Passive Text Analysis: Learn What Makes Your Audience “Tick”

(Originally published in Online-Behavior)

At the heart of any marketing initiative is one goal: to influence an audience. The tough part is figuring out 1) who the audience is, 2) what will influence them, and 3) the best way to do so. The Myers-Briggs Type Indicator (MBTI) has long been considered a standard to help answer the first question in corporate HR, for example, but its popularity among marketers is growing rapidly.

There are significant problems with using the MBTI for marketing, such as its lengthy survey format, prevalence of bias, and arguable inconsistency. The problems with survey-based insights have led these methods to finally be questioned as useful means for understanding an audience. While some marketers may rely on these insights, that could be a costly mistake.

The MBTI, like any survey, has statistical problems — have you ever taken the test twice and gotten different results? Additionally, there’s an inherent bias associated with survey-taking. Think about it:

People rarely take surveys just for the fun of it. Whether they’re expecting to earn something, gain acceptance to a group dynamic, or prove a point, human nature causes people to bring a level of bias to their answers.

Why Passive is Preferable

By using personality analysis that’s 100 percent passive — meaning audience members don’t know they’re being “tested” — you learn the deep-rooted characteristics or temperament of your audience. Text analysis of the words they use, as well as the actions performed while using social media, show temperament, which can be thought of as how people are hard-wired.

As “moment” advertising using social media gains in popularity, its risk can be too high for many brands to consider. The qualitative insights from passive analysis can provide profound comfort to product strategists and marketers who wring their hands over questions like, “What tone will they respond to?” and “Will they be turned off by this copy or these images?”

The Big Five: Take a Seaside View

My preferences for passive audience segmentation focus on the Big Five personality traits, which make up the acronym OCEAN: openness, conscientiousness, extraversion, agreeableness, and neuroticism. By assessing the diction choices of your audience, you can gain insight into their personality types and understand what makes them tick.

1. Openness to Experience

Distinguishes imaginative, creative people from those who are more down-to-earth and conventional. Open people are intellectually curious, appreciative of art, and sensitive to beauty.

Related words: Receptive, innovative people differ in their language use from those who are more closed-off and conservative. People who use more money-related words, like profit, pricing, and refund, tend to be less open to new experiences.

2. Conscientiousness

Refers to the way we control, regulate, and direct our impulses. People who lack conscientiousness tend to be more impulsive.

Related words: Conscientious people use social words like roommate and secret, as well as human identifiers like female, citizens, and infant. Impulsive people tend to use more swear words, as well as perception words like scent and sour and seeing words like beauty and views.

3. Extraversion

Marked by pronounced engagement with the external world. Extraverts enjoy being with people, are full of energy, and often experience positive emotions.

Related words: People who use perception words, such as scent and seeing, tend to be more introverted. Those who use work-related words, such as educate, duty, and executive, tend to be more extroverted.

4. Agreeableness

Reflects a person’s concern with cooperation and social harmony.

Related words: Agreeable people value getting along with others and tend to use words like hero, helps, and wins. They also use positive words, such as luck, merit, and OK, as well as biological words, like rehab, exercise, and drink, more frequently.

5. Neuroticism or emotional stability

Refers to the tendency to experience lower levels of strong negative feelings, such as anxiety, anger, or depression.

Related words:People who use words such as miser, restless, and scary tend to be less emotionally stable. Those who use ingestion words, such as cook and ate, also fall into this category.

Be Smart with Social

You can use social media to determine where on the Big Five your audience members fall. But this goes further than just monitoring the words they choose; you have to look at actions as well as diction.

Start with understanding the Big Five, but pay attention to how activity markers (like changing one’s profile picture a lot or frequent commenting) correlate with an action. This information has been well-documented and can be used to tune your copy, imagery, and media to those underperforming — but promising — segments.

For example, if Red Bull wanted to expand its consumer base to target middle-aged women, it could use a brand personality formula to find those women who were classified as “daring.” The personality traits that define this category are high levels of openness to new experience and low levels of conscientiousness. Once you add in all the other activity and interests markers, you can create products, packaging, and campaigns that appeal to them.

Using passive text analysis to learn what makes your audience “tick” is still a relatively new concept. Legacy approaches — like surveys, focus groups, and polling — are dying off quickly; the new, “moment-based” generation of marketers has little patience for the lengthy project times needed for surveys.

Your consumers are telling you so much about who they are and what they want — and not through a Myers-Briggs lens. With passive text analysis, you can finally start to both listen and cut through the noise to provide some value for those big social media budgets.

About Mattr

Segment your audience in hours — not weeks or months — all without asking questions. Craft campaigns and products that appeal to their personalities and unique interests.

How to Diversify Your Market Research Data to Find the Right Balance

(Originally published in iMediaConnection)

Ask anyone who’s struggling to stay on the cutting edge: Technology is evolving at light speed, and that means new innovations, new gadgets, and new data sources are hitting the market every day. It also means that America’s 300,000 market research analysts are feeling more confused and overwhelmed than ever.

It’s no wonder: With every technological advancement — whether it’s a new gadget, wearable device, or social network — new data points grow exponentially. And business executives are challenging researchers to use that data to make smarter marketing decisions right now. After all, it’s just data; how hard can it be?

But here’s what C-suite executives don’t know: The big data they need for thorough quantitative analysis is buried in Oracle databases. And once it’s exported — only after asking a company’s IT department very, very nicely — it’s just millions of rows of raw, uncorrelated atomic data. Agencies can have an even harder time getting transactional or behavioral data from a company, and no single social platform provides every metric, which translates to acquiring access to several tedious, expensive platforms.

There is hope, however. First, companies need to create a new, diverse portfolio of research approaches. Then, they need to explore how these approaches complement each other — and which combination fits their business best.

The Right Questions

Every day, all over the world, millions of consumers are leaving trails of comments, posts, pins, and cookie-based data. So why are concrete, accurate consumer profiles so hard to create?

Most companies already know they should start by getting a complete, in-depth profile of who’s buying what they sell and why they’re buying it. Ideally, a complete portfolio of source data would show:

  • Who bought your product
  • What they bought
  • When they bought it
  • Where they bought it
  • Why they bought it
  • How they bought it

To answer all of these questions in a complete, meaningful way, you need more than one kind of data analysis. After all, five out of six of these require quantitative data.

In the past, ascertaining a concrete answer for “why” was the hardest question on this list. But now, with so many different ways and places to buy, the other five questions are difficult to answer completely, too. Your best bet for netting answers to all of these questions is to blend several different kinds of marketing research together.

The Right Balance

The case for mixing in-depth, analytical research with real-time market research is growing stronger every day, and big-time industry leaders are beginning to share the news.

At GigaOM’s Roadmap 2013, Instagram’s founder and former CEO, Kevin Systrom, had a unique angle on why businesses need a more diversified research strategy. He said that “hunch-driven design” works when you have 100 users, but you need data-driven design when you have 150 million.

It’s true. Big brands with a reputation to maintain can’t gamble on a hunch. They need hard-and-fast numbers to base their actions on, and that’s where a balanced mix of these two kinds of research can come into play.

The Right Process

Creativity and research can live together beautifully and help you create a smarter, more tailored product. But there’s a clear order you should take when pairing these two together. 

After all, smart development comes from a creative place, but that doesn’t mean creativity can’t be informed by facts. When it’s used correctly, data shouldn’t hinder your company’s innovation. It should guide it.

That’s why creating a mix of in-depth, analytical research with real-time market research is a necessity for any growing, evolving company. It can give you the information you need to innovate — and the justification you need to take that next great leap forward.


About Mattr

Segment your audience in hours — not weeks or months — all without asking questions. Craft campaigns and products that appeal to their personalities and unique interests.

3 Reasons Time is the Achilles Heel of Consumer Research

3 Reasons Time is the Achilles Heel of Consumer Research

Screenshot 2013-11-19 at 14


(Originally published in B2B Marketing)

Imagine a scenario in which you want a new book. You go to an online book retailer and sign up to fill out a survey. A customer service rep mails you a questionnaire or calls to ask what kind of person you are, what books you like, and how much time you spend reading. He may even talk to your friends and family to make sure you gave an accurate portrayal of yourself.

After the form is filled out, a week is required to manually collate the data. Then, the company sends you an email with a suggested book. You hop over to the site and pay for your personalized recommendation, and five days (13 since you started looking) later, you receive it.

See how crazy that is?

Yet most companies conduct consumer research with the same lengthy, manual-intensive approach.

Consumer Research Takes Too Long

Remember how we used to check email? We’d hook a computer up to the phone line, wait for the dial-up shrieking to end, and download our email. Consumer research today is the dial-up Internet of yesterday.

Think about it. It can take weeks at best — usually more like months — to complete a consumer research project. How relevant is the data at that point? What if the competition uncovers a solution everyone’s looking for while you’re still gathering data?

Did you miss the “moment”? Almost surely.

And It’s Just a Snapshot

Most consumer research is done manually, right? Think about those survey people you avoid at the mall or the surveys you receive via email. (Cue dial-up screeching.)

Opinions could have changed drastically while you were waiting on survey data. Today, three to six months is a lifetime in the modern marketing world.

Marketing windows are indeed compressing. CMOs today are intrigued by the target moment — the Oreos Super Bowl tweet that won such acclaim. But, for good reason, moment advertising is inherently dangerous and seat-of-the-pants assumptions just aren’t good enough to risk your brand’s reputation. You need hard data, and you need it now.

Making Assumptions is Too Risky

And yet, to capture that ‘moment’ that’s exactly what’s happening with some brands.
Big Data is a very possible answer. And yes, Big Data may be an unsexy cliché, but it’s the Bill Gates of the marketing world. In an unassuming way, it will be running the world of advertising.

But like Bill Gates’ achievements, it’s very hard work that you need to undertake with dedication. It’s onerous, but possible, to sift through tweets and Facebook posts to look for insights. Listening tools like HootSuite will let you follow mentions of your company or product across several social media platforms. Platforms like Spredfast and Salesforce are decent tools to find quantitative data and a few qualitative metrics like sentiment analysis, which can be very helpful when your audience is segmented.

Making assumptions about the tastes and personalities of your target segments without data can be too risky. Yet it’s happening more and more because the comfortable alternative is relying on outdated, expensive market research, which means you’ve missed the moment. Falling back on traditional consumer research will have you downloading that file at dial-up speeds while your competition flies past.

It’s about time market research left the dial-up era.


About Mattr

Segment your audience in hours — not weeks or months — all without asking questions. Craft campaigns and products that appeal to their personalities and unique interests.

What is Your Focus Group NOT Telling You?

What is Your Focus Group NOT Telling You?



Originally published in Everything Business Corp!


Marketers have used focus groups to get feedback from real consumers for decades. But often, a focus group just reaffirms marketers’ beliefs or gives a distorted view of how customers really feel. The truth is that focus group research often fails — or at least fails to deliver any groundbreaking insights.

What your focus group can’t tell you
Because of the nature of focus groups — a small sampling of people led in a group discussion by a moderator — a focus group cannot tell you about:

  •  Benchmarking. Benchmarking allows you to better leverage your marketing or R&D dollars and uncovers extreme data points, either to avoid negative response or capitalize on positive response.
  • Personality. You can fill your focus group with young women, but if you want to target compassionate young women, you’re going to need something other than a focus group to find them effectively.
  • Trends. Focus groups provide you with detailed information that is frozen in time. You get a one-time snapshot of a group’s response, but this doesn’t allow you to see trends.
  • Unaffected Responses. Probably the most significant drawback to focus groups is that the participants’ responses are greatly affected by two parties: the moderator and dominant voices. Research studies on interviewer/response bias have appeared since 1955, and the moderator’s attitudes and behavior can have a significant effect on the group.

About Mattr

Segment your audience in hours — not weeks or months — all without asking questions. Craft campaigns and products that appeal to their personalities and unique interests.

Why Aping Oreo’s Super Bowl Moment Is A Pointless Risk [3min read]

Why Aping Oreo’s Super Bowl Moment Is A Pointless Risk [3min read]

Stop Reacting to Others’ Real-Time ‘Moments’ and Do Your Research to Create Your Own


There’s a lot of hub-bub over ‘target moments’, or ‘moment advertising’. But how much should you worry about so-called real time marketing? Is it worth the risk to your reputation and budget? You’re no doubt under a lot of pressure to spool up a team to react to these moments others create. You should re-consider.

There’s good reason for you to be feeling this pressure. The reactive, real-time moments, both the wins and losses, are the ones that I seem to hear about most. Here’s a byte that had the audience tittering at a recent conference:

“Advertisers aren’t looking at target markets anymore. They’re looking at target moments.”

The discussion continued to talk about those real-time, opportunistic moments like when the lights went off during the Super Bowl and Oreo tweeted this:

The Dunk in the Dark story

Now forget the story for a moment and focus on the image and copy. What does it make you feel?

If you know the story behind it (and that’s a bigger “if” than you think), you may feel envy or admiration. After all, it was a herculean effort to make this happen with such apparent ease.

This was the Addy of real-time moments.

Moments You Create Can Offer Far More Value

Here’s another conquest from the Oreo team:

How does this one make you feel? For me, I nearly teared up with pride that a major consumer brand would lead with its chin, hands-on-hips, and shout out the message of equality. They voiced a very strong opinion, that their research told them would work, rather than reacting to a real-time moment with a bit of safe copy.

“It made me feel great. I bought some Oreos.”

From a consumer’s point of view, the Gay Pride ad made me buy Oreos. Dunk in the Dark? It excited marketers.

Moments You Create are Less Risky

What’s the win-lose ratio for reactive moments? Does it matter?

The point of view of the brand manager can win out over creative. Why would she risk her brand’s reputation to win the praise of the advertising media?

Reacting to other people’s moments is the desperate action of a cornered cat. It’s always playing offense — and rarely an opportunity for anything but an innocuous response. It’s why the Oreo copy is so milquetoast.

Except in this case, you’ve not been backed into a corner. You don’t need to dedicate the people or money to react to someone else’s moment. It’s a phantom crisis.

Create your own moment. Let others react.


About Mattr

Mattr’s Web-based software is for people who want sophisticated segmentation analyses on a brand or campaign but need it sooner than a long-term research project.