“If there’s no research behind a persona, I prefer a more direct name for it. I like to call it a bullshit persona. You start with a bullshit persona, you do your research, and only then do you have a persona. If you skip research, you’re just left with the bullshit.” - Tomer Sharon, Validating Product Ideas
Business school taught me to create customer personas, so I did.
Post-graduation and after a few years working in marketing, I realised that when I created a persona document, I'd file it, then never once refer to it.
This all changed when I read the quote above from Tomer Sharon. In their most common form, customer personas are bullshit. They consist of assumptions, light demographic research, and little else.
There's another way. It takes more time, but you're left with a valuable resource to use when creating and selling your products.
Start with the BS
You need a starting point to create a persona, which comes from your general assertions of who you’re making a product for.
I’ll use a persona I created in the past as an example throughout this piece. We were building a lead generation platform for specialist contractors in the UK, like a roofing company working for a main construction contractor. The starting point looked something like this:
You can see that there are four sections in the template:
- A persona name you'll reference internally
- The demographics, which are self-explanatory
- Problems - put down the fundamental issues you assume the persona has regarding the domain you are interested in
- Solution - the product that you offer
You know the type of person you're building a product for here. The purpose is to understand the kind of person you need to find to conduct qualitative research (interviews).
Finding interview participants
You want to speak with real people that represent the BS persona. You may have created two or three personas, which is fine. Five or six is too many.
Conduct a minimum of six interviews, maximum of twelve. Less than six, and you won't have enough material to extrapolate reliable assumptions. More than twelve is too much work, as each interview accounts for probably between 3-4 hours of work. You'll see patterns forming at around four or five interviews and repeated phrases and concerns from participants.
Finding participants can be difficult depending on whether you have an existing network of people that match the BS persona. You'll need to budget for an incentive. I offer something generous to grab attention and ensure people want to turn up - something like a $50 Amazon voucher for a 30-minute interview. $500 is a fantastic investment to get this insight.
If you have no immediate contacts, find participants via second-degree connections. You could also DM suitable candidates through Twitter. If that's not an option, go down the advertising route on Facebook or LinkedIn. Facebook is better for B2C, LinkedIn for B2B, as you can create relevant audiences on the respective platforms.
Some people like participating in surveys. Others may tell you anything you want to hear to get that $50 voucher. So, you need to make sure that the participants genuinely fit the criteria. You need a screener.
Here's the structure of the screener:
List of assumptive attributes
You create assumptive attributes to identify the characteristics that any respondent needs to fit our BS persona. This doesn't go out with the screener questions. It's for your eyes only.
For our example, these would be:
- Director of a Limited Company
- Limited company provides subcontractor construction services
- Company is less than five years old
- Clients include Main Contractors or Property Developers
- Lives in the UK
- Under 45 years old
Now we create screener questions. We don't want the respondent to know what type of person we're looking for. In our example, the screener questions would be something like this:
Which of the following is most accurate about yourself?
- I am the Director of a Limited company
- I am self-employed
- I work for an employer
- None of the above
Which of the following is most accurate about the company you work for or own:
- We are a general building contractor
- We are a specialist trade contractor (e.g. plumbing, roofing, etc.)
- We provide building materials
- We offer design services
Please select the type of business that best represents the category of clients you work with:
- Main Contractors
- Home Owners
- Local Government
The company you work for or own is
- Less than one year old
- Less than three years old
- Less than Five years old
- Less than eight years old
- More than eight years old.
Select which is true
- I live in the UK
- I live outside the UK
Select your age bracket
- 65 and over
You'll see from the screener questions above, as a respondent, it's not evident which answers I'm looking for.
If you have zero budget, use Google Forms to create this questionnaire.
Once you've got the forms back from your initial group of respondents, you can ensure that you only schedule participants who meet your criteria.
I'm not going to write a breakdown on conducting an interview. A section in this article isn't enough to cover all nuances involved there. Interviewing Users by Steve Portugal is a good starting point if you'd like to learn about conducting interviews.
I tend to conduct a 30-45 minute recorded interview on Zoom. By recording the interview, I don't have to create notes, which means I can focus on the conversation and ask open questions where the opportunity arises to delve deeper into responses.
Once I've conducted the interview, I get everything into a transcript. There are plenty of tools around to do that. I use Descript.
Once everything is down on a transcript, I'll break the information into a new document that categorises feedback (e.g. references to specific problems, good descriptions of the problem, wording to reuse in copy, etc.).
Creating the Persona
With the interviews complete, you can build your final personas. One option is to merge all respondents into a fictional persona. Or, if one respondent in particular best captures a group, you could lean on that person’s responses to build the persona and add a few other answers from other respondents to round it off.
For our example, here’s a final persona.
How to use the Persona
The first benefit of creating a persona based on actual research is you can focus on communicating your solution and addressing real problems identified.
In the previous example, with our BS persona, we assumed that the user’s problem was getting sales leads. We were wrong. The persona was getting plenty of sales leads and didn’t need more. What they needed were better sales leads. They needed leads with sufficient information to help them accurately forecast their chances of winning the sale.
This insight was critical. It informed the product we were building (what information do we need to provide with the lead?), and it helped us create relevant copy (address the real problem). It also meant we could use the customer’s voice in the copy, which is incredibly persuasive.
It’s also helpful to have personas when you’re part of a team. Many of the insights you’ll get from the research seem counterintuitive. If you were to present them as your ‘qualified’ assumption, you’d likely get pushback from people who may be less qualified than you to make the assumption. Still, nonetheless, in my experience, everyone thinks they know marketing.
When you present a persona backed up by real research, you don’t get dissenting voices from investors, employers, or clients. They accept the persona and decisions you need to make on the back of that persona.
You now have a concrete user in mind to build a product, test features, create landing pages, and write ads.
Join the newsletter to receive the latest updates in your inbox.