What are your brand’s best and worst qualities? Your customers know best. We show you how to use social media to understand brand perception online.

One of the biggest takeaways from the recent Gap logo fiasco is that brand perception is wholly defined by the customer. No doubt this caught Gap North America President Marka Hansen by surprise, who early in the debate said that Gap chose the design because it’s “contemporary and current” (hufftonpost.com).

But had Gap studied its brand perception before going through with the logo, they would have learned that its fans value qualities that are not at all contemporary or current. Gap customers like the brand because it’s “classic”.

The Gap logo story illustrates how brands are built (or broken) on customer perceptions, not the tellings of its president or brand manager. But this doesn’t mean that branding is hopeless.

The wonderful thing about the internet is that it’s a free database of customer opinion. Brands can mine those opinions to better understand their core values and then build their branding around those values.

Simple measures of brand perception

The best way to understand brand perception is to study the language people use when they talk about brands. This understanding can come from something as simple as monitoring the brand’s Twitter Stream or Facebook page, or searching for “i hate [brand name]” or “i love [brand name]” on Google.

Unfortunately, this measure of brand perception is anecdotal at best. After all, just because one person says your cardigans are “comfy” or another says your jeans “rip within weeks” doesn’t mean that these are core values shared by the marketplace. So how do you get to the bottom of brand perception across the whole spectrum of consumers who are talking about you online?

Get fancy with language and mathematics

A more robust approach to measuring brand perception takes advantage of the HUGE wealth of conversations happening on the web. The method is called “linguistic analysis”, and it’s all about digging deep into the mountain of conversations and pulling out all of the words that are most significant to the brand.

Let’s look at a concrete example.

Nielson reported last week on an interesting case study about Greek yogurt. Despite the 121% higher average price of Greek yogurt, sales are up 203% over the past 52 weeks.

Buzz about Greek yogurt suggests something of a triple threat. The category appeals to consumers to satisfy three core needs: health, convenience and taste…Remarkably, apart from a relatively small subset of total category buzz related to deals and coupons, most of the conversation deals not with price but instead with the benefits of Greek yogurt.

Yogurt brands can do more than watch and marvel at the upward trend in buzz around Greek yogurt – they can actually take advantage of that buzz to gain a better understanding of their brand perception.

Consider a brand such as Fage Total Greek Yogurt, the best Greek yogurt on the market (according to this author).

Fage could use language analysis to study the buzz around its corner of the Greek yogurt conversation as follows:

1. Search the web for “Fage Greek Yogurt”, “Fage Yogurt”, “Fage Total” and related phrases that describe the brand, then pull the text from, say, the top 1,000 pages that contain those terms. That’s a pretty chunky set of data – what do they have in common?
2. Bring on the computers and the mathematicians. Plow through the data and pull out the most common words that appear across all of those documents
3. Bring in language analysis. Look at those words and decide which ones are most relevant. For example, Fage probably doesn’t care about “the” or “it”, but it would care about words like “creamy” or “revolting”.
4. Score the words by sentiment. That is, isolate all positive and negative words about the brand and sort them by relevance.

Language analysis requires some powerful computers, fancy mathematics and complicated algorithms, but it can be done, and lots of people are doing it.

We’ve been playing with language analysis ourselves here at Market Sentinel Labs and thought we’d have a go at understanding Fage’s core values. To do so, we applied Skyttle‘s language analysis suite to a large set of webpages discussing the Fage brand. Our analysis revealed a list of common sentiment words associated with the brand.

The results show that Fage’s core values align with overall consumer perception about Greek yogurt as reported by Nielson: it’s “healthy” and “delicious”.

Sentiment analysis also point to qualities of Fage that are unique to the brand. For example, the word “sour” begs the question – do consumers see this as a positive or negative aspect of the Fage brand?

This brings us to one of the most important points about measuring brand perception, and something that all too many brands miss:

Getting to the heart of brand perception requires looking beyond stats and sentiment words; you have to look at the contexts around those words.

In other words, our list of sentiment words and Nielson’s buzz chart tell us nothing about Fage’s core values as perceived by consumers. We need to drill down on the contexts of conversations around those top sentiment words to get to the bottom of things.

Digging deeper into contexts

If I were Fage, I’d want to understand why the word “sour” features so prominently in the list, and what people actually mean when they apply the word to the brand.

A deeper look at the conversations around “sour” reveal that Fage is frequently recommended as a substitute for sour cream (I personally recommend crème fraiche):

So by digging into the contexts, we discovered that one of Fage’s core brand perceptions is that it’s a healthy substitute for sour cream. The word “sour” alone didn’t get us there; it was the contexts that delivered real insight and actionable intelligence. Don’t be surprised if Fage starts getting in on conversations about Tex-Mex and Mexican food. ¡Ay, caramba!