How To Create Customer Personas (with Real-Life Data)

The risks of failing to understand your customers

According to a study by the Edelman Group, brands are failing to understand some of the fundamental motivations and concerns of their customers.

Edelman’s consumer marketing study surveyed 11,000 people across eight countries who had taken part in at least one brand-engaging activity (e.g. following a brand on Facebook) in the previous year.

Some 51% of respondents felt that brands under-performed when it came to asking about their needs. Only 10% considered brands to be doing it well.

It gets worse. Two additional research studies paint an even bleaker picture. Responsys surveyed over 2,000 U.S. adults to find out how they felt about their relationships with brands. Some 34% said they had “broken up with a brand due to receiving poor, disruptive or irrelevant marketing messages.”

The customer engagement specialist Thunderhead conducted a similar study. They found that one quarter of U.S. consumers would switch to a different provider after one negative experience.

As the data shows, too many people are fed up with how businesses interact with them. It’s not all risk—there are rewards for companies that get it right. A 2017 study showed that 79% of U.S. consumers are loyal buyers from brands that “understand and care about me.”

And yet, only 6% of senior executives believe that their companies understand their customers’ needs extremely well. No wonder customer acquisition and retention is such a problem.

So what can you do about it?

The case for building customer personas with data-driven research

Patching together actionable information about your customers with gut feelings, good intentions, and some duct tape isn’t a recipe for success.

If the above studies aren’t enough to make you quiver in your boots, take a look at the debacle of JCPenney’s 2012 rebranding.

Within a month of becoming CEO, Ron Johnson launched a radical rebranding. He changed the look and feel of stores—getting rid of lucrative private-label brands and replacing them with designer-inspired ones that, it turned out, where priced too high for most customers.

Johnson also overhauled how the company did business. JCPenney went from a sales model based on constant coupons and markdowns to “everyday low prices.” It was a disaster. Sales collapsed within months of Johnson’s takeover.

Johnson explained the failure:

“I thought people were just tired of coupons and all this stuff. The reality is all of the couponing we did, there was a certain part of the customers that loved that. They gravitated to stores that competed that way. So our core customer, I think, was much more dependent and enjoyed coupons more than I understood.

Johnson admitted that he didn’t understand what his customers wanted. But he also made it clear why he didn’t understand. When asked to consider rolling out the new changes on a limited basis, his response was, “We didn’t test at Apple.”

He blindly instituted a plan based on what worked elsewhere without testing or, apparently, taking into account what drove his customer base to shop. A greater appreciation for customer personas could’ve avoided the debacle.

So what are customer personas?

In case you’re unfamiliar with the term, let’s begin with a definition from a leading expert in the field of buyer-insights research.

“Buyer personas are research-based archetypal (modeled) representations of who buyers are, what they are trying to accomplish, what goals drive their behavior, how they think, how they buy, and why they make buying decisions.” – Tony Zambito

In essence, personas are fictional representations of segments of buyers based on real data reflecting their behaviors. Their purpose is to put the people in charge of company decision-making in the shoes of the customer.

The problem with many personas is that they’re either based on irrelevant data, poorly sourced data, or no data at all. As B2B marketer Ardath Albee notes in an interview:

“I see a lot of personas that are what I kind of call “Ouija Board” personas, because they are based on stuff that marketers would never know.”

While some basic demographic information, such as gender and age, may be applicable, other very specific attributes (e.g. what the family dog eats—unless you’re selling dog food) you may glean from research or anecdotes is useless.

What does a good customer persona look like?

Customer personas can be as basic or complicated as you like. They can take various forms, but at the end of the day, their value lies in how clearly they reveal what drives different types of buyers. While there are plenty of templates and examples to follow online, think about modeling personas from your available qualitative and quantitative research. Focus on:
  • Behavioral drivers. These encompass your customers’ goals, what they want to accomplish, and their journey to finding your business.
  • Obstacles to purchasing. Take into consideration the hesitations and concerns of your customers. How do they view your product or service? How does that impact the information they need to make a decision?
  • Mindset. Your customers come to the buying experience with expectations and preconceived notions. Are they shoppers who want the thrill of the bargain, or do they expect a refined experience? Selling a weight-loss program will be more emotionally charged than, say, selling routers.

Giving your personas names and faces is less important than ensuring they are based on real people, not stereotypes. As Tony Zambito notes, ineffective buyer personas “read like job descriptions and offer little insights.”

Conduct qualitative research for buyer personas

To understand segments of your customer base and what motivates them, begin by asking them questions. There are three ways to get started.

1. Customer surveys
Conducting surveys online or off with open-ended questions is critical to understanding how your customers frame their motivations and needs.

The goal is to get inside your customers’ heads and make sure your personas are based on what real people think, not just your idea of what they think.

Ask between 7 and 10 questions about their behavioral drivers, obstacles to purchasing, and mindset. Depending on your business, the questions vary. But the end goal is always the same—actionable information that serves your needs.

For example, survey questions can include:

  • When did you realize you needed a product/service like ours?
  • What problem does our product/service solve in your life?
  • What doubts or hesitations did you have before buying?


2. Phone and in-person interviews
18 Tips on Conducting Killer Customers Interviews from Zachary Cohn
Talking to your existing customers can provide valuable information into their buying habits, what motivates them, and the words they use to describe your product or service.

Conducting interviews can be expensive and labor intensive. However, the answers can be illuminating. You can go back and ask your respondents to elaborate, getting details not available through surveys.

Sean Murphy has some great tips for conducting customer interviews.

3. Web and exit surveys
These surveys are designed to have a single question pop up on your site at a designated time. They’re particularly good for finding out why your customers aren’t completing a purchase.

The question to ask depends on your goal. Do you want to know if your site or products/services meet their needs? Or do you want to understand sources of friction that keep them from buying?

Experiment with your question to see what gets the most responses and which responses return the most insights. For example, if “Why didn’t you complete a purchase today?” isn’t as successful as you anticipate, try, “Do you have questions you weren’t able to answer today?

Distill your qualitative research

Segment users based on commonalities you find. Look first to intent, then to possible hesitations and the ways in which customers are susceptible to persuasion.

You may find two personas that you can clearly define; you may find four. The number depends on what the research supports.

Say you sell organic household cleaners. After combing through the data, you identify a persona: Beth, a 35-year-old woman who is worried about her family’s exposure to chemicals in the environment.

She cares about reducing her carbon footprint and is willing to pay a little extra to make sure she’s buying a product that’s sustainable.

  • What are Beth’s behavioral drivers? These products give Beth a sense that she’s doing something right for her family and the environment. She can easily order them online which works for her hectic schedule.
  • What are Beth’s obstacles to purchasing? She’s concerned that the information about the ingredients’ sources is correct. She worries about the packaging that houses the products and how they’re shipped (i.e. if they contain harmful substances).
  • What are Beth’s expectations from the buying experience? The presentation is important. She wants a product that reflects her values. Getting a bargain is not as important as getting a product she trusts.

While picking a name and age isn’t a must for personas, it helps visualize a person behind the persona. Beth feels real as opposed to “Persona #1.” You may be more likely to ask what that persona needs and wants when creating things like copy or design.

Use quantitative data to back up your qualitative personas

You’ve created a few personas with your qualitative research—segments based on goals, behaviors, and attitudes. Google Analytics can round out your personas with quantitative findings. Segments in Google Analytics can showcase the on-site behavior of key customer groups. Create segments for:
  • Average revenue per user;
  • Transactions per user;
  • New versus repeat customers;
  • Frequent customers.

Even on its own, quantitative data can help you find and market to your key personas more efficiently. Watchfinder, a UK online watch retailer, found that less than 1% of their visitors completed a transaction on the first visit to the site.

As a result, they decided to create a remarketing campaign using Google Ads. They started with customer segment insights from Google Analytics—creating lists based on user language, location, and stage of the purchase funnel.

In conjunction with a traffic performance analysis, they realized that much higher engagement and conversion rates came from ISP addresses in the London Financial district.

By retargeting these site users with messages tailored specifically to employees at large investment banks, they increased the average order value by 13%, with an overall return on investment after 6 months of 1,300%.

If they had wanted to take this effort to the next level, they could’ve added qualitative persona modeling to further refine their marketing efforts.

Apply customer personas to buyer behavior

Remember the cautionary tale of Ron Johnson and JCPenney? To him, a “fair and square” pricing model and relative transparency made perfect sense. To his entrenched customer base, not so much.

The JCPenney shopper expected to see markdowns and use coupons. When they could no longer viewed prices through the same lens—no matter the rationality of the vantage point—they no longer saw the value.

Kahneman and Tversky ascribe this economic behavior to “Prospect Theory.” People evaluate outcomes relative to some reference point that usually involves their current situation. Gains and losses are viewed through the prism of perceived outcomes, not absolutes.

In 1981, the two researchers posed this survey question to a set of randomly selected respondents:

68% of the respondents were willing to make the extra trip to save $5 on the calculator. When the question was posed to another set of respondents but the pricing reversed—the calculator was on sale for $120 elsewhere—only 29% of people were willing to drive across town.

The savings was the same, but the framing of the question different. As William Poundstone notes in his book, Priceless: The Myth of Fair Value,

“The price of being so acutely sensitive to ratios and contrasts is a relative insensitivity to the absolute.”

If only Ron Johnson had spent a bit of time thinking about that.

How Facebook used persona research to improve their reporting system

Facebook fields millions of user complaints each week. Many complaints are anonymous, making it difficult and time-consuming for Facebook to address.

When investigating these anonymous complaints, Facebook focused on teenagers of both genders to understand a common issue—when a user wanted a photograph, posted by another person, to be removed.

After talking with different segments of teenagers, they found that the word “report” as a click trigger caused friction. Kids didn’t want to get their friends in trouble. When they changed the phrasing to “This post is a problem,” it made it easier for teens to cite the specific issue.

Facebook also tested a change that allowed the person complaining to name the recipient and the emotion the post triggered.

According to the data, they found nearly an 85% likelihood that the originator of the photo post would reply back to the person offended or take the photo down when the words “It’s embarrassing” were used.

This is a good example of how you can take your qualitative research for personas, apply it to another area of your business, then test it using quantitative means.


Customer personas are a tool. As with all tools, they’re only as good as the people using them. They can provide tremendous insights into how to create better user experiences, persuasive copy, or pricing models.

The key is to do the qualitative and quantitative research that enables you to model personas out of data, not intuition.

Most importantly, remember your personas should reflect real people with real motivations, desires, and concerns. When we lose sight of the human element, the customer isn’t far behind.

No Comments

Post A Comment