The Basement

So what can blogs reveal about brand strength? Before I take a stab at that, I need to take a step back...

If you're a Creating Passionate Users zealot, and I am, then you know how to spot a passionate user. If not, you can read all about it in Reverse-Engineering Passion. In this article Kathy outlines the seven components of passion. I think that in order to blog about something – good or bad - you automatically meet four of seven of the passion requirements: Evangelize, Connect, Show Off and Spend Time. Based on the notion that bloggers are passionate about their subject matter, one might infer the converse, namely that the level of “buzz” in the blogosphere about a particular product, service or company is loosely representative of the level of consumer passion. I think there is some significant merit to this line of thought.

What about the casual mention? The “I looked up this or that on Google” or “I was just hanging out listening to my iPod...” statements that float around a blogger's primary point. Is it valid to count these? If a brand is not the primary focus does its incidental mention still indicate passion? I say not just yes, but hell yes! These casual mentions make brands inseparable from consumer identity and they are much more credible and sincere that straight-on evangelism. Once a product or service becomes a verb such as “Google” or a generic product category like “Band-Aid”, “Xerox” or “iPod” then the race is over. Invoking the use of a specific brand name deeply ingrained in conversation, be it in a blog or at a flea market, requires extra effort, it's showing off, it's trolling for a connection ... it's subversive evangelism.

Before we jump in, a word of warning -while buzz level may be an accurate indicator of passion, it may or may not be a particularly accurate measure for market share, adoption rate or even whether the effects of the buzz are positive or negative. Since we are using lexical analysis of free text in a really big text space, we also need to take some reasonable steps to ensure that we're measuring the right brand. A simple search for “apple”, for example, gets all the mentions of the fruit as well as folks in Cupertino.

Let's start by looking at two big technology brands: Apple and Microsoft. First, a simple BlogPulse Trend for the words “apple” and “microsoft” reveals:

At first blush, Apple seems to get much more buzz than Microsoft. This is really impressive considering Apple only has something like 4% of the desktop market share. But are we talking about the right “apple”? Certainly there are some mentions of “apple and orange” comparisons, or mentions of places like “The Big Apple” or “Apple Valley” or “an apple a day” comments. So are these mentions significant and how can we weed them out? My approach was to search for the word “apple” with the appearance of at least one other word that would indicate that discussion was about the apple of Steve Job's eye. I compared the results with all other mentions of “apple”, i.e. the mention of “apple” without the mention of any of the words on the list. I continued to tweak this list of terms until the two trends no longer resembled each other – indicating that I had separated the apples from the Apples.

My first Apple-apple comparison was simple. I used the search phrase “apple and (computer or imac or ipod)” and its inverse “apple and not (computer or imac or ipod)”. This BlogPulse Trend looks like this:

Notice the similar characteristics in the two trends. The most obvious being the spikes in January (Macworld Expo and the announcement of iPod Shuffle) and in June (the Apple-Intel announcement). These spikes should not affect the mention of the fruit so we're still not done. After several iterations, I settled on the following lengthy search phase :

apple and (osx or imac or imacs or powerbook or powerbooks or ipod or ipods or ibook or ibooks or emac or emacs or g4 or g5 or quicktime or computer or computers or intel or laptop or laptops or desktop or desktops or keyboard or cupertino or numbers or nokia or shuffle or mac or macs or itunes or os or x or cinema or google or gui or "Steve Jobs" or wozniak or woz or microsoft or adobe or macromedia or imovie or "Garage Band" or garageband or tiger or imail or imall)

The following BlogPulse Trend shows the signal (likely Apple Computer mentions) and the noise (mentions of “apple” that are not likely about Apple Computers). I also added total mentions of the word “apple” for scale and the word “banana” as a fruit that's not commonly used as a business name in blogs as a sanity check.

Because the average use of the word “apple” in a non-computer sense is reasonably static (as are most common words) it's use doesn't effect the overall trend but does effect it's magnitude. Applying our filtered “Apple” to our original comparison with Microsoft we get this:

Whew! Finally, we can start our analysis. With our filtered Apple trend, we see that while Apple holds its own with Microsoft in the blog buzz race, it's not a fruit ahead as we might have initially thought. Still, considering that, no matter how you measure, Apple's market is tiny compared to Microsoft's – Apple's buzz is disproportionately large for it's market. For this reason, I call Apple a buzz brand. Being a buzz brand is a good indicator that a company has passionate users. Again, this is a good thing. Why? Well, let's think about the recent Apple-Intel announcement. Since Apple users are passionate they easily buy in to the Intel deal and are instantly prepared to make the transition. They believe and trust that they are in good hands with Apple despite any rational notion that might indicate otherwise. My prediction is that Apple will loose few if any customers due to their “Intel Deal”. Furthermore, if this deal lowers the barriers to entry for “switchers”, i.e. lowering computer costs, increasing Windows software compatibility (speculation on my part) – Apple increases their value proposition for non-Mac users who are less passionate (or worse yet, not passionate at all) about they current system. Furthermore, these users will be inundated with buzz and encouragement from current passionate Apple users. When the time comes, this buzz may be the allusion of mainstream adoption that Geoffrey Moore prescribes in Crossing the Chasm that Apple will need to regain significant desktop market share against Microsoft. But I digress...

Second interesting point – Microsoft is an church-friendly brand meaning that it surges during week days and hits weekly lows each Sunday. Apple's dips on weekends are significantly less dramatic. This could be because Apple has a much higher penetration of “personal” computers and iPods than it does of “business” computers and as such get discussed less in the professional context that tend to surge during the work week. It could also, however, be another indicator of passion – passionate users don't evangelize because it's their job, they do it because they are passionate about the products they discuss. Passion doesn't end when the end-of-shift whistle blows. Hmm, does this mean Apple users are all heathens and sinners? Didn't Eve use an Apple? Anyway...

Next point – as a buzz brand, Apple generates really big bursts of buzz when stuff happens. Apple has a reputation of keeping secrets right up until a product is ready to show or ship. The main spikes you see here are Macworld Expo/iPod Shuffle, iPod Mini, OS X 10.4 released and the Apple-Intel announcement.

In contrast, Microsoft has a reputation of over-promising on features years before the product exists. In this regard Microsoft is a like the little boy that cried wolf – when they do announce something important, no one comes running. Microsoft spikes of interest are the announcement and release of their virus/spyware detection products and the Xbox360 unveiling. Also keep in mind that this is the Microsoft brand overall. There are certain verticals where Microsoft excels (if you will excuse the pun). For example, the Microsoft Xbox team seems to really get it. The unveiling of the Xbox360 was a picture perfect buzz spike that is still sustaining increased buzz for the Xbox brand – something I'll go into more detail on in an upcoming post.

In the next day or two, I hope to apply these techniques to looking at several brands and products in different markets. I also hope to discuss a technique for dissecting brand strength by product – like showing the contribution of Xbox to the Microsoft brand.

Man, this stuff is cool.