Taking Flight with Brand Share


By Jessica Carneiro, Senior Director of Analytics & Insights

One of the benefits of being a company with a multi-petabyte scale data operation is that we have an unparalleled view into actual consumer behavior. Using data we buy from third parties and those that we collect internally, we can look at the actual URL string of sites that consumers visit (anonymously of course) and we thought it would be interesting to compare how often major brand names showed up in these URLs. For our purposes, we’ll call it brand share.

For this first comparison, we looked at seven airlines over 11 weeks and saw some fascinating results.  In a mature industry like air travel, market share changes slowly, so we expected brand share to be similar. In fact, it appears to be exceptionally volatile.  Within individual airlines, brand share varied as much as 51% between its high and low. So what caused such changes? News. Especially negative news.

airline brand share dec 2014
Five of the seven airlines tracked had notable spikes that correlated with major news.

In the last few months, the top three spikes in brand share correlate to the following stories:

  1. Flight attendant refuses to hang Army Ranger’s jacket (US Airways)
  2. Ebola passenger on flight (American)
  3. Plane makes emergency landing (Southwest)

Each of these events, in and of itself, was sure to make news. But it’s interesting to note the relative spikes. The general public apparently cares more about a member of the armed force being mistreated than an international health scare or a major flight mishap.  American Airlines captured 57% brand share the week that the Army Ranger story made the news, compared to an average of 10% brand share for other weeks.  That’s nearly a 6x increase in brand share from this one story alone.

We also found that promotions can drive brand awareness as it did for Virgin America when they ran an offer on Living Social. But promotions aren’t as newsworthy as celebrity-related news stories, like when Jose Conseco posted negative tweets about Delta Airlines before he accidentally shot his own finger.

Another observation: The public moves on, but not as quickly as you might imagine. In today’s 24-hour news cycle, it surprised me to see that it took more than two weeks for American to return to their pre-controversy brand share of ~8%.

What can we learn here?  Brands can’t control events that may drive large shifts in the public awareness of their brand. But they can better monitor these trends to take advantage of the times when their brand share does change (for good or bad reason) and manage their brand image.

There are existing tools to measure brand buzz like Google Trends or the many tools analyzing Twitter feeds. But the drawback to these tools is that they only capture a small segment of consumers who perform certain activities – search for the brand or tweet about the brand.

Analyzing pages that consumers visit online is a much broader way to measure brand buzz. This type of analysis captures higher level brand awareness because a person may read some news story about a brand but never search or tweet.