Local Search Ranking Factors

Our Goal

This is the first in a five-part Bizible study where we set out to discover, through statistical analysis, the most influential local ranking factors in improving business’s rankings in Google Places.  This report focuses only on local ranking factors from a business’s Google Places page.  The next four reports will cover the other categories of ranking factors, namely:

1. On-site optimization
2. Citations
3. Reviews
4. In-bound links and off-site

If you would like specific questions answered in our next studies, use this Google+ thread to suggest and vote for topics.

There are other reports like this one, so where does this fit in?

Before our study, there was a notable hole in the search ranking analysis world. SEOMoz does an excellent job of discovering ranking factors for Google’s web results through both a statistical analysis (like ours) and a survey of the opinions of SEO experts.  David Mihm also has a great survey on the opinions of Local SEO experts (please check them both out.)    But a statistical analysis of the local ranking factors was missing.

Our Motivation

When we left Bing’s ad platform to start Bizible (about us), we looked across the local search industry for a data-driven analysis of what drives ranking changes in Google Local.  Although there is some great information out there about local SEO, no one had answered this question holistically with a data-driven approach.  As a data-driven organization, we are well-equipped to answer this key question. We hope you benefit from our work.

The Process

We surveyed 22 local business categories (i.e. photographer, hotel, etc.), 22 major U.S. cities and searched Google for the given localized term (i.e. “Seattle photographer”).  We grabbed the first 30 results in Google Local and all the local results from the integrated web results (1/3/5/7 packs, universal results, etc.) We then inspected each Places page for all of the ranking factors we could extract (see below for the complete list).  In total, we analyzed 477 queries, 14,309 individual search results, and 457,888 data points.  We removed the seven queries that did not generate integrated local results, like “Phoenix insurance.”  Note that our analysis was performed before the recent “Venice” Google search update in late February. We will run a followup study to see what has changed since then.

We looked at each ranking factor in isolation and accounted for variation in competitiveness across search terms.  We went out of our way to ensure that our methods were sound. For completeness, we performed both univariate analyses (using both non-parametric Spearman and Pearson correlation), and multiple regression analysis.  The latter is what yielded the results you find here.

We conducted the searches from 1,350 different IP addresses across multiple C classes and originating in multiple U.S. cities to ensure we received the results an average user would see.  We did not send geo information in with the queries, nor did we simulate logged in users.

The Results

There are two pieces to our results:

1. The raw statistical results for the most influential ranking factors that we set out to discover.
2. Some interesting takeaways that we uncovered along the way.

The later revealed some very interesting takeaways we did not expect.

Top Local Ranking Factors

Upon analyzing these signals, we realized that there was a sharp difference between signals associated with improved ranking for results on the main SERP page (in the integrated results) and those in the top 30 of Google Local, but not in the integrated results.  This makes sense: The things that got you to the bottom of the 7-pack are not the same things that will get you to #1.  It also makes a nice prescription for what needs to be done to optimize your Google Places page, depending on where your current ranking.

There are two components to each term we searched for:

1. The search city – “Seattle” when searching for “Seattle pizza.”
2. The search category – “pizza” when searching for “Seattle pizza.”  We also generated a list of synonyms for the search category

Factors in Integrated Results

For those pages in the integrated results, the top local ranking factors while holding all other variables constant are:

1. Having the primary category match a broader category of the search category was associated with a 1.42 improvement in rank. For example, primary category is set to “restaurant” and the search category was “pizza.”
2. Having the search category or a synonym in the business name was associated with a 0.64 improvement in rank.
3. Having the search category or a synonym in “at a glance” was associated with a 0.36 improvement in rank.
4. Having five or more Google reviews was associated with a 0.31 improvement in rank.
5. Having photos (atleast 1) was associated with a 0.25 improvement in rank.

Listings with all of these signals showed an improvement in ranking of about three positions, which is pretty high considering that on average there were five integrated local results in the main search page. Moving up three out of five positions is significant.

Factors in Pure Local Results

For those pages not in the integrated results, the top local ranking factors while holding all other variables constant are:

1. Having five or more Google reviews was associated with a 1.47 improvement in rank.
2. Having the search city in “at a glance” was associated with a 1.42 improvement in rank.
3. Having the search category or a synonym in in review content was associated with a 0.97 improvement in rank.
4. Having the search category or a synonym in the business description was associated with a 0.85 improvement in rank.
5. Having the search category or a synonym in “at a glance” was associated with a 0.85 improvement in rank.
6. Having the primary category match the search category was associated with a 0.79 improvement in rank.
7. Having the search category or a synonym in in the business name was associated with a 0.75 improvement in rank.
8. Having a secondary business category that was a broader category than the search category was associated with a 0.68 improvement in rank. i.e. secondary category is “restaurant” when searching for “Seattle pizza.”
9. Having at least one photo was associated with a 0.66 improvement in rank.
10. Owner verified was associated with a 0.52 improvement in rank.

Listings with all of these signals showed an improvement in ranking of about nine positions.  Given that they were in the top 30, an improvement of nine is significant.

Surprising Takeaways

During our analysis, we also uncovered the following:

• Having a physical address in the city of the search did not turn out to be a strong ranking factor, only distance from centroid seemed to matter.  So, if you are just outside the city and your address is not officially in the city, this didn’t seem to hurt any more than a business whose address was in the city, but just as far from the centroid.
• Not having any Google reviews or having an average review score of one hurt ranking (as expected), although the incremental increase in ranking as the review score increased from two through five was negligible.
• The presence of a business description alone did not help ranking, but having the search category in the business description did help.
• As expected, getting your fifth Google review significantly helped ranking, although incremental reviews between one and four and above five had a very small impact on ranking. You have to get 100+ reviews to again have a significant impact on ranking.
• On average, for every mile away from the centroid, ranking dropped by 0.4 of a position.  Note that this is the average across all 22 cities.  In very dense cities like New York, this number was higher and in sprawling cities it was lower.
• Seven of the 484 queries did not have local results on the main page (we removed those 7 from our analysis). On average, there were five local results on the main page.

Notes About Statistical Analysis

These results are from statistical correlation which does not imply causation.  In other words, we don’t have the source code to Google’s search engine; we are statistically analyzing it.  Also, Google’s ranking algorithm is obviously very complex and has many signals we cannot control or measure (like click through rates.)  This study is meant to help better understand the impact of the ranking factors we can control.

More Information

If you are interested in contracting us to do a custom statistical analysis of your market or want more information on our analysis, email aaron+stats@bizible.com.

If you would like specific questions answered in our next studies, use this Google+ thread to suggest and vote for topics.

Subscribe to our blogto receive updates on our future studies.

About Bizible

Our all-in-one inbound marketing platform for small business is currently in private beta.  If you would like to sign up for our beta, or are interested in our affiliate or reseller programs, please go here.

Contributors

Special thanks to Patrick Danaher at the University of Washington’s Bio Statistics department for doing the statistical analysis and Dr. Jia Zhou for reviewing it.