faq & Anleitungen
FAQ and guides
Why do I even need to code my interviews?
Basics
If you’re currently working on your bachelor’s thesis, master’s thesis, or another research project and have collected qualitative data – for example, through expert interviews, guided interviews, or open conversations – you’re probably facing the big question: 👉 “How do I analyze these interviews? How do I get from my transcripts to insights that answer my research question?”
This is where coding comes in. But what exactly does that mean – and why is it so important?
This is where coding comes in. But what exactly does that mean – and why is it so important?
What does “coding” interviews mean?
Coding is about systematically structuring qualitative data. This means: you carefully read through your interviews, highlight important statements, and assign them to specific categories or themes – the so-called codes.
The goal is to create an organized picture from a large amount of unstructured text (e.g., interview transcripts) that shows:
✅ Which topics appear most frequently in the interviews.
✅ How the participants’ statements relate to your research question.
✅ What patterns or differences emerge between different interviews.
In short: Coding helps you make qualitative content tangible and derive meaningful, evidence-based conclusions.
The goal is to create an organized picture from a large amount of unstructured text (e.g., interview transcripts) that shows:
✅ Which topics appear most frequently in the interviews.
✅ How the participants’ statements relate to your research question.
✅ What patterns or differences emerge between different interviews.
In short: Coding helps you make qualitative content tangible and derive meaningful, evidence-based conclusions.
Why is coding so important in qualitative research?
Especially in qualitative methods – such as Grounded Theory or Mayring – coding is a central step.
📌 Grounded Theory: You develop theories directly from the data by creating codes and gradually refining them into concepts.
📌 Mayring: In this structured content analysis, you work with a system of categories that is either predefined (deductive) or developed from the data (inductive).
Both approaches aim to reduce the complexity of qualitative data, make it comparable, and find precise answers to your research question.
Without coding, it would be like trying to find all relevant passages in a huge book – without bookmarks or a table of contents.
📌 Grounded Theory: You develop theories directly from the data by creating codes and gradually refining them into concepts.
📌 Mayring: In this structured content analysis, you work with a system of categories that is either predefined (deductive) or developed from the data (inductive).
Both approaches aim to reduce the complexity of qualitative data, make it comparable, and find precise answers to your research question.
Without coding, it would be like trying to find all relevant passages in a huge book – without bookmarks or a table of contents.
Coding means: Ensuring quality & making themes visible
Through coding, you can:
🌟 Create thematic categories – e.g., “product satisfaction,” “criticism of the service”
🌟 Represent frequencies – how often a topic was mentioned in the interviews
🌟 Identify relationships – which themes are connected or contradict each other
This transforms what was initially “just” text into a reliable foundation for your analysis – and often also builds a bridge to quantitative representation (e.g., how frequently a theme was mentioned).
🌟 Create thematic categories – e.g., “product satisfaction,” “criticism of the service”
🌟 Represent frequencies – how often a topic was mentioned in the interviews
🌟 Identify relationships – which themes are connected or contradict each other
This transforms what was initially “just” text into a reliable foundation for your analysis – and often also builds a bridge to quantitative representation (e.g., how frequently a theme was mentioned).
Coding is important – but often time-consuming
Manual coding takes time. Normally, you would:
🕒 Read through your interviews multiple times
🕒 Create a codebook
🕒 Assign all relevant statements to codes by hand
That’s laborious – and this is exactly where Cody supports you:
👉 Cody automates the coding process based on established scientific methods (e.g., Mayring, Grounded Theory)
👉 Cody saves you many hours of work – and gives you a solid first assessment of which topics in your interviews are most important
👉 Cody is specifically designed for qualitative research – unlike general AI tools, which often lack the precision needed for coding
Sign up for Cody now and speed up your analysis!
🕒 Read through your interviews multiple times
🕒 Create a codebook
🕒 Assign all relevant statements to codes by hand
That’s laborious – and this is exactly where Cody supports you:
👉 Cody automates the coding process based on established scientific methods (e.g., Mayring, Grounded Theory)
👉 Cody saves you many hours of work – and gives you a solid first assessment of which topics in your interviews are most important
👉 Cody is specifically designed for qualitative research – unlike general AI tools, which often lack the precision needed for coding
Sign up for Cody now and speed up your analysis!
Coding is the key to successfully analyzing your interviews
Without coding, qualitative data remains difficult to evaluate. With a structured coding process, you can:
✅ Directly answer your research question
✅ Make qualitative content visible and traceable
✅ Draw scientifically grounded conclusions
With Cody, it’s faster, easier, and more reliable – try it out now:
👉 Start coding your interviews with Cody
If you’d like to know how to interpret the results, you’ll find more information in the next section.
✅ Directly answer your research question
✅ Make qualitative content visible and traceable
✅ Draw scientifically grounded conclusions
With Cody, it’s faster, easier, and more reliable – try it out now:
👉 Start coding your interviews with Cody
If you’d like to know how to interpret the results, you’ll find more information in the next section.
Interpreting the Results: How to Read Your Analysis with Cody
The Initial Results at a Glance
Right after the analysis, Cody shows you:
✅ Type of analysis – We always evaluate based on Mayring + Grounded Theory.
✅ Number of sources – An overview of the data material used.
✅ Date of analysis
✅ Engine version – For reference and citation purposes
✅ Your research question – So you always see the results in the context of your original inquiry.
✅ Type of analysis – We always evaluate based on Mayring + Grounded Theory.
✅ Number of sources – An overview of the data material used.
✅ Date of analysis
✅ Engine version – For reference and citation purposes
✅ Your research question – So you always see the results in the context of your original inquiry.
Categories in Cody: How Content Is Structured
Our software automatically assigns all content to one of three main categories:
1️⃣ Relevant Classifications
Here you’ll find all coded sentences that directly help answer your research question. These form the core of your analysis and should be interpreted with special attention.
2️⃣ Irrelevant Classifications
This includes parts of the conversation such as small talk, transitions, or other segments that are not directly related to the research question. Cody helps you clearly separate these elements to keep your analysis focused.
3️⃣ Moderation Classifications
This category includes elements like open questions, follow-up questions, or other moderating remarks by the interviewer. They help you better understand the flow of the conversation.
1️⃣ Relevant Classifications
Here you’ll find all coded sentences that directly help answer your research question. These form the core of your analysis and should be interpreted with special attention.
2️⃣ Irrelevant Classifications
This includes parts of the conversation such as small talk, transitions, or other segments that are not directly related to the research question. Cody helps you clearly separate these elements to keep your analysis focused.
3️⃣ Moderation Classifications
This category includes elements like open questions, follow-up questions, or other moderating remarks by the interviewer. They help you better understand the flow of the conversation.
Interactive Visualization: Your Sunburst Diagram
The Sunburst Diagram gives you a visual overview of the thematic structure:
🔸 You can hover over the segments to see what percentage of the data each theme represents.
🔸 Click on a segment to dive deeper into its subcategories and analyze their share of the overall data.
👉 This feature makes it especially easy to identify patterns and key focus areas.
🔸 You can hover over the segments to see what percentage of the data each theme represents.
🔸 Click on a segment to dive deeper into its subcategories and analyze their share of the overall data.
👉 This feature makes it especially easy to identify patterns and key focus areas.
Table View: Structured Numbers & Facts
In the Table View, you’ll find:
📌 The main categories, which can be expanded
📌 The corresponding subcategories – including:
- Number of mentions
- Percentage of the overall corpus
This gives you a precise, quantifiable foundation for your analysis.
📌 The main categories, which can be expanded
📌 The corresponding subcategories – including:
- Number of mentions
- Percentage of the overall corpus
This gives you a precise, quantifiable foundation for your analysis.
Document Viewer: Context Directly in the Text
In the Document Viewer, you can switch between your uploaded sources and examine them in detail:
✅ Color highlights show you directly in the text which sentences are relevant, irrelevant, or part of the moderation.
✅ A legend helps you quickly understand the color codes.
This way, you always have the full context in view.
✅ Color highlights show you directly in the text which sentences are relevant, irrelevant, or part of the moderation.
✅ A legend helps you quickly understand the color codes.
This way, you always have the full context in view.
The Results Section: Interpretation Made Easy
Cody divides the results into four areas:
1️⃣ Key Findings – A brief summary of the most important points.
2️⃣ Analysis in Relation to the Research Question – How your data answers the question.
3️⃣ Interpretation & Implications – What the results mean and what conclusions can be drawn.
4️⃣ Possible Limitations – Notes on the limits of interpretability.
➡ With one click, you can export the results and use them in your own work.
Here is a sample excerpt from the analysis of the example project.
1️⃣ Key Findings – A brief summary of the most important points.
2️⃣ Analysis in Relation to the Research Question – How your data answers the question.
3️⃣ Interpretation & Implications – What the results mean and what conclusions can be drawn.
4️⃣ Possible Limitations – Notes on the limits of interpretability.
➡ With one click, you can export the results and use them in your own work.
Here is a sample excerpt from the analysis of the example project.
Why Cody Makes Your Analysis Easier
Our method is based on proven scientific approaches such as:
📌 Qualitative content analysis according to Mayring – clear structuring along the research question.
📌 Grounded Theory (Strauss & Corbin) – systematic development and linking of categories.
👉 Cody automates this process and gives you a well-founded initial analysis – fast, precise, and easy to understand.
Use Cody for your evaluation!
🚀 Save time and improve the quality of your analysis:➡ Start with Cody now and get structured results for your interviews!
📌 Qualitative content analysis according to Mayring – clear structuring along the research question.
📌 Grounded Theory (Strauss & Corbin) – systematic development and linking of categories.
👉 Cody automates this process and gives you a well-founded initial analysis – fast, precise, and easy to understand.
Use Cody for your evaluation!
🚀 Save time and improve the quality of your analysis:➡ Start with Cody now and get structured results for your interviews!
How Does Cody Work – and What Should You Pay Attention To?
How Cody Works – Step by Step
✅ Purchase Credits
Cody gives you full flexibility: simply buy credits and use them for your analyses.➡ 1 credit = 1 interview up to 60,000 characters
If your interview is longer, an additional credit is automatically charged. In practice, a 45-minute interview usually contains around 40,000–50,000 characters.
✅ Upload Your Interview
Upload your interview transcripts or other text documents – it’s just a click away.
✅ Enter Your Research Question
Formulate your research question clearly and precisely. It forms the basis of the analysis.👉 Here’s how to phrase your research question for Cody
✅ Start the Analysis
Cody shows you a preview – then you confirm the analysis.Within just 5–10 minutes, you’ll receive the results, even for a batch of 10 interviews.
Cody gives you full flexibility: simply buy credits and use them for your analyses.➡ 1 credit = 1 interview up to 60,000 characters
If your interview is longer, an additional credit is automatically charged. In practice, a 45-minute interview usually contains around 40,000–50,000 characters.
✅ Upload Your Interview
Upload your interview transcripts or other text documents – it’s just a click away.
✅ Enter Your Research Question
Formulate your research question clearly and precisely. It forms the basis of the analysis.👉 Here’s how to phrase your research question for Cody
✅ Start the Analysis
Cody shows you a preview – then you confirm the analysis.Within just 5–10 minutes, you’ll receive the results, even for a batch of 10 interviews.
What Should You Pay Attention To When Using Cody?
⚠️ Data Quality
Cody follows the principle: “What goes in is what comes out.”If your data is incomplete, poorly transcribed, or lacks substance, Cody cannot produce high-quality results from it.👉 Prepare your data carefully.
👉 Here’s the guide: How to prepare your data properly
⚠️ Automation Is Not a Substitute for Your ExpertiseOur analysis is fundamentally based on the following concepts:
📌 Qualitative content analysis according to Mayring
📌 Grounded Theory by Strauss & Corbin
📌 and Hsieh & Shannon (2005)
⚠️ Exploratory, Not Final
Cody’s analysis reveals initial patterns, frequencies, and thematic focuses – but it is not a replacement for a theory-driven, manual analysis by professionals.
⚠️ Cody Does Not Replace a Thorough Manual Content Analysis
The automated analysis is based on statistical and semantic models – not human interpretation.Category formation and coding are rule-based and data-driven – they offer a starting point to help you recognize patterns and focus areas.
The results should always be critically reviewed, validated, and if necessary, supplemented or adjusted using your own coding framework.
➡ Cody provides impulses and orientation – but the final scientific analysis, according to the requirements of your institution or organization, remains your responsibility.
Cody follows the principle: “What goes in is what comes out.”If your data is incomplete, poorly transcribed, or lacks substance, Cody cannot produce high-quality results from it.👉 Prepare your data carefully.
👉 Here’s the guide: How to prepare your data properly
⚠️ Automation Is Not a Substitute for Your ExpertiseOur analysis is fundamentally based on the following concepts:
📌 Qualitative content analysis according to Mayring
📌 Grounded Theory by Strauss & Corbin
📌 and Hsieh & Shannon (2005)
⚠️ Exploratory, Not Final
Cody’s analysis reveals initial patterns, frequencies, and thematic focuses – but it is not a replacement for a theory-driven, manual analysis by professionals.
⚠️ Cody Does Not Replace a Thorough Manual Content Analysis
The automated analysis is based on statistical and semantic models – not human interpretation.Category formation and coding are rule-based and data-driven – they offer a starting point to help you recognize patterns and focus areas.
The results should always be critically reviewed, validated, and if necessary, supplemented or adjusted using your own coding framework.
➡ Cody provides impulses and orientation – but the final scientific analysis, according to the requirements of your institution or organization, remains your responsibility.
Literature Basis & Citation
Cody’s analysis is based on proprietary systems and algorithms developed by us. However, it draws on insights and frameworks from the following key academic publications:
Mayring, P. (2000). Qualitative Inhaltsanalyse: Grundlagen und Techniken. Weinheim: Beltz.
Mayring, P. (2014). Qualitative content analysis: theoretical foundation, basic procedures and software solution. Klagenfurt.
Strauss, A., & Corbin, J. (1990). Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Sage.
Hsieh, H. F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qual Health Res, 15(9), 1277–1288.
If you want to reference the use of Cody in your academic work, you can cite it as follows:
APA 7
tebbl UG. (2025). Cody [Computer Software]. In QDA - Software (0.8). https://www.tebbl.com/cody, Retrieved YYYY.MM.DD (Tag der Analyse)
AMA10
tebbl UG. Cody. QDA - Software. Juni 2025. https://www.tebbl.com/cody. Retrieved YYYY.MM.DD (Tag der Analyse)
Mayring, P. (2000). Qualitative Inhaltsanalyse: Grundlagen und Techniken. Weinheim: Beltz.
Mayring, P. (2014). Qualitative content analysis: theoretical foundation, basic procedures and software solution. Klagenfurt.
Strauss, A., & Corbin, J. (1990). Basics of Qualitative Research: Grounded Theory Procedures and Techniques. Sage.
Hsieh, H. F., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qual Health Res, 15(9), 1277–1288.
If you want to reference the use of Cody in your academic work, you can cite it as follows:
APA 7
tebbl UG. (2025). Cody [Computer Software]. In QDA - Software (0.8). https://www.tebbl.com/cody, Retrieved YYYY.MM.DD (Tag der Analyse)
AMA10
tebbl UG. Cody. QDA - Software. Juni 2025. https://www.tebbl.com/cody. Retrieved YYYY.MM.DD (Tag der Analyse)
How to Prepare Your Data for Cody
What Makes Good Input Data?
The quality of Cody’s results depends largely on how well you prepare your data.
Because: Only those who input clean, structured data will receive usable results in return.
To help you get the most out of your analysis, here’s a compact guide on what constitutes good input data – and what you should pay special attention to.
Because: Only those who input clean, structured data will receive usable results in return.
To help you get the most out of your analysis, here’s a compact guide on what constitutes good input data – and what you should pay special attention to.
🗣 Which Interview Transcripts Work Best?
✅ Clear, coherent sentence structure
Your transcripts should be grammatically correct. Avoid fragmented or error-ridden sentences – they significantly hinder the coding process.
✅ Separated speaker contributions
Speaker turns should be clearly distinguishable. Overlaps or constant interruptions reduce semantic accuracy.
✅ Speaker labeling
Clearly indicate who is speaking – for example:
Moderator: "How do you see that?"
Participant: "I think that" …It’s especially helpful if you label the moderator consistently throughout. This will help cody differentiate between statement made by you and the interesting juicy responses for the analysis.
✅ Digital, readable text (no scans!)
Cody can only process real, machine-readable text – no image files, no scanned PDFs, and no handwritten notes.
❌ Photos,
❌ OCR text with errors,
❌ jumbled copy-pasted fragments → lead to unusable analyses.
✅ Thematically consistent interviews
Use semi-structured interviews where all participants are asked similar questions – ideally aligned with your research question.Avoid improvised, unstructured conversations with constantly changing topics.
✅ Minimal moderator intervention
Let your interviewees speak freely. Stick to your interview guide and only ask follow-up questions when necessary. This encourages the depth needed for meaningful category formation.
Your transcripts should be grammatically correct. Avoid fragmented or error-ridden sentences – they significantly hinder the coding process.
✅ Separated speaker contributions
Speaker turns should be clearly distinguishable. Overlaps or constant interruptions reduce semantic accuracy.
✅ Speaker labeling
Clearly indicate who is speaking – for example:
Moderator: "How do you see that?"
Participant: "I think that" …It’s especially helpful if you label the moderator consistently throughout. This will help cody differentiate between statement made by you and the interesting juicy responses for the analysis.
✅ Digital, readable text (no scans!)
Cody can only process real, machine-readable text – no image files, no scanned PDFs, and no handwritten notes.
❌ Photos,
❌ OCR text with errors,
❌ jumbled copy-pasted fragments → lead to unusable analyses.
✅ Thematically consistent interviews
Use semi-structured interviews where all participants are asked similar questions – ideally aligned with your research question.Avoid improvised, unstructured conversations with constantly changing topics.
✅ Minimal moderator intervention
Let your interviewees speak freely. Stick to your interview guide and only ask follow-up questions when necessary. This encourages the depth needed for meaningful category formation.
🎯 The Research Question – How Should I Formulate It for Cody?
The research question is the core of your analysis.
Cody relies entirely on it – which means it must be:
Clearly formulated
Explicitly based on your interviews ✅✅
Not too general -> ❌ "Why are people lazy?"
Not reliant on other cross-method from your thesis (e.g., mixed parts of the question are answered by other methods and not the interviews you upload to cody)
Avoid simply copying your thesis question if it’s too broad or theoretical.
✅ Good Example:
“According to my interview participants, what everyday factors prevent them from switching to a vegan or reduced-meat diet?”
🟢 This question is concrete, interview-based, and focused on causes that can be identified in the data.
❌ Bad Example:
“Why don’t people become vegan?”
🔴 Too vague, too general, and not clearly grounded in interview-based data.
Tip: Think of your research question as the title of a pie chart. Cody will later show you the thematic portions – for example:
Social pressure
Habit
Lack of availability
If your question isn’t precise enough or your interviews stray too far from the topic, you won’t get clear, usable results.
Cody relies entirely on it – which means it must be:
Clearly formulated
Explicitly based on your interviews ✅✅
Not too general -> ❌ "Why are people lazy?"
Not reliant on other cross-method from your thesis (e.g., mixed parts of the question are answered by other methods and not the interviews you upload to cody)
Avoid simply copying your thesis question if it’s too broad or theoretical.
✅ Good Example:
“According to my interview participants, what everyday factors prevent them from switching to a vegan or reduced-meat diet?”
🟢 This question is concrete, interview-based, and focused on causes that can be identified in the data.
❌ Bad Example:
“Why don’t people become vegan?”
🔴 Too vague, too general, and not clearly grounded in interview-based data.
Tip: Think of your research question as the title of a pie chart. Cody will later show you the thematic portions – for example:
Social pressure
Habit
Lack of availability
If your question isn’t precise enough or your interviews stray too far from the topic, you won’t get clear, usable results.
Fazit: Gute Daten = Gute Ergebnisse
🧠 Cody is a smart tool – but not a magician.
👉 Only with well-prepared, thematically consistent data will you receive truly meaningful and useful results.💡 Take your time with preparation:
• Cleanly prepare your interview transcripts
• Formulate your research question precisely
• Upload only relevant, clearly readable text
👉 Only with well-prepared, thematically consistent data will you receive truly meaningful and useful results.💡 Take your time with preparation:
• Cleanly prepare your interview transcripts
• Formulate your research question precisely
• Upload only relevant, clearly readable text
Sonstige FAQ
Why can analysis results differ even with the same input?
Cody uses non-deterministic, AI-driven methods for qualitative content analysis. This means that even with identical input data, slight variations in results may occur – much like with human coders.
We are continuously working to make the results more valid and consistent, but some degree of variation is inherent to exploratory methods.
We are continuously working to make the results more valid and consistent, but some degree of variation is inherent to exploratory methods.
Can I cancel an analysis once it has started?
No, currently an analysis cannot be canceled once it has started. Please carefully check the confirmation window before starting to ensure that all data is correct and you truly wish to proceed with the analysis.
What are credits – and how do they work?
Credits are used to pay for your analyses.
1 credit = 1 document with up to 60,000 characters (including spaces).
If your document is longer, an additional credit is automatically charged for every started block of 60,000 characters.
The total number of required credits is clearly displayed before you start the analysis.👉 You purchase credits in advance and use them flexibly.
1 credit = 1 document with up to 60,000 characters (including spaces).
If your document is longer, an additional credit is automatically charged for every started block of 60,000 characters.
The total number of required credits is clearly displayed before you start the analysis.👉 You purchase credits in advance and use them flexibly.
Can I get a refund for my credits?
Generally: No.
Credits cannot be refunded or converted back into money – even if you’re dissatisfied with the results.
Exception:
If a technical error on our side prevents results from being generated despite credits being deducted, you will of course receive your credit back.
Please contact us at support@tebbl.com in such cases.
Credits cannot be refunded or converted back into money – even if you’re dissatisfied with the results.
Exception:
If a technical error on our side prevents results from being generated despite credits being deducted, you will of course receive your credit back.
Please contact us at support@tebbl.com in such cases.
In which language will my results be displayed?
The language of the results is based on the language of your research question.
If you enter a question in German → results will be in German
If you enter a question in English → results will be in English
💡 Note: If your research question contains words from multiple languages, this may cause confusion in the analysis. So make sure to formulate it clearly and in a single language.
If you enter a question in German → results will be in German
If you enter a question in English → results will be in English
💡 Note: If your research question contains words from multiple languages, this may cause confusion in the analysis. So make sure to formulate it clearly and in a single language.
Which file formats can I upload?
.PDF
.DOC/DOCS
.TXT
It is important not to use scanned documents or files in which the text is only available as an image.The file must contain real, embedded text that Cody can read and process.
.DOC/DOCS
.TXT
It is important not to use scanned documents or files in which the text is only available as an image.The file must contain real, embedded text that Cody can read and process.
How long can my research question be?
The maximum length is 300 characters.
This is sufficient for a precise, clearly formulated question – make sure to use clear language without unnecessary wording.
This is sufficient for a precise, clearly formulated question – make sure to use clear language without unnecessary wording.
Can Cody analyze images?
No, Cody is a purely text-based QDA tool and does not process images or other complex file contents.
Where can I find a receipt for my purchased credits?
You can find your invoices and receipts by clicking on your user icon in the top right corner, selecting Settings, and then navigating to Invoices & Billing.
There, you can view and download them at any time. We recommend saving them regularly for your records.
There, you can view and download them at any time. We recommend saving them regularly for your records.
Can I delete my Cody account?
Yes, you can delete your account at any time via the user menu.
Important:
Once deleted, your account will be completely and irreversibly removed within 30 days – including all stored data, analyses, and any remaining credits.
Important:
Once deleted, your account will be completely and irreversibly removed within 30 days – including all stored data, analyses, and any remaining credits.
Can my interviews be in different languages?
Yes – this is generally possible. Cody always conducts the analysis in the language of the research question.
This means that even if your interviews are in German, English, or a mix, the coding will follow the selected analysis language.
💡 For best results, ensure consistency in language across your data whenever possible.
This means that even if your interviews are in German, English, or a mix, the coding will follow the selected analysis language.
💡 For best results, ensure consistency in language across your data whenever possible.
© cody is a product run by the Tebbl company. Tebbl's terms apply.
