Google Local Search Stats for Black Friday

As promised in our post about the impacts of the October 27th Google update, we have put together a quick look at the traffic trends we saw for retails outlets on Black Friday and Cyber Monday.

About The Data

  • Sample (hundreds) of retail chain locations across the United States (other verticals responded differently and this analysis focuses only on retail stores)
  • Click and ‘click action’ data from Google Places for November 12th through December 12th
  • All locations have clean, complete and verified listings in Google Places

Our first step was to take a look at the number of clicks generated per day across the sample locations.  We took this data and indexed the performance to the click totals on November 12th:

bf-vs-cm1

Few things standout on this graph:

  • Not all that surprisingly, Black Friday was the best day in terms of the number of clicks generated
  • Local traffic popped on Cyber Monday as well
  • Big slowdown going into the Thanksgiving holiday
  • Black Friday is just a launching pad for the season with local traffic accelerating strongly starting in early December

Now we wanted to understand a bit more about the Black Friday burst versus the Cyber Monday burst – specifically if Black Friday is truly a ‘physical’ and in-store event while Cyber Monday is an online event.  To do this, we simply looked at the click actions (as reported by Google) for clicks to the location’s website versus clicks for directions to the location.  We used this data as a metric to help understand what people’s intent was once they found the local listing:

bf-vs-cm-type

The data showed a material difference in the click behavior with a full 10% of user clicks shifting from directions to website on Cyber Monday.  While this data is not perfect (for example, many people going to the website just ended up using the on-site functionality to get directions) is does provide a reasonable proxy for the behavior of shoppers on these two days.

In early 2011 we will be posting an analysis of the entire 2010 holiday shopping season.  Follow us on Twitter or Facebook to get the latest updates as they become available.  Thanks!

Google Local Updates: Performance Impact Analysis

Right around October 27th Google rolled out an update to both how they rank their local listings and how they integrate those local listings into the main search result pages.  While the 3-pack and 7-pack still exist, this layout is starting to become the norm for local listings on the main Google results page:

blog-google

To help understand what these changes mean to local businesses listed on Google, we have put together an analytical summary of what we saw during and after this update.

About The Data

  • Sample of retail chain locations across the United States (other verticals responded differently to the change and these number apply only to retail stores)
  • Click and Impression data for October 18th through November 14th, as reported in Google Places
  • Removed outliers based on total number of clicks and/or impressions
  • All the locations have clean, complete and verified listings in Google Places

Summary of Results

  • Significant CTR improvements
  • Some of the CTR improvement is driven by increased impressions for branded traffic
  • The number of reviews for a location seems to have impacted overall traffic growth for the location
  • The average review rating does not seem to be connected to overall traffic

Details and Graphs

The impact of the change is clear when looking at the clicks and impressions:

blog-avg-perf1

Clicks grew over 50% in the ~2 weeks following the update while impressions grew only ~15% — implying that a majority of the click benefit came from increased CTR.  This leads to the question of “did the query stream change?”.  To help answer that, we took a look at the growth in clicks and the growth in the percentage of branded traffic:

blog-perf-branded

While the update did drive a noticeable and material change in the percentage of branded traffic (up over 10%), that change alone did not account for a majority of the increase in overall traffic to the locations.

So, good clean listings seem to have benefited from the update…but how do reviews play into the performance equation?  To help answer this we took a look at the click growth for locations that had received 3 or more reviews in the last 6 months versus those that have received 2 or less:

blog-perf-num-reviews

The locations with 3 or more reviews performed better after the update.  Now, this does not mean more reviews = better ranking.  It could simply be that more reviews is just a good proxy for the overall number of citations a location has.  That said, this data does provide additional evidence that the more times Google can find matching information about your business location on the web the better it is for your ranking.

With the recent discussions around how rating could/should/does impact ranking, we also took a look at the click growth based on the average rating for the sample locations.  We split the locations into two buckets, one for those with a rating over 3.5 and the other for those with rating of 3.5 or lower.  Each bucket had an average right around 5.5 reviews per locations (to help deal with the impacts that total number of reviews could have on the analysis).  Here is what we saw:

blog-perf-rating

Based on this data, the poorly rated locations actually saw better click growth than the strongly rated locations.  While not definitive, it certainly points to the average rating being ignored by Google at this point (but not by your potential customers!).  And this actually makes a lot of sense given the issues we have seen with Google importing 3rd party reviews correctly (which we will take a look at another time) and the lack of credibility of the ratings/reviews on many 3rd party sites.  However, as Google (tries) to gather it’s own rating data via Hotpot they will be able to build a data set that they can trust — and can then start to use more actively in the ranking process.

If you are still reading…thanks and we will be posting an update on the holiday shopping traffic in the near future.