Segmentation and Personas "Product people - Product managers, product designers, UX designers, UX researchers, Business analysts, developers, makers & entrepreneurs 6 December 2011 True Personas, Segmentation, Mind the Product Mind the Product Ltd 962 Product Management 3.848
· 4 minute read

Segmentation and Personas

The No Bullshit Guide to Product ManagementThis is a guest post written by Shahid Hussain, author of The No Bullshit Guide to Product Management.

Segmentation and Personas are tools to help you and your team understand your users better.

First, segmentation. Ready?

Before you built your product, you probably considered who might use it. That group of people is your market. Let’s say you’re making a cooking website – you might have a target market of … anyone who needs help with their cooking and is comfortable with a laptop in the kitchen.

Not everyone in your market is the same. There will be:

  • Different genders
  • Different ages
  • Different locations
  • Cooking in different environments
  • Cooking for different people
  • Different levels of cooking ability

… and so on. You can look at your market in each of these different ways, and divide it up so you can understand it better. For example:

  • Men / Women
  • 0-6, 7-12, 12-15, 15-25, 25-35, 35-45, 45-55, 55+
  • USA, EMEA, Asia
  • At home, at work (i.e. professionally), in a tiny kitchen, in a huge one
  • Cooking for themselves, cooking for one other, cooking for a family, cooking for a group of people
  • A professional, a regular cook with talent, a mediocre occasional cook, a blunderer in the kitchen

This is called segmentation.

I made all of these up – fresh from my rear end. They are a starting point, and desperately need to validated with some data. You can do that by interviewing prospective customers, doing surveys, or examining data from a product you already launched. The data will confirm or deny your segments and how to divide within them. For example, you might discover that cooks in a tiny kitchen and a huge kitchen behave pretty much the same way as far as you’re concerned, so there’s no need to worry about it. On the other hand, you might discover that cooks who earn differently look for different types of recipes, so it’s important to examine a segmentation you didn’t have already.

It’s tempting to think – why bother? You probably know your market well enough that any list you pull out of your rear end will be fairly accurate. Here’s why – it’s not the 90% you get correct that matters. It’s the 10% that will surprise you. That will set you apart from your competition, and makes this work worthwhile.

Once you have a list of segmentations, and you’re happy, there are a couple of obvious things to note:

  • Your users aren’t evenly spread throughout your segments. For example, looking at your cooking app users by age, you might find that most are around the 35 mark. You might even find there’s more than one bump in the graph – for example, an application that appeals to both mums and children might have age bumps at 12 and 40. (This is called a bimodal distribution.)
  • Some segmentations go well together. For example, if you’re seeing bumps in age around 35, mostly female, and evidence they are cooking for a family, it’s likely that you’ve nailed the housewife market. Once you start talking to your users, you’ll be able to validate that theory.

If you have some segments you’re happy with, congratulations! And, you’re not done. There are two problems:

  • Holding all of these different perspectives in your head at one time is hard.
  • Communicating it all to your team, and others within the company, is also hard.

The solution is personas. Personas are imaginary people that represent a group of your users. You’ll usually have a few personas, in the same way that you usually have different groups of users.

Let’s say that after examining the usage data from your cooking app, and talking to customers to validate your theories, you discover there are basically two types of users, which contain three of your segmentations (age, gender, who they are cooking for):

  • 35 year old mums, who are cooking for a family.
  • 20-30 year old single males, who are cooking for themselves.

You need to keep these in mind when thinking about the design of your product, what type of recipes to include, and so on. How can we do that? By naming them. Just create two imaginary people who are the average of these two groups. Research their behaviour – what phone does the average 35 year old mum use these days? What does she do with her day? Add all that into your persona. It gets more helpful as you add detail, but try to capture the variation of

  • Cherise, a 35 year old mum from Chesapeake, Virginia. Married. She has three kids, cheers them on at soccer games. She loves her iPhone. She uses cooking as an expression of affection for her family. She owns her family home with her husband, and they’re paying off a mortgage.
  • Rick, a 25 year old professional from Berlin. Single. He cooks for himself sometimes when he’s at home, but combines this with eating out during the week, and likes to get in and out of the kitchen quickly. He doesn’t cook anything fancy. He rents his place, and has a reasonable amount of disposable income.

Is it accurate for all of your users? Nope. Is it a reasonable average? Yes. Most importantly, it will help keep you and your team on the straight and narrow. Is feature X going to help Cherise? Would Rick hate that recipe? Create little posters with pictures of Rick and Cherise, some description of their habits and statistics, and pin them up around the office.

Personas will help remind you that you’re not designing for incremental traffic in Google Analytics. You’re designing for real people, and understanding them will help you understand they way they feel about your product.

If you liked this post, try the book: The No Bullshit Guide to Product Management. It’s cheap, short and helpful.

Comments 2

Excellent idea. I shall pick through my data and try to invent some personas based on the emails I’ve received from users. Hmm, I wonder if someone smarter than me could write an algorithm that would automatically generate personas based on Google Analytic data…

I digress.

Do you have any thoughts on how to (or whether to) apply some sort of weighting to each persona? For example, if feature A will be great for Cherise, but Rick will go potty over feature B then I think it’s important to consider if 90% of your users are Cherises. Maybe the size of the poster should be proportional to that persona’s percentage of your userbase.

Excellent idea. I shall pick through my data and try to invent some personas based on the emails I’ve received from users. Hmm, I wonder if someone smarter than me could write an algorithm that would automatically generate personas based on Google Analytic data…

I digress.

Do you have any thoughts on how to (or whether to) apply some sort of weighting to each persona? For example, if feature A will be great for Cherise, but Rick will go potty over feature B then I think it’s important to consider if 90% of your users are Cherises. Maybe the size of the poster should be proportional to that persona’s percentage of your userbase.

Join the community

Sign up for free to share your thoughts

About the author