There are two main types of information that you can get from research that involves people: subjective and objective.
The word “subjective” means that it is an opinion, or a memory, or your impression of something. The feeling it gives you. The expectations it creates. Not a fact.
i.e. — There is no right answer.
To get subjective information you have to ask people questions.
The word “objective” means a fact. Something you can prove. Your opinion doesn’t change it, no matter how hard you wish.
If people had perfect memories and never lied (especially to themselves) we could ask them about this stuff. Hence, objective data comes in the form of measurements and statistics. But just because you can count something doesn’t make it objective.
For example: If 102 people say something is good and 50 people say it’s bad, the only objective information you have is the number of people that voted. Whether it is “good” or “bad” is still subjective.
As a general rule, more people makes more reliable information, even if it is subjective. One opinion could be completely wrong. If a million people agree, it is a good representation of the crowd’s beliefs (but could still be false, objectively). So collect as much info as possible for your research.
Lots of subjective info can become… almost objective?
If you ask a lot of people to guess the answer to something objective — like jelly beans in a jar — the average guess will often be pretty close to the real, objective, answer.
There are 3 basic types of questions:
Open Questions — “How would you describe me?” — This allows for a wide range of answers, and works well when you want all the feedback you can get.
Leading Questions — “What are this site's most useful features?” — This narrows the answers to a certain type. My example assumes that the site has useful features which might not be true! Be careful: this type of question also excludes answers you might want to know!
Closed/Direct Questions — “Which is more useful, site A or site B?” — This type of question offers a choice. Yes or no. This or that.
Remember, always ask Open Questions in your User Research, and avoid the other two types.
First of all, let’s nail down what personas or profiles are NOT:
So what is a persona / user profile?
It describes the goals, expectations, motivations, and behavior of real people. Why do they come to your site? What are they looking for? What makes them nervous? And so on.
All the information you need should be in your research and data. If you can’t back it up with research or data, it's not a good sign.
Bad Profile: Persona A is a female, between the ages of 35-45 with an above average income and education. They have at least one child and own at least one new vehicle. They are outgoing and career-oriented, and tend to be right-brain thinkers.
Why it’s bad: That might be great if you’re selling ads, but as far as UX goes, that profile is not useful. Why? Because it doesn’t allow you to say “no” to any feature ideas. What sort of features does a female between 35-45 need? It could be anything!
Useful Profile: Persona A is an experienced manager, mostly interested in one or two areas of expertise. They visit often, but they are pressed for time, so they focus on “collecting” content to read on the weekends. They tend to be prolific social media sharers, mostly to Twitter and LinkedIn. They consider themselves thought-leaders, so public image is important.
Why it’s useful: Now you have a lot of information to use! You know that fluffy content will not be popular, self-curating will be a big deal and you have a basis for setting up content categories. They need easy access to sharing, and only certain types of social sharing will be relevant.
You also get to say “no” to a Facebook campaign, because these users don’t spend time there, and digest emails will be better than frequent notifications because these people are already pressed for time.
Think of “Ideal” Users. Several of them!
When you think about features, think of the most valuable version of the users you see in real life. You’re not trying to support the current behavior; you’re trying to nudge those users toward an “ideal” version of themselves.
Also remember that all users are not alike! You will probably have a few different behavioral groups, and they all deserve a good profile.
Here's an example of a real persona: