Analyze your research

In today’s lesson, we’re going to show you how to analyze your research data.

Remember how we set up the email surveys and website polls at the beginning of the course?

Well, that data has been trickling in over the past few weeks.

Now it’s time to analyze those responses!

By the end of the lesson, you’ll know:

  • How to download your email survey and website poll data
  • How to analyze your data to find common themes
  • An alternative way of using AI to help

Ready? Let’s dive in…

Video:

Notes:

By this point, you should have conducted the following research:

  • Analytics
  • Email surveys
  • Website polls
  • Heatmaps
  • Become the customer
  • Journey mapping
  • Competitor analysis
  • User tests
  • Social proof

And put all your findings in your research document, except:

  • Email surveys
  • Website polls

These spreadsheets need to be analyzed.

The email surveys and website polls are written responses in a spreadsheet that needs to be analyzed to find common themes.

Here’s how to analyze the written responses:

  • Download your data in a spreadsheet
  • Read each response
  • Identify common themes
  • Examples: too expensive, free shipping, easy to use, etc.
  • Add a new column for each theme. Type “1” in the column when it’s mentioned. Count the totals at the end.
  • Update your research document

AI alternative

  • Use AI (e.g. ChatGPT, Claude, MonkeyLearn, etc.) to identify common themes.
  • Upload the spreadsheet to AI to analyze the data.

Note: AI is fast, but their summary may not be accurate. We recommend reading each response yourself to truly understand your visitors.

Use this AI prompt:

Objective: To extract and categorize key themes from customer survey responses, including likes, dislikes, questions, and objections.

Sentiment Analysis:

  • Identify the sentiment of each survey response (positive, negative, neutral).

Theme Extraction:

  • Extract common themes related to:
    • Likes: Positive sentiments, features, or experiences that customers appreciate.
    • Dislikes: Negative sentiments, pain points, or aspects that customers find challenging.
    • Questions: Queries or uncertainties expressed by customers.
    • Objections: Stated concerns or objections that need addressing.

Keyword Identification:

  • Identify and compile keywords associated with each theme to provide context and depth to the analysis.

Categorization:

  • Categorize responses into broader segments (e.g., product features, customer service) to understand the context of the feedback.

Frequency Analysis:

  • Determine the frequency of each theme to highlight the most commonly mentioned aspects.

Contextual Understanding:

  • If applicable, understand the context of sentiment shifts within a single response (e.g., positive about features but negative about pricing).

Additional Insights:

  • Provide additional insights into any emerging patterns or unexpected themes that may not fall into predefined categories.

Recommendations:

  • Based on the analysis, suggest actionable recommendations to address identified issues or capitalize on positive aspects.

Notes:

  • Utilize natural language processing (NLP) techniques for accurate sentiment analysis and theme extraction.
  • Consider the nuances of language, such as sarcasm or context-dependent meanings.
  • Iterate the analysis as needed, refining the model for ongoing improvement.

Add all feedback in your research document, so you can reference it later.

Comments:

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *

Video transcript:

Diving into Audience Research Analysis

Analyze your research. Congratulations. By this point, you should have conducted all of the following research. You’ve collected analytics, email surveys, website polls. You’ve looked at heat, click, and scroll maps. You’ve become the customer yourself. You’ve conducted a journey mapping session. You’ve also done competitor analysis, run several user tests, and compiled all of your social proof into one doc for easy access.

At this point, all of your research should be in your research document except for your email surveys and your website polls. 

Both of these will be in CSV format. So you need to put them into a spreadsheet and analyze them. This is going to be huge for you to have a deeper understanding of how your visitors operate within your website and become your customers.

It also gives you an easy way to share these insights with the rest of your team.

Step-by-Step Guide to Analyzing Survey Responses

Here’s how you analyze all of the written responses. First, you’re going to download all of the data from your polls and your surveys into a spreadsheet. Go ahead and read each response, and then you’re going to identify the common themes among them.

So, for example, you might see several people that are saying something along the lines of it being too expensive, or how they really like the free shipping, or that something is really easy to use. Every time you come across a common theme, you’re going to add a new column into your spreadsheet, and then you can enter a 1, or a Y, or an X, or whatever the heck you want.

You can put a marker in there so that you can count the number of times that that’s mentioned for each of those different themes.

Then you can simply count the totals at the end and add your insights into your research document.

Here’s a quick example. What you see here on my screen would be the results of an email survey. As you can see, we’ve got a column for the email address, and then every other column to the right, column C, D, E, and F, are various questions that you can ask.

For a full list of those questions, be sure to refer back to the email survey section of this course. 

So, let’s take column C for an example. The question is, what do you like most about our offer? And you see that there are a bunch of different responses. It’s easy to use, there’s great value, it’s super fast, sleek design, amazing features, love the colors.

All these are different responses. So here’s how you’d go through and analyze that.

All you have to do is copy that question into a separate tab in your spreadsheet, and then as you identify common themes, just add a new column. So you can see here, some of the themes that show up are it’s easy to use, it’s affordable, it’s fast, the design, or the features. Then as you go through each of the responses, you just mark a 1 under any column that is applicable.

So the first one is easy to use, so shockingly, that goes under the easy to use column. Great value would go under affordable, super fast would go under fast, etc. 

Over time, you’ll start to notice that some themes are more popular than others, and that’s why you have a total that’s at the very bottom.

So very quickly, we can see here at the responses, if you look at the total row at the very bottom in yellow, affordable is mentioned more than anything else. So we know that that’s something that people really love about this particular offer. This is a great insight to have. Now you can go ahead and do this for this question, as well as all of the other questions that you have, both in your email surveys, as well as your website polls.

And remember, once you get that content, you’ll want to add all of these insights directly into your research doc.

Leveraging AI for Data Analysis: Pros and Cons

Now, you may be thinking, all right, Adam, I get it. This seems like a lot of work. I have to go through a spreadsheet. I have to put in a bunch of numbers. I have to read all these responses. Surely, AI could just do this for me. And the answer is that it can. You can use tools like ChatGPT, Claude, or MonkeyLearn, for example, which will analyze all these responses and then break it down for you and identify those common themes.

This is great if you absolutely need something done in like the next five minutes. However, I would caution against relying on this too much. In my experience and what I’ve seen from others, the summary is fast, but it’s not necessarily accurate. It may be off a bit. So in the spirit of accuracy, you definitely want to do this yourself.

The Value of Personal Analysis and Final Thoughts

More importantly though, there’s something internal that happens to you when you go through and you read all of these responses, you really internalize the thoughts, the questions, the concerns, the desires that people have, you internalize all of that so that when you start to improve your product or change your web designs or write copy, you have that not just as some like abstract notion in your head, but it’s in your bones.

It’s in your muscle memory. 

So that you can write better copy, you can improve your products because you know exactly what people are thinking, you know the exact words that they’re using, and you know how to improve your products, your websites, and your copy based on those insights. So yeah, if you want to do AI, you certainly can, but it’s really not that hard to go through and analyze the responses yourself.

Trust me, it’s worth it.

And once you’ve gone through and done all this research, Be sure to go and add all of your feedback into your research document so you can reference it later.

On this page: 

  1. Video
  2. Next steps
  3. Notes
  4. Transcript

Need help?