How Business Names Might be Used by Google in Local Search Ranking Signals

I love local search. In many ways, it’s similar to Google’s Web Search, but with its own unique features. In addition to Googlebot, Local search has street view cars. In addition to looking at links, local search also looks for mentions of businesses that appear with location-based information. Instead of robots.txt files, local search is stopped by signs like “military base,” or “private street.”

I also appreciate the local search ranking factors that a good number of people who are involved in local search have been putting together every year lately, but I’m also a little apprehensive about those, and I’m going to illustrate why with this post. Imagine for instance that Google considers the names of businesses in the way that it ranks them in local search, but that it doesn’t treat every name the same. For example, “Frost Diner” might be treated one way by Google Local Search.

Frost Diner in Warrenton, Virginia

And, because it has a somewhat longer name, “Red Truck Bakery” might be treated differently by the algorithms that use business names as a ranking signal by Google’s Local Search:

The Red Truck Bakery in Warrenton, Virginia

Google has come out pretty clearly and told us that Local Search rankings are primarily based upon relevance, prominence, and distance.

So how does the length of the name of a business fit into local search rankings?

A Google patent granted last week tells us about how they might use a “webscore” based upon the amount of “estimated” search results that are returned on a search for the name of a business. But, if the name of the business is on the shorter side, such as less than a certain number of characters (for example, 10 or less), or less than a certain number of words (such as 3), that web search might not just include the name of the business but also the location name as well.

So Google might take the shorter “Frost Diner” and add “Warrenton Virginia” to it, to create a query of [frost diner warrenton virginia], which returns 16,500 (estimated) results

For the longer business name, Google may just use it in a query, such as [Red Truck Bakery], to see that there are 602,000 (estimated) results for that query.

A search for [frost diner] without the city and state name included in the query ends up with 3,340,000 (estimated) results.

As an aside, this is the first patent I can recall seeing from Google that includes the number of estimated results returned for a query as part of a ranking score.

The patent is:

Title based local search ranking
Invented by Jiang Qian, Ben Luk, Xinghua An
Assigned to Google
US Patent 8,122,013
Granted February 21, 2012
Filed: January 27, 2006

Abstract

A method for performing a local search includes receiving a local search request that includes at least a search term and a geographic identification. Business listings matching the received local search request are identified. The business listings are then ranked based on at least a webscore associated with each listing. Each listing’s webscore is based on the listing’s web popularity. In this manner, local search listings are ranked and presented in a more accurate manner.

The patent tells us that other features might be to this webscore or or to a location prominence score for a business:

  • Review scores or the sources of reviews associated a business listing might increase or decrease a location prominence score or web score.
  • Language within a review for a business may also increase or decrease a location prominence score or web score.
  • Financial information for a business, such as annual sales, employment base, longevity, etc. may also influence those scores.

Redundant Business Names

Because there is only one “Frost Diner” and one “Red Truck Bakery” in the area, they may also get treated differently than “McDonalds” does as well.

The local McDonalds in Warrenton, Virginia

So imagine that I do a search to find lunch in the area, and McDonalds is a possible option.

Since McDonalds is so short a business name, the query used as part of the webscore might be expanded to [mcdonalds warrenton virginia], and that web search gives us an estimated result count of 1,690,000. But, there are two McDonalds in the area considered, so that number might be reduced by half. The patent tells us that where there are x business listings in an area that have the same business name, that the webscore for each of the businesses in that area might be 1/x. In my local example, that would mean that each McDonalds would have a webscore of 1,690,000/2, or 845,000. If there were 10 McDonalds in the area, it would be one tenth of the raw score, or 169,000.

Location Prominence

I linked to a post I wrote about a Google patent on location prominence above, but this patent provides a list of factors that might be considered in determining a location prominence score as well:

A location prominence score might be a combination of factors such as:

  • A search ranking value for an authority page associated with the business listing
  • A highest search ranking value for any page referencing the listing address
  • The number of pages referencing the listing address
  • The number of scraped page references (listings at places like Citysearch and Superpages)
  • The number of reviews for the listing
  • A scaled webscore for the listing using a method like that described above.

Takeaways

Of course, Google may not be using this particular algorithm, but then again they might be, or may be using something similar to it.

This patent seems to be based upon an assumption that the longer a business name is, the more unique it might be, and the more likely that many or most of the web search results returned on a search for the name are going to be about that specific business.

As for business chains, I’m wondering if it’s helpful for them to all share the same name, such as McDonalds, or to give them somewhat longer and more unique names, such The Hilton Manhattan East.

After reading this patent, I’m probably going to be paying a lot more attention to the lengths of business names that I see in search results.

Share

37 thoughts on “How Business Names Might be Used by Google in Local Search Ranking Signals”

  1. Interesting to see that they state to incorporate language properties from review text to determine the prominence of a business. Does that mean that they filter out spam reviews?

  2. Another point that favors smb’s doing a lot of offline advertising (TV, etc) to increase the amount of brand searches. Gut and data from clients has been saying this has been being used for some time.

  3. Google are cutting down on locations and now categories added to registered or prevalent business names from the real world.

  4. Jan – Google is getting pretty good at filtering out spam reviews. The only issue is that they also tend to filter out perfectly legit reviews…

  5. Interesting article. There is also something to be said about Google’s new “offers” page, where they advertise deals that are being held by local businesses near your zip code. That and local search provide two different ways to get recognized locally.

  6. Matt G – Do you have some more information about the criteria which Google uses to determine if a review is fake? Does it take in account how active a user is on the internet for example?

  7. When you say an “authority page associated with the business listing” I’m not sure what you mean. Can you clarify that point? I’m a little dense!
    Otherwise, I think it is safe to view these patents as strong indicators of how G evaluates websites and provides good validation for get listed in a lot of local directories as a part of an SEO strategy.
    Thanks,
    Kris

  8. If a business is in a small town, and optimizes for a nearby larger city in hopes to earn more business they might want to try and get reviews from customers in the target cities. Sometimes people are willing to drive farther if it means better customer service.

  9. Interesting article Bill. But one thing is a little confusing which is whether unique business names will get more priority over common business names ? Could you please throw some more light on this. Thanks.

  10. Nice as always before.

    But as you mentioned Local Search rankings are primarily based upon relevance, prominence, and distance. And that’s completely true, but here might be one more factors needs to be come in play to increase or decrease Local search listing that is “Listing popularity”.

    Like if visitors are more feeling good to click/visit specific listing & rising click through rate for that listing. So definitely that particle local listing will stand in top of search results.

    Thanks.

  11. I believe what actually impacts, or should impact, or what Google will have impact the local listing results in the near future is not only the Clickthrough rate like Rajesh above me mentioned, but actually study the clickthrough rates/reviews ratio. Meaning if Listing A gets averages 30 clicks a day but 0 reviews a day and then Listing D gets only 10 clicks a day but averages 2 reviews a day, it really means Listing D is actually providing the service that people were searching for, hence he deserves to go up on ranking order because at the end of the day, that’s all Google wants, to provide accurate results!

  12. I always assumed local search worked by figuring out your location from your IP address and showing search results based on that – or GPS data, in the case of smartphones. Aren’t all modern browsers location-aware these days? Surely the physical location of these businesses in relation to the person doing the search would take precedence over the pagerank.

  13. Hi Jan,

    I suspect that the intent behind that language isn’t so much on “spam” reviews as it is on considering the language within reviews, and possibly even the sentiment of reviews, when generating a score. The patent really isn’t so much about reviews and filtering them, so its focus isn’t there, and it doesn’t really discuss reviews much. The patent is also originally from 2006, and it’s possible that they didn’t have enough experience with review spam to have spend too much energy thinking about it.

  14. Hi Jacob,

    Being involved in your local community can play a role as well, from a listing on your local chamber of commerce website, to news about you and your business (mentioning its location) in the local newspaper, to your hosting community events, and more. Those types of things show that you are involved in your community, your business is one that is known there, and are indications of the prominence of your business in that area.

  15. Hi Andrew,

    Not sure what you mean. I do know that Google doesn’t want to list businesses too prominently that don’t have a location that people can visit in person, or a specific area that they provide services to people within. They also don’t want people adding locations where they aren’t actually located, and might only have a post box address at, instead of an actual location.

  16. Hi Neal,

    In Google’s recent post on their Inside Search blog about 40 new algorithmic changes in February, they also mentioned that they made a change in their algorithms to show local results more often in cases where they probably should have been, but hadn’t. Not sure what the impact of that has been, but it definitely seems that they are aiming at showing local results when there seems to be an intent on the behalf of searchers.

  17. Hi Kris,

    Google shows Maps or Place pages for many businesses without the owners of those businesses verifying the information contained within them, and Google will often determine which website should be associated with a particular business when they do that. That website is one that they might consider to be the “authority” website for that business. I wrote about that kind of association back in 2006, and it appears that the kinds of things they were considering back then are very similar or the same as what they are looking at now. Here’s a link to the post:

    Authority Documents for Google’s Local Search

  18. Hi Scott,

    It probably wouldn’t be a bad idea to try to get reviews from the nearby larger city, and even do other things like provide driving directions from that city to the location of the business on the website of the business as well.

  19. Hi Aaron,

    I’ve been thinking about that, and I think sometimes having a unique business name might help, and sometimes it might hurt.

    Under the approach outlined in the patent, it doesn’t seem to matter if any, many, or all of the search results that show up are actually about your business. So, for example, you might have a long and unique business name that is composed of somewhat generic terms that are somewhat common, and may appear on the same page but are rarely used together as a name of a business, and the total number of search results might actually be fairly high for your business even though most of those pages aren’t about your business. On the other hand, you might have a unique business name that uses somewhat rare terms, and the amount of search results might be much smaller, resulting in a smaller webscore for that business name. Chances are though, that more of those result might actually be about your business.

    So, naming your breakfast diner the “To be or not to be cafe” could result in a pretty big web score under this approach. How many of those might you find though? If I put quotes around the name, not too many. :)

    I suspect that someone at Google might have tweaked this in some ways if they are using it. What I find interesting about it is how Google might consider something like a business name as a ranking signal, but apply interesting twists to it, such as adding location information when it might be under a certain size.

  20. Hi Rajesh,

    It definitely is possible that Google also incorporates other signals into how it rates listings as well. And it might add some twists to doing that, too.

    For example if three businesses are displayed as place results within Google’s web search results for a particular web search query, and people tend to click on the “show more local results,” instead of one of those, and then click on the fifth or sixth business listed in Google Maps, then maybe that listing ends up getting boosted in search results over time.

  21. Hi Khaled,

    I’m not sure how helpful a negative signal like a lack of reviews might be in many cases. There are many business types, for example, that people tend not to leave reviews for, like insurance companies for instance. But the patent does mention that it might consider reviews in rankings as well, so it’s possible that they might look at the kind of thing that you mention.

  22. Hi Alex,

    Your location might influence what local search results might be blended in hen you search the Web, and might be impacted by information about your actual location, the location that you might enter into Google as your preferred location, or whether or not you included location information within your query.

    Your location might also influence the results that you are shown when you search in Google Maps, and haven’t included geographical information in your query, and again that might be based upon your actual information if you’re searching from a mobile device, or a preferred location if you’ve set one with Google.

    If you’re going on a trip to somewhere else, it’s not a bad idea to include some geographic location in your query or change your location preference with Google, and relying upon IP or GPS or cell tower triangulation won’t be of much help for geographically related queries then.

    PageRank may place a small role in how something like location prominence influences rankings, but as representatives from Google have noted a few times, the primary signals tend to be distance, relevance, and location prominence.

  23. This thought process behind business names in local seo sounds like it is on the right track. It obviously has some bugs but the thought process is intriguing.

  24. Hi Charles,

    There are some chains, like hotels, that do attempt to give their different locations unique names, even if it’s just appending the location to the name. That makes sense from a practical standpoint as well as possibly helping when it comes to online business directories like Google Maps as well. Ask a cabbie in New York City to take you to the Hilton, for instance, and he’s going to ask you which one.

  25. Hi Johnathan,

    It was pretty exciting to see how Google might take a signal like a business name, and apply different rules to it based upon something like the length of the name. I suspect that Google could be doing something a little more sophisticated than this now based upon experimentation and testing, but I think it’s something that people trying to understand ranking signals in local search, and other types of search have to keep in mind – different factors and signals that Google might use may have some unexpected aspects to how they might be determined and used that can be very hard to guess.

  26. Great post, I’m trying to improve my understanding of local search and how to make the most of it, particularly relevant for mobile search of course.

  27. Hi Dave,

    Thank you. Local search and mobile search seem made for each other. Even Apple seems to be pretty serious getting in on the interaction between the two.

  28. Hi Bill,
    Just one doubt , pardon me if I am being noob , will this change affect the small businesses in a positive way?

  29. Hi Neil

    I have no idea. It’s possible that Google has been using something like this for years. It’s possible that they never used it. It’s possible that it’s something they might start using tomorrow.

    I would expect that they would test results that might go through a process like this, but I don’t think Google does an analysis of whether a certain algorithm benefits small businesses or big brands. The purpose is to try to help people find certain types of businesses at certain locations, rather than biasing results to favor some types of businesses over others.

  30. Pingback: Changes to Ranking Factors on Google Places: What You Should Know | Webologist
  31. Bit late stumbling on this Bill but very interesting post. I had to read it three times to fully grasp it to be honest, it’s late! (or maybe I’m just a little thick!). I still can’t decide whether or not it would be better to have a long unique name or a short name using generic words that have multiple meanings.

    As if the Google Places algo wasn’t vague enough, they throw this little chilli into the pot! I’ll definitely be taking a lot more notice of business names from now on though.

  32. Hi Sean,

    I wrote the post, and that’s truly one of the questions that I have about the patent myself. As I mentioned towards the end of the post:

    This patent seems to be based upon an assumption that the longer a business name is, the more unique it might be, and the more likely that many or most of the web search results returned on a search for the name are going to be about that specific business.

    I’m not sure that’s true.

    My thoughts are that it best having a unique name, regardless of the length, and to have great marketing that gets that name out into the public and on the Web, so that when it’s mentioned, it’s likely that the people doing that mentioning are referring to that particular business.

    What I wanted the most to point to with this post wasn’t so much about how Google might be ranking businesses based upon name length, but more the folly of coming up with a list of ranking signals and trying to rank them, when we don’t have a clue as to what kinds of twists and tricks Google might include in how they use those signals.

Comments are closed.