Archive for November, 2010

How Reviews Flow Around the Web

Click on the image for a big version, suitable for framing.

My colleague Kris wrote a fine post a couple of weeks ago on the importance of online reviews for local business, and how to get them. Following that post, we did a little research to figure out how reviews get passed around between major local search sites in the US. This graphic illustrates some of our findings. If an arrow is pointing from one site to another, it means reviews from the source site are showing up on the destination site. An arrow with a dotted line means we found very inconsistent results, i.e. reviews from the source site might show up on the destination site, but often don’t. We sampled a variety of business categories around the country in our research, but may well have missed some relationships. The number in parenthesis next to the site name is our estimate of monthly visits to the site. Some of these are taken straight from Compete.com, but some required a little more figuring. More on estimation below.

A Few Interesting Relationships

While researching, we noticed a few interesting relationships:

• AOL recently started showing Mapquest results for local category and business name searches in their search engine.
• While Google gathers review content from many sites, Citysearch, Yelp, Insider Pages and Judy’s Book tend to be more visible than other sites mentioned here.
• Insider Pages and Citysearch (both owned by IAC) almost behave like two front ends to the same source data. Most content is shared back and forth between the two, and almost all sites that show Citysearch reviews also show Insider Pages reviews.

How We Estimated Visits

We thought it would be useful to include traffic estimates in our diagram, but take them with a good sized grain of salt. Measurement of web traffic is an inexact science in any circumstance, and becomes even more difficult when you are trying to understand the local intent of searchers on a general search site like www.Google.com. For Yahoo, Bing and Google, we estimated that 10% of general search traffic has local, commercial intent. We also added 10% of AOL search traffic to Mapquest, since Mapquest results are shown for local results on AOL. You can see the logic behind using 10% in our “What Percentage of Search is Local?” post, but it’s just a ballpark. In any case, we believe industry estimates often underrepresent how much local search is happening on the big search engines.

There is also some difference between the nature of local search happening on the sites we included. For example, people going to Mapquest are less likely to be looking for a new dry cleaner, and more likely to be finding directions on a map. Conversely, people going to Yelp are a prime target for influencing – the site is all about helping people make commercial decisions. So again, take volume numbers with a grain of salt.

If We Were To Pick Three…

…that are more important than the rest, we’d say Citysearch, Yelp and Yahoo Local. Citysearch is  a slam dunk – their reviews show up on almost every site, and get prominent placement on many. Yelp is our second choice because the Yelp community is very active and reviews are all that matters on Yelp. Yelp reviews also get good placement on Google Place Pages and get syndicated to a number of smaller sites that are not on this chart. Yahoo Local is important because 3rd party reviews are ghettoized on Yahoo. You have to click the “web reviews” tab to see them, and the star rating displayed in Yahoo search results is based on Yahoo reviews.

Google is also putting more emphasis on reviews in their results. For the time being Citysearch and Yelp will have you covered in Google, but once Google gets enough review content, they may start to favor their own reviews.

Niche Sites

We looked at major local search sites serving the US market, and a few smaller properties that are influential in the review space. When encouraging reviews for your business, you should also look for sites that are specific to your vertical or location, as these can be important to your target customer. For example, Urbanspoon and Opentable are important restaurant review sites, and Tripadvisor is very important for travel. To find sites in your niche, search for your business category in Google Maps, and see where your competitors are getting reviews.

Do you have experience with web reviews? We’d love to hear your thoughts!

So, while the definition of “local” is decidedly inexact, we can infer that roughly 10-20% of queries on search engines are local, depending on how strict your definition is. And based on my analysis, approximately a third of those are related to business/commercial interest.

If you know of any other good data sources for this metric, please comment below!

With Internet trends, it can be hard to tell when hype is merited and when it isn’t. Below is my attempt to answer this question for some sites of interest to local businesses. The metric traffic-to-hype ratio measures the proportion of actual traffic to how often a site gets talked about. A high ratio means there’s more steak than sizzle, a low ratio means there’s more sizzle than steak.

So what?

I ran these numbers mostly to satisfy my own curiosity. At Two Octobers, we help businesses get found, and we like to focus on what will drive real results. Sheer volume of traffic is certainly a part of that, but there is also merit to looking for under-recognized opportunities. Anecdotally, I find that the ratio of marketers to non-marketers on Twitter is very high. Given the traffic-to-hype ratio of Twitter, that should come as no surprise. It may still be worth marketing on Twitter, but you will have to work pretty hard to stand out. Yelp, on the other hand, isn’t getting a lot of hype but it is getting decent usage, so you should be able to stand out with less effort. And the very unsexy Internet yellow pages also get good traffic and very little hype.

I also think the YouTube number is interesting – YouTube gets a lot of hype, but it also gets A LOT of usage. I’m going to be thinking more about how to leverage YouTube for local businesses after writing this post.

Where does the data come from?

The visits numbers come from Compete.com September 2010 U.S. visit statistics. Search results are based on a Google query for the site name without the “.com”, excluding results on the site itself. The search results were also filtered to only show pages that were indexed in September, 2010.

example of Google search query

The traffic-to-hype ratio is simply the ratio of the two. I recognize that there are some flaws to this methodology, for example most of these sites also have mobile applications that aren’t represented in the Compete statistics. Twitter and Foursquare in particular are affected by this problem. But all have web interfaces that account for a significant percentage of usage and I don’t know of a good source to account for application visits. If you do, please let me know!

And I would love to hear your thoughts on this topic in the comments below.