Surveys are most useful when elegantly deployed; having a working understanding of survey segmentation and targeting best practices will help you ask the right questions and get actionable answers.
The best way to keep your product roadmap under control (and avoid playing guessing games) is by making informed, data-driven decisions about your customer needs and consequently, your investment in new features. Thankfully, in-app surveys make it easy to collect specific data from your users (so easy, in fact, that they’re happy to contribute!) contextually right inside your app.
In this guide, we’ll quickly recap and then explore actionable strategies you can take.
How survey segmentation will help you get better feedback
Segmentation is the process of dividing your users into groups with similarities—you do this for multiple reasons. In short, it’s for both marketing and analysis.
You want to entice the right users into trying your app, and when you reach out to them for feedback, you want your response rates as high as possible. Let’s learn how you can use segmentation to your advantage, and target your users strategically. It’s not just about attracting users, but keeping them too—untargeted, annoying messaging is the leading cause of app uninstalls.
By zooming in on specific subgroups, you can learn more about your customers and better cater to their needs. People in general are less likely to respond to generic advertisements and questions. But when you tailor your messaging to a specific group of people, you increase your odds of hearing back from them—and your odds of selling to them. Basically, segmentation makes sure you’re targeting the right people with your outreach. Your objective is to stay relevant and useful in their lives and show them that you’re doing so.
Let’s get more specific. Segmentation can be as general or granular as your product requires (or as your budget allows). Some companies might have only three segments. Some might have dozens. There are some segmentation tactics that most companies have in common, but the right combination of tactics will always be unique to each business.
Some very common launch points for segmentation include:
- Demographic information
- Purchasing behavior
- Location or geography
- Usage behavior
- Professional affiliations
We’ll get into specifics later on in this article, where we’ll brainstorm survey questions as well.
Segmentation allows you to choose better groups not only for marketing, but beta testing as well. Product manager Ashton Rankin told us about how segmentation helped Kik build a balanced beta testing group for their social cryptocurrency.
On fine-tuning Kik’s beta testing team with survey segmentation:
“[W]e put in a survey to try to get users that fit some kind of criteria, whether it be for regulatory reasons or the type of users we wanted to test with because they were a representation of the largest percentage of our target users.
For a while our marketing team worked really hard to develop who the main demographics of Kik users were before we actually started figuring out who do we want in this program… As we worked with some of these users, we started to realize that some of them were more power users than the average user.
Their opinion is super valuable and you get great insights, but you also have to understand that when you’re testing your products that you might also want to go to a different person in your beta community or do a user test with someone that’s new to your app to see if the opinions are still the same as your power users.
Get to know the users you add to your beta program. If you are going to build and iterate on products with their feedback in mind you need to understand who they are and if they represent your target audience.”
The kinds of data you’ll need for powerful survey segmentation
To really let the data guide you, you’re going to need to do some research, both quantitative and qualitative. Quantitative research is one of the first places many companies start, but it’s not all you need — your users and their behaviors are more than numbers and statistics. If you’re not sure where to start with qualitative research, be sure to check out our qualitative research manual.
Don’t forget to test your findings regularly—unfortunately, user segmentation is not something you can “set and forget.” Check in with your data and see if you notice any changes or patterns forming. Remember that user habits and your app evolve over time, and you may find your users transitioning into different segments as their behavior changes. Create relevancy at every stage of your user’s journey by adapting to changes in your users’ behavior. Your app will evolve, just as your users do.
Here’s a deeper look into some of the details you might use to draw your segments. You might use just one variable or combinations of them to create segments.
- Demographics: statistical information about types of populations.
- Marital status
- Education level
- Purchasing history: What has their relationship with your app been like so far? What do buyers have in common? What about non-buyers?
- Never bought anything
- Bought once
- Repeat buyer (2 or more purchases)
- Physical address
- Current location
- Time zone
- Device Information: There are so many variables for optimization to consider here, and you’ll want to think about the ones that are most relevant for your app.
- Device type (phone, tablet or PC)
- Operating system
- App version
- Application Event: They might be positive, might be negative. Experiencing an event may influence your user’s emotional state at the time.
- User has just completed a level
- User achieved a difficult goal
- User matched with another user
- User Behavior: This is becoming one of the most effective ways to segment users because past behavior is a better indicator of future behavior than raw demographic data.
- Number of days between logins
- Times used per week
- Active or Inactive users
- User goals and how the app helps them achieve them
- Preferred features
- Usage patterns
- Preferred times of day
- Customer lifetime value
- Acquisition source
- Occupational Details: If your app is a service for professionals, knowing about their workplace will influence your development.
- User’s industry
- User’s job position
- Agency: decision-maker vs. specialist
- Company size
- Profile Information:
- Language preferences
- Personal interests
- Social media profiles
- User Frequency (this is just an example, these designations will vary depending on your app):
- Power user: 5+ times a week
- Average user: 2-4 times a week
- Light user: 1-2 times a week
- Disengaged user: Regular users who have disappeared recently
- Dormant user: Users who installed but never or rarely returned
You might look for segments based on your business priorities if you know what you want your users to do: identify your needs and goals, what types of users you want, and which metrics matter most to your app specifically.
Asking segment-specific survey questions
Once you’ve created segments, what do you do with them? Here are a few examples of ways you might break your customers into groups and some examples of what kinds of questions you might ask them or steps you might take — what works best for you will depend on what kind of app you have!
When creating survey questions, always have a goal in mind: what are you going to accomplish by getting the answer to this question?
Know who you’re going to ask; whether it’s a particular segment or persona, or several of them, if you’re trying to determine what differs some groups from others.
|Segment Type||Variable||Sample Questions or Ideas|
|Demographic||Age||Ask if certain parts of the app or game were difficult|
|Location||Live location||Send a targeted offer|
|Location||Home address or ZIP||Ask for reviews of nearby restaurants|
|Device information||Tablet||Invest in optimizing tablet experience|
|App Information||Has latest version||Ask what they think of your latest feature|
|Purchasing Behavior||Bought once||What could we do to improve your shopping experience?|
|Demographic||Income level||How would rate this app’s value for the money?|
|Demographic||Career field||What type of customer support communication methods do you prefer?|
|User Behavior||Hasn’t logged in in over 2 weeks||When is this app most useful to you?|
|User Behavior||Free user, hasn’t converted to paid||Are there any features you expected to find but didn’t?|
|User Behavior||Downgraded subscription||How would you rate the app’s value for money?|
|User Behavior||Upgraded from free to premium||What made you decide to upgrade your subscription?|
|Application Event||User has just experienced a happy in-app event||How likely are you to recommend this app to your friends? Would you be willing to leave us a review in the app store?|
Learning from your segments is a powerful way to personalize your user experience for specific groups, which will in turn make them feel noticed, acknowledged, and like your app adds special value to their lives. Once you know what certain groups of people have in common, you can use this information to predict future behavior or improve upon existing features. If you need more ideas about survey questions, you’ll find tons of templates in our in-app survey question guidelines.
By learning more about your segments, you can also create personas, an important part of getting to know your users—and once you do, you can use empathy to design delightful user experiences based on specific needs and desires. Dr. Lene Nielsen’s 10 Steps to Building Personas (available in multiple languages) is a simple step-by-step guide that clearly breaks down the process of persona creation from A to Z.
Personas are helpful because they help you understand your users on a personal, relatable level, and they might have different demographic details yet share values or expectations about your service. Personas might combine traits from multiple different segments and identify usage patterns across markets. Understanding your most common user personas will help you to build user journeys that make sense based on what they really need and experience from your service.
Survey targeting considerations
It’s not enough to just create a custom survey and send it off to a particular segment of users. For best results, approach your survey deployment with strategy. Keep the following considerations in mind:
Keep the user journey in perspective when sending surveys. Don’t prompt brand-new users for their opinions on your app; they don’t know what they think yet. Also be sure not to interrupt them while they’re in the middle of completing a task or otherwise occupied in-app.
Location and time of day
If you’re sending out a survey at a specific time, think about the time zone of the group you’re sending it to (here’s where knowing your user’s location is important information). Few people will respond to a survey received at 3:00 in the morning. But it goes deeper than time zones—some populations may use your app at different times of day or have certain times they’re more likely to answer surveys. For example, perhaps your US West Coast users are more likely to respond to surveys at 11 am, while your East Coasters might be online and active closer to 9 am local time. Maybe your British users are more likely to respond to a survey at 7 pm. This kind of information requires trial and error; once you find your magic window for each locale, take note.
If your user has just experienced a happy event in the app, that might be a good time to trigger an NPS survey asking how likely he or she is to recommend your app to a friend. If they respond with a rating of 9 or 10, they’re considered “promoters,” and now is a great time to link to the app store and ask if they’d like to leave a review. Tinder is an excellent example of this use case; the moment after a user has made a match, they’re more likely to feel positive about themselves and the app. This is a great time for Tinder to ask the user about their experience because this moment is a high point in the user’s journey.
If you’re curious about why large numbers of users are abandoning their shopping carts, for example, you could reach out to this particular segment of people and ask a question about how to improve the checkout process. You can do similarly for people who have downgraded or upgraded subscriptions, do or don’t use a particular feature, or drop off in usage. By reaching out to segments of people based on their behavior, you can learn more about areas where there is friction in your app and improve user experience issues.
They say if you fail, try, try again. That’s fine—try once. Instabug prompts users with surveys a maximum of two times, because we don’t want to interrupt their app experience. If a user has dismissed a survey twice, it won’t come back. And that’s a good thing because you don’t want to leave your users with a negative impression of your app experience.
Your users are more likely to respond to quick, short surveys than they are long, winding questionnaires that interrupt their experience. But if you’ve done your segmenting, this is actually a great thing. You can ask short, specific questions that you know will be relevant if you’re targeting the right segment. Progressive profiling is an effective way to collect detailed data about a group of people—every so often, they’ll see a new survey pop up asking for just a little bit more information—but never too much at once.
Don’t forget to close the feedback loop by following up on survey responses! Instabug users can immediately respond using in-app chat. By digging deeper to uncover their real thoughts and motivations, you’ll begin to understand your users and their needs better. You can even prioritize requests based on what you’re learning from your most valuable segments. If you’re not sure how to break down all of your feedback, you’ll find plenty of analysis techniques in our user feedback manual.