfish) all explaining how they are using Geo-Ranking and all giving their top tips on geo-optimization.
The whitepaper has a foreword of Kristjan Mar Hauksson, co-author of the book “Global Search Engine Marketing”. He states:
“For me as an old school search marketer and redundant “tool” guy, I found the urge to look into the new Linkdex addition further and believe me, I was blown away by the possibilities it presented.”
Research by Linkdex: the Geo-Variance score
To figure out the size of this issue Linkdex did extensive research into what types of keyword rankings vary most by location and by how much. With that they created a segmented dataset of 2,000 keywords and phrases in 10 UK locations, made up of a mixture of business sectors. They then “applied an algorithm based on ‘Pooled Variance’, a measure of how spread out the data is, to the rankings of every URL that ranks for the keyword anywhere in the 10 locations.“
Linkdex then creates the “Geo-Variance score”, a score between 1 and 10. They found that on many keywords there are a lot of differences. See below some example data (more in the whitepaper):
The research showed some more interesting findings amongst which:
“67% of the time, if you rank in the top 30 for a keyword in one location you will NOT rank for that keyword across all other locations.”
“When location is not set, the average keyword ranking deviates by 4 positions compared to the locations specific ranking.”
One of the biggest eye openers maybe that of the 2,000 keywords researched, none of them actually had the same position geographically.
The whitepaper also shows how Geo-Variance keywords trigger Google Places. From the whitepaper:
“Keywords that trigger Google Places results have high Geo-Variance scores while most keywords with steady rankings, i.e. those with low Geo-Variance scores, don’t trigger Google Places results.”
Linkdex did further extensive research on Geo-Variability by Market Sector. In the whitepaper they looked at for example “Restaurants and Cafés”, in which it showed that even bigger sites like Tripadvisor are sensitive to these changes and are not ranking top on every location tested. They also looked at “Tradesmen” (lots of Google Places there) and “Professional Services” where the greatest geo-variability of any vertical they reviewed was found throughout positions 4 – 10. Other areas you can find information on in the whitepaper is Top Brands, Furniture, Used Cars, Flights searches, Hotels, Poker, and Finance.
All the research showed there are significant changes to be spotted, some verticals showing more differences than others.
“In each vertical, we found high value terms with significant local differences in performance that if optimized, would drive more revenue.”
The Linkdex solution
To figure out the differences between these areas we need Geo-Ranking information. Data which shows us exactly how “we” rank in these areas. You could do different searches in different tld’s, or use the tool we described a few weeks ago, but that won’t be sufficient because it will only show you the differences in countries. You could then off course try and travel to the cities to do the searches there yourself, or ask people who live there (probably easier 😉 ).
From the whitepaper:
“Most ranking position services only conduct searches from single location data centers and those few suppliers that allow you to specify the location struggle to deliver the scale required by brands and agencies.This means that ranking reports are often inaccurate, which can negatively impact a business’ market awareness and investment decisions.”
Linkdex now offers in their tool the options to track Geo-Rankings for any location – chosen by the user at zip/postal code, town, city and country levels – for any number of keyword phrases. It also offers you to check your competitors.
The whitepaper which I’m talking about above can be obtained at Linkdex, get in touch with them for a copy.