carson smith’s blog

More unshaven people ordering McDonalds? Online fantasies about Louis Vuitton handbags? These are just a few inferences one might take away from my rather unscientific investigation into the performance of top brand names in Google search.

I started my research with data from Millward Brown (MB), a major brand marketing consultancy. Each year, they attempt to determine the top 100 worldwide brands. Using combination of hard and soft data sources from research companies, monetary values are calculated. So are the percentage changes from the previous year. You can view the 2009 report here.

As a web marketing analyst, I was particularly interested in how the top brands performed online.  I wanted to:

• Discover which brands people are more popular in Google searches compared to last year.
• Determine any correlation with MB’s reported changes in brand value between 08 and 09.
• Interpret the data in context of the recession.

After some basic research of MB’s top 100 brands using Google Insights for Search, I came up with my own list of the biggest winners and losers over the past 12 months. A positive % means more people are typing the term into Google compared to last year. I contrasted these Google search volume changes with Millward Brown’s % changes in brand value, which are calculated using more traditional sources.

With the exception of Nissan, the only discrepancies between positive and negative % changes in the Google searches and MB data sets are the American bank brands. In some cases, this makes sense because many banks underwent mergers. Chase, for example, swallowed up Washington Mutual in 2008. Chase’s (blue) search volume saw a marked increase as WaMu (red) customers started searching for their new bank brand. The defunct WaMu brand plunged. See graph below:

Over half of the best-performing brands on Google between 08 and 09 were foreign. Again, this makes some sense considering the recession has affected domestic companies most severely.

Another interesting thing to look at are the luxury brands. Surprisingly, Louis Vuitton and GUCCI saw some of the largest increases in online interest this past year. Whether this was due to increased marketing spend or escapism from economic hardship, it’s clear people are not about to start organic farms in their back yard just yet.

Now let’s take a look at the 20 brands who saw the biggest declines in online interest between 2008 and 2009:

Here, the relationship between Google and MB data is not as strong. For example, IBM saw a 20 percent decline in search volume, yet according to MB, their brand value increased by 20%. Both data sets agree on Porsche, whose attempt to take over VW seems to have had adversely affected them. And fewer people are searching for Gillette. After all, who needs to shave when you’re unemployed?!

It is also interesting to note here that several major American banks did see declines in search volume, in accordance with MB’s brand value percentages. There are likely reasons other than brand value for these changes for the correlations.

How strong are the correlations?

To determine the relationship between Google search and MB brand value data, I plotted the percent changes in Excel and generated an R squared value. R squared is a value from -1 (perfectly negative relationship) to +1 (perfectly positive relationship). And 0 means there is no relationship at all.

In an effort to minimize anomalies attributable to factors other than brand value, I removed all the American bank brands from the data set. Below is a scatter plot showing the relationship between Google search data and MB data:

With an R squared value of 0.14, there is a weak positive correlation between two data sets. This does not definitively mean that Google search volume is an indicator of brand value, but simply suggests a minor relationship is possible. Given that MB’s data is a rough estimate of something that’s intangible (and disagrees on many points with a similar study by another agency called Interbrand), a weak correlation is to be expected.

Rather than providing answers, I hope this little study provokes some questions about the online behavior and brand value. More people than ever are searching for products and services online. Identifying changes in keyword search patterns offers insight into consumers’ interests unlike any other medium. While the relationship shown here may be small, there is no doubt companies concerned with their brand value should be paying attention to Google searches.

If not predicting the impending collapse of our economy one day, the media is forecasting terrorist attacks the next. The press loves to depress. “How ’bout ya quit makin’ things up?” asked a frustrated Sarah Palin upon resigning last week.

Harsh. How ‘bout we all cool off and step back for a sec?

What if we could zoom out and look at all this clatter and chatter in perspective? With the internet we can. Every hour of every day, Google’s non-partisan computer code aggregates all the news article published online – all around the world.

One great feature Google offers is called “Google Trends,” a visualization of keyword frequency in both search and news articles. I want to show you Google’s monthly news reference volume for two of the biggest words in news: “economy” (in blue) and “Iraq” (in red) since the beginning of 2007.

“Iraq” has been steadily decreasing for the last three years. Meanwhile, the “economy” grew slightly through mid 2008, leaping passed Iraq as the crisis hit. It grew very quickly, thousands upon thousands of new articles crawled by Google each day as the stock market tanked. You’ll notice that just before 2009, the economy took a big dip in news coverage. This is probably due to an emphasis on the election, as it quickly climbed back up after Christmas.

But the overall trend for both the economy and Iraq since October 2008 is a steady, albeit slow decline. Let me emphasize that this is not  a reflection of sentiment towards the economy or Iraq, it is simply a report of news mentions.

So when you read those fire and brimstone articles about the collapse of the economy and chaos in Iraq, take a step back and look at the overall news volume reference. If bad-news-hungry journalists are chattering about it less and less each month, maybe things are getting better?

Let’s hope so.

PS, here are Sarah Palin’s news mentions to-date:

According to recently leaked internal memos, Twitter hopes to become the “pulse of the planet” with a billion users in 5 years. With 25 million visitors a month today, Twitter’s vital signs are already looking pretty good. What can we learn about the users who keep it pumping?

I decided it would be fun to see what personality types characterize its “Tweeters.” According to my very unscientific research this week using TweepSearch, there are a total of 903 users who self-report a Myers-Briggs* personality type in their Twitter bio. By entering this information into Google Fusion Tables, I sorted and aggregated the results.

Let’s start with the big question: are Twitter users more likely to be introverted (I) or extroverted (E)?

Result:  63% report they’re introverted. This compares to a general population that, according to MBTI, is about 51% introverted.

So what does this say about introverts? Perhaps “introverted” does not have anything to do with being social in the traditional sense? It may simply indicate introverts prefer a different medium (online) than their extroverted counterparts (face-to-face).  As new online media tools change how we communicate, perhaps they will give a greater share of the voice to introverts?

In addition to introvert vs. extrovert, Myers-Briggs also tells us about three other personality dimensions: sensing vs. intution, thinking vs. feeling, and judging vs. perceiving.

How do  these types come out on Twitter?

Sensing (S) vs. Intuition (N)

Apparently, Twitter users are much more intuitive than the general population. This means they form their impressions about people and things less on tangible evidence than on their own inner perceptions. Why would Twitter attract vastly more intuitive users? Perhaps it is because Twitter is not based in reality – it’s a virtual world where there are few tangible “senses” to grasp onto?

Thinking (T) vs. Feeling (F)

Twitter users are more likely to be thinkers than feelers. As thinkers tend to be more concerned with impersonal facts than do people-focused feelers, their majority presence on Twitter indicates that micro-blogging attracts more users who are interested in sharing information rather than social bonding. As Twitter is very scant on personal information and profiles and focuses almost entirely on the content of the Tweets, this makes some sense.

Judging (J) vs. Perceiving (P)

With a majority of perceiving types, Twitter users have slightly more flexible lifestyles than the judging-skewed general population, who tend to be more structured and care about order and punctuality more. Twitter is indeed a free-flowing medium, and could possibly frustrate someone who likes more structure.

Of the 16 possible types, which one wins overall?

To my surprise, the relatively uncommon INTJ personality type came out on top with a 17.5% share, compared with 2.1% presence in the general population. In fact, the top four personality types on Twitter are all very uncommon when compared to the statistical average across the board. And by contrast, the three least popular personality types on Twitter seem to be much more common in the overall population.

If Twitter users are indeed a unique subset of the population, then what does this say about its mass appeal? Do certain personality types drive trends more than others? If investigated further and with more scientific rigor, I think this could have huge implications for start-up companies as well as the marketing/media industry.

*Myers-Briggs is a personality theory based on Carl Jung’s work, Psychological Types. The assessment measures how people interface with the world around them. And it does this through 4 dimensions with two possible outcomes each:

Introverted vs. Extroverted

Sensing vs. Intuition

Thinking vs. Feeling

Judging vs. Perceiving

So, if I were extroverted, sensing, thinking and judging, my type would be ESTJ. The opposite of that would be an INFP. To learn more about Myers-Briggs, I recommend checking out this link.

To discover your own Myers-Briggs type, try the MyType Facebook application.