8 Tips on How to Use Feedback Data Effectively
Mobile is the new battleground.
Mobile is the place where users go for convenience and speed, so make your mobile experience count.
Developing a product and growing a business is hard yet it’s even more difficult when users start to complain. Although nobody likes to get negative comments, such feedback is actually a good thing as it can help you to improve your business and take it to the next level.
The trick is to know how to use such feedback data to make real changes in your company.
The best method to understand your users’ needs and to respond to those needs in your app or a website is to collect user feedback. You should gather feedback on mobile to optimize user experience and gain an advantage over competition by giving your users what they want.
Many companies are abundant with feedback data yet stay unsure how to leverage such data for their consistent growth.
WHAT ARE USERS INSIGHTS?
User insights are interpretations of the data gathered from user feedback and other sources, compiled and analyzed to support business decisions better.
The main goal is to identify behavioral trends to improve your product/services and to make your sales and marketing more effective.
Getting insights from user feedback gives you a chance to understand your users better, to improve their digital experience and at the same time, to increase your revenue.
Analysis of such feedback involves identifying the needs or issues of your users so that you can improve user satisfaction and reduce churn rates.
HOW TO UTILIZE USER FEEDBACK?
Before analyzing, make sure you have gathered all the feedback in one place.
There are some common ways on how to act on user feedback and turn it into your biggest strength.
Here’s how you can turn user feedback into an advantage for your business:
Collect and Consolidate Your Data
Firstly, sort all user feedback that you want to analyze into a spreadsheet and add key metadata about each user (you can include how long a person has been a user, how much they spent, the date and source of the feedback etc.).
Your table could look similar to this:
After you collect the feedback, such systematization will consolidate incomplete or inconsistent data. There are many different segments of users and grouping them per different aspects (frequency etc.) will be helpful as each user has their own expectations from your company.
Upon carefully examining the data, discover common themes as well as repetition of the feedback. For example, if many users from different segments complain about the same thing (e.g. onboarding process etc.), then you should act on it.
Categorize Feedback
If you take your time to review the comments and try to uncover patterns within responses by categorizing them, you will get even more valuable insights.
Once you start reviewing the comments, you will notice categories which can contain elements like speed, availability, registration, price and so on.
Create categories and review the comments again to mark them with an appropriate category. If a single comment refers to more than one category, you can divide it into parts and mark each part of the comment.
Expect that comments will be both positive and negative and you can start dividing them into matching columns in your table (e.g. Excel) with proper sections.
Some common rule to help you to understand the feedback is to group it by:
Feedback type
Categorizing feedback into different types is helpful if dealing with inconsistent feedback from your customer support or customers who wrote anything they wanted in your survey field.
Some common categories that can be useful are:
- usability issues
- new feature requests
- bugs
- pricing
- generic positive or
- generic negative comments
- junk (nonsense feedback) or
- other (for feedback hard to categorize where you can later come back and re categorize if any patterns appear within the rest of the data)
Feedback theme
Sorting out feedback per theme can be useful if you’re trying to make sense of the high volume of different feedback. If you have a small data set (around 50 pieces of feedback), you may not need it.
The themes you create will be unique to the feedback you get and will relate to the aspects of your product.
For example, if you received a lot of user feedback, your themes can look like a list of product features such as:
- Profile
- Mentions
- Newsfeed
- Onboarding
- Etc.
Such categorization can help you even more when you have to provide your insights to more than one team (if one team works on mentions and another on your profile etc.).
You can also create team-related themes (marketing, sales, customer support etc.) for better efficiency — try to create some themes and see if such category types can be useful to your teams.
Feedback code
Feedback code is necessary to clarify unsorted feedback from users and to rephrase it in a more actionable way.
Create the feedback code detailed enough so that even an inexperienced person can understand the subject of the feedback.
The feedback code should be concise and matching user feedback as possible — you must refine the feedback whether you like it or not.
An example can be:
In order to uncover such details, you should segregate your feedback data. Common practice is to create a table with Excel or Google Sheets or any other tool you find suitable and import all your feedback data.
Your feedback will come in both structured (ratings or ranking indicators) and unstructured form (comments, complaints etc.), so you can also add a category to help you consolidate your data more efficiently.
You can make a list of all complaints or praises in one place.
After you fix it and test it, update the app and request more feedback again.
When new feedback arrives, assess the situation again for better insights — if there’s no improvement then the issue might be elsewhere, and if the problem is resolved, it’s a win for you.
This process is repetitive — analyze the data, fix specific issues, and repeat the feedback loop until everything is working properly.
The better you are at gathering and analyzing feedback data on a regular basis, the more effective the entire process will be.
Get a quick overview
Before you start coding your feedback data, you should review the feedback to understand the variety in responses.
If you review feedback and end up with almost each customer giving different feedback, you will most likely have to analyze a higher volume of data to discover the patterns.
However, if you scan through the first batch of feedback and the majority relate to a particular issue within your app, you will have to review less.
Code the feedback
There are different ways to code your feedback but the usual manual procedure is to create two spreadsheets — one to keep your code frame and the other for coding each piece of feedback per specific code frame.
You can structure your feedback by code frame to separate positive from negative comments or you can code each piece of feedback per several codes from the code frame.
Another method of manual coding is to use Feedback Types as one of hierarchy levels such as usability issues, bugs, new feature requests etc.
When you go through the entire feedback and carefully code each row, your exact feedback codes will be specific to the product/service that the feedback relates to.
For example, a few analysis codes you can use to a new feature request can be:
- The ability to send emojis
- Assigning tasks to multiple clients
- Onboarding process
If any piece of feedback talks about multiple points (two different feature requests etc.), it is useful to capture these two separate points in separate columns.
Whether positive or negative feedback, if you spot certain words that appear frequently (e.g. negative: bad experience or positive: easy to use etc.) it will instantly indicate widespread use.
It can also help to color your spreadsheet to easily see the emerging patterns or to run a sentiment analysis and get an overview whether customers are mainly satisfied, neutral or dissatisfied.
Refine Your Coding
You can start with higher level codes and break them down later.
A good thing is to focus on the exact language people use as sometimes, similar issues can turn out to be quite different issues after all. As you go through more feedback, you can realize that you have to break a single popular code into several pieces of more specific code.
For example, you realize you have a lot of user feedback related to Email Issues, but as you read more, you realize that these can be divided into separate issues like: ‘Email delivery bug’ or similar which are quite different.
Always remember to go back and recode the earlier codes if necessary.
Analyze the Popularity of Each Code
Once you’ve finished coding your feedback, the next step can be to determine the total amount of feedback per a single code — it will help you to see what kind of feedback is the most common and which patterns exist in your user feedback.
A great way to do this is to sort the data into ‘feedback type’, ‘feedback topic’ and ‘feedback code’ columns and group similar items together. After that, you can highlight all items with the same feedback code and a total count will appear — create a summary table to record all total counts for each feedback code.
If you have a large data set (more than 500 pieces of feedback), you can create a pivot table for these calculations — it is valuable to analyze the other user attributes that you collected. For example, put your user attributes (user type, etc.) into a spreadsheet and look for other correlations you received e.g. what is the monthly spend of users demanding a new feature?, and similar.
Summarize and Share
You’ve coded your data and now you can create a summary of user feedback data based on the popularity and discuss it with your team.
If you have little feedback (less than 50 pieces), you can summarize actionable feedback in a simple table or a document.
A large set of feedback can be broken down by other variables like ‘feedback type’ or ‘feedback topic’ etc.
Such practice will make it much easier for you to identify different groups of feedback and channel them to different people in your team who should act on it.
Keep in mind that the most powerful thing you can do with feedback is to prioritize.
You can create a top 10 user issues that you can then use to adjust your product roadmap.
Take Action
Implementing feedback successfully is a significant part of product development.
If poorly managed, user feedback can divert you from the real problems and you can end up wasting time and resources.
However, if you manage your feedback properly, you can come up with new ideas and plan your improvements more effectively.
Summarize previously reviewed feedback into a short document and share it with your team for further planning.
An effective method is to try to incorporate user feedback at each stage of the development process leading to the final ‘thank you’ email. Never forget to thank your users no matter if the feedback is positive or negative.
Unstructured insight is the one that will help you to build a better product but other data can help as well, like product purchase history, lifetime value and more.
AUTOMATE USER FEEDBACK ANALYSIS WITH 3RD PARTY SOFTWARE
There are many specialized solutions for automating user feedback analysis where some are more DIY than others.
Some of these solutions are:
- Text analytics solutions like Lexalytics, RapidMiner, Thematic and similar.
- Qualitative data analysis solutions like NVivo, Dovetail, ATLAS.ti, Quirkos etc.
- Voice of customer (VoC) platforms like Medallia, Qualtrics, Clarabridge, Confirmit and more.
You should check each solution and confirm its features and if they correspond to your needs for analyzing the feedback. For example, some tools have both manual and automatic methods for coding feedback, while others can unify user data no matter the feedback channel etc.
Check each of the solutions and decide which one suits your needs best.
FINAL WORD
Without a feedback process in place, your company might overlook some apparent mistakes which can prevent you from achieving the revenue you want. When it comes to feedback, the more details you have, the better.
Start with consolidating the data for a preliminary analysis, look for patterns and correlations and then tackle the issues — afterwards, test until you notice the improvement in feedback.
In case you get stuck, you can always use some tools that can give you further insight.
Keep in mind that the feedback process is a never ending process. You should ensure that your feedback process is a loop — listen to your users, make improvements and continue to collect feedback.
