What is Verbatim Coding?
Verbatim coding is the process of organizing and categorizing open-ended responses by assigning specific themes, or “codes” to the text.
This technique helps to identify patterns, trends, and key insights within qualitative data like open ended survey responses, customer reviews, or social media comments and turn them into quantitative data that can be transformed into charts, graphs, and actionable insights.
By systematically coding the responses, researchers can easily analyze large volumes of text and transform unstructured data into meaningful, actionable insights that drive business decision-making.
For example, several open-ended answers to the question “What’s your guilty pleasure snack?” could be coded with “Chocolate Obsession”:
This is helpful information for a market research firm because it shows that their client’s audience responds well to chocolate products.
The process of coding verbatim responses like this used to be done manually using tools like Excel spreadsheets.
As you can imagine, manual coding is a time-intensive and expensive process. You have to read through every single response to extract a codeframe, and then go back through the responses one by one to assign codes to each.
It requires several rounds of revisions and reviews to avoid data errors and human biases.
Because of this, market researchers and consumer insights analysts are often drowning in an ocean of open-ended text responses, lacking the time or budget to sift through all that data. As a result, many avoid using this valuable data altogether, missing out on critical insights and untapped potential. It's a tedious process that requires patience, care, and attention to detail, but the missed opportunities can be even more costly.
Luckily, AI verbatim analysis software like Blix exists to help you speed up the verbatim coding process.
Simply upload your spreadsheet of open-ended responses, select the columns you want to analyze and review the suggested codebook.
After the software does its job, it will create a report with a chart showing the code distribution, a summary of each theme, direct quotes from the data, and a fully coded data file.
Here’s a video overview of this process:
The Verbatim Coding Process
Whether you decide to code your qualitative data manually or use software like Blix, here are the steps to coding verbatim:
Step 1: Collect The Data
You can’t analyze data you haven’t collected! You can collect data via surveys, customer reviews, customer support chats, or even social media or forum comments.
Gather the text verbatim into a spreadsheet.
Step 2: Organize And Clean The Data
Once you have the data in a spreadsheet, it’s time to clean and organize it.
Take the time to delete any duplicate or irrelevant responses and begin to look for common themes and trends.
Step 3: Create A Codebook & Define Your Themes
As you’re reviewing the data, you’ll likely see common words, phrases, and themes in the responses.
This is how you begin to code the verbatim responses. Create themes — words or phrases that describe what the response is about — and start to mark the responses with those initial themes, or “codes”.
For example, let’s look at a coding project that uses the question, “What’s your guilty pleasure snack?”.
Your codebook, then, is a list of all the different codes or themes you created to categorize the data.
Here’s what our codebooks look like at Blix:
Blix can automatically generate these themes and create a codebook for you using AI-powered text analysis, saving you a ton of time and headaches.
Step 4: Assign Codes
Once you have a list of all the different themes in your data, it’s time to go through and assign those codes to each response.
This is the most time-consuming part of verbatim coding. You have to carefully pour over each result and categorize it, being careful to avoid biases or inaccuracies.
Once you’re finished, you may need to review it several times and have more than one person look over it. You may even want to involve your customer(s) in the process, to make sure that you create informative & actionable codes that relate to your target audience.
Again, this is the power of AI verbatim coding software like Blix. It can create your codebook and assign your codes to the data automatically with more accuracy and in far less time. It also allows you to recode if changes are needed, without having to manually review all of the coding from scratch.
Step 5: Analyze The Results
Finally, we get to turn all that hard work into measurable, actionable insights.
You can pull all of this data to create visual charts that show the percentage of responses tagged with each type of code to see trends and gain insight into what your customers like or dislike about your product.
Blix is able to generate a report showing the breakdown of each code, a summary of what each code means, and direct quotes from the verbatim comments to see exactly what your survey participants are saying.
Benefits of Verbatim Coding in Market Research
So why spend the time, money, and effort coding verbatim responses?
I’ve already mentioned that verbatim coding helps you to understand the WHY behind the numbers. Open-ended comments give you far more data and understanding of how customers think and their pain points and desires than simple “yes or no” questions.
Beyond that, coding gives you a deep understanding of customer sentiment and actionable insights to enhance your business.
Challenges Verbatim Coders Face
Verbatim coding is a tedious and time-consuming process, prone to mistakes and human biases if not handled carefully.
It can also be difficult to manage extremely large datasets with thousands of open-ended survey responses. And sometimes answers are in multiple languages, making the process even more difficult.
One solution to some of these problems is hiring out coding services from a verbatim analysis company.
While this frees up time internally, it can still be expensive and prone to human biases. Plus, by hiring out the process, you don’t get as strong of a connection with your customer’s responses and it requires a lot of back and forth communication with the service provider.
So what should you do instead?
