Main illustration: Rachel Denti
In part one of our best practice guide to understanding customer feedback, we looked at how to analyze and understand the types of feedback most important to your business.
Why is feedback analysis important?
Feedback analysis helps businesses understand customer needs and sentiments, as well as specific customer pain points. Feedback analysis is used to make more informed company-wide decisions and set the direction for product roadmaps.
7 steps to analyze customer feedback
But once you’ve decided which feedback you want to pay attention to, how do you transform customer feedback into something you can act on as a company? How can you take a jumble of open-ended feedback and use it to inform your product roadmap?
Follow these steps, and you’ll have a prioritised list of customer insights you can act upon with confidence. You can even use the output of your analysis to inform your product roadmap.
1. Collate your data
First, collate all the open ended customer feedback you want to analyze, plus key metadata about each customer, into a spreadsheet. Ideally, the metadata will include attributes such as how long the person has been a customer, how much they spend, date the feedback data was submitted, and the source of the feedback e.g. open ended customer survey question. Of course, you can use Intercom to help gather this data. Your column headings should look something like this:
2. How to think about categorising the feedback
A general rule that you can apply to help you make sense of customer feedback is to group it by:
- Feedback type
- Feedback theme
- Feedback code
Let’s break these down.
Categorising your feedback into different types is particularly helpful if you’re dealing with unclassified feedback from your customer support team or situations where customers could write anything they liked in a survey field (e.g. “Any other feedback for us?”)
Here are some categories you may find useful:
- Usability issue
- New feature request
- User education issue
- Generic positive (e.g. “I love your product!”)
- Generic negative (e.g. “I hate your product!”)
- Junk (this is useful for nonsense feedback like “jambopasta!”)
- Other (this is useful for feedback that’s hard to categorise. You can go back and recategorise it later as patterns emerge in the rest of the data)
Breaking feedback down into themes can be useful when you’re trying to make sense of a high volume of diverse feedback, so if your data set is small (roughly speaking, 50 pieces of feedback or less) then you may not need this.
The themes you come up will be unique to the actual feedback data you’ve received and will usually relate to aspects of the product. For example, let’s say you work on a popular product like Instagram and you’ve received a bunch of customer feedback. Your themes might look like a list of specific product features, like this:
- Photo stream
This type of categorisation is particularly useful when you’re working in a situation where you’re likely to have to feed your insights back to multiple teams to take action on (i.e. if you have one team that works on Stream, another on Stories, etc).
Sometimes themes can by team related (e.g. customer support, sales, marketing) or they could be related to unmet needs that customers are experiencing. Try coming up with some themes and see if these types of category are useful to you and the data you’re making sense of.
The purpose of the feedback code is to distill the raw feedback the customer has given you and rephrase it in a more concise, actionable way.
Your goal is to make the feedback code descriptive enough so that someone unfamiliar with the project can understand the point the customer was making. The feedback code should also be as concise and true to the original customer feedback as possible. Your job is to distill the feedback as objectively as possible, whether you agree with it or not.
Here’s an example:
3. Get a quick overview
You want to get a feel for the data before starting to codify it. Scan through the feedback to get a sense of how diverse the responses are. As a general rule of thumb, if each customer is giving you very different feedback, you’ll likely have to analyze a higher volume of feedback in order to see patterns and make it actionable. If you scan through the first 50 pieces of feedback and they all relate to a specific issue in your product, then you’ll likely have to review less.
4. Code the feedback
Time to roll up your sleeves and focus. Find a place you won’t be disturbed and start reading through each piece of user feedback, carefully coding each row.
The exact feedback codes you create will be specific to the product that the feedback relates to but here are a few analysis codes for some fictitious new feature requests to give you a flavour:
- Ability to assign a task to multiple clients
- Ability to add complex HTML to tasks
- Ability to add or remove teammates from any screen
- Ability to send emoji to clients
If one piece of feedback is communicating multiple points (e.g. two different feature requests), it’s useful to capture these two separate points in separate columns.
5. Refine your coding
Pay attention to the exact language people use
It’s OK to start with higher level codes and break them down later. Pay attention to the exact language people use. Issues that sound similar upon first glance might actually be separate issues.
For example, imagine you initially see a lot of customer feedback related to “Email issues”. However, when you read more feedback carefully, you realise that these break down into separate issues: “Email composer bug” and “Email delivery bug”, which are quite different.
Sometimes, as you read more feedback you realise that you need to break one popular code down into a couple of more specific codes. That’s fine, go ahead and split it up into sub-codes. For example, “More control over visual design” could be broken down into “Ability to add fonts” and “Ability to control alignment of images”. Remember to go back and recode the earlier rows.
6. Calculate how popular each code is
Once you’ve coded everything, the next step is to calculate the total amount of feedback per code. This will help you see which feedback is most common, and what the patterns are in your customer feedback.
One super simple way to do this is to sort the data in your “feedback type”, “feedback theme” and “feedback code” columns alphabetically, which will group similar items together. Then highlight all cells that have the same feedback code and a total count will appear in the right hand corner of your spreadsheet. Create a summary table to record all the total counts for each feedback code.
If you have between 100-500 pieces of feedback, add a new column next to your “Feedback code” column, and enter a “1” for each row that has the same feedback code (e.g. add a 1 next to all cells that say “Ability to crop image”. Then add up how many times that code appears. Repeat for the other feedback codes.
If you have a larger data set, you can create a pivot table to do these calculations. With large data sets, it’s also valuable to dig deeper at this point and analyze the other customer attributes that you collected in the spreadsheet (e.g. customer type, customer spend) and look for other correlations with the feedback you’ve received. For example, which customers are complaining most about X? What’s the monthly spend of the customers demanding X new feature?
7. Summarise and share
Now you’ve coded your data, you can create a summary of customer feedback data based on issue popularity and discuss it with your team.
If you’ve got 50 pieces of feedback or less, you can summarise actionable feedback in a simple table or one page doc. If you have a larger set of feedback, you can break the data down by the other variables we discussed earlier (“feedback type” and “feedback theme”). This will make it much easier for you to take the different buckets of feedback you’ve identified and channel them to different people in your company who can take action on the feedback.
One of the most powerful things you can do with customer feedback is to create a Top 10 list of feature requests or Top 10 customer issues that you can then use to inform your product roadmap.
It can be hard to know how to go about analysing customer feedback, especially if you don’t have researchers or analysts at your company who can help. However, if you follow the advice in this post, anyone can turn a jumble of customer feedback into a clear summary. Best of all, you can then use that summary to make informed decisions in your company and, in turn, improve your products.