So, you’ve decided to learn more about your users with some quick research. That’s a great move. User insights guide good decision-making and ensure we focus on solving the most important problems.

There are plenty of tips on how to run effective interviews, surveys and user tests, but that’s not where research ends. It takes time and energy to synthesise our findings, and for teams without much experience, this task can be overwhelming.

While AI struggles to produce original ideas, it is very effective at summarising masses of data. Here are three ways teams can use AI to speed up user research.

1. To transcribe interviews with accuracy

If a single team member is running your interviews, we suggest recording the calls and transcribing them afterwards, as note-taking may lead to inaccuracies and missed quotes. This allows the interviewer to stay fully engaged and attentive.

Transcription tools have been around for a while, but struggled with accents, dialect and slang, often requiring extensive editing. New AI tools have greatly reduced this effort, and we have a few favourites.

At Neverbland, we use Supernormal, an AI assistant that joins online calls and produces editable transcripts. It even summarises the call, making it easy to revisit key discussion points. Alternatively, if you miss the chance to transcribe your call in real-time, consider using Notably or Dovetail. These platforms allow you to upload audio or video files of past interviews and generate transcripts within minutes.

Most AI transcription tools offer their basic service for free, making them a no-brainer for smaller teams. But check their limitations, as some restrict the number of free hours or monthly usage.

2. To summarise your users’ behaviours and preferences.

After running interviews, you’ll need to summarise the key findings. Traditionally, this involves extracting key quotes and sorting them into an affinity map. However, this is a lengthy process and it can be tough to justify the time it requires.

Here’s how we use ChatGPT for rapid synthesis of user interviews:

• Upload your transcript files to ChatGPT. The free version provides this feature.
  
  
• Craft a thorough prompt. This is what your outcome hinges on, so be specific. Make your research goals clear, outline the insights you want, and ask for direct supporting quotes.
  
  For example, ‘Summarise these user interview transcripts to help me understand user attitudes and behaviours around [insert topic]. Please provide detailed summaries of user values, pain points, and any feedback and opportunities arising from their current experience. Provide quotes to support each point’.
  
  
• Keep asking questions. Don’t stop there - if an insight needs more exploration, keep probing for more information. Use phrases like ‘provide more detail on ‘give another example of…’ to achieve depth in your synthesis.

3. To pull statistics from your qualitative data

When summarising survey responses, we like to use AI to extract statistics. Drop your survey files into ChatGPT and ask it to rank answers by frequency, identify correlations between questions, or segment the data by user group for a comparative analysis. This helps identify patterns and nuances in your audience’s feedback.

The same approach applies to interview transcripts. Ask data-oriented questions like, ‘what were the most common barriers for users wanting to continue their subscription and what % of participants cited each one?’ This quantifies the prevalence of themes, helping you prioritise issues based on their impact and urgency.




Using AI as a research sidekick makes synthesising data much easier. But we must acknowledge that AI is not enough on its own. We’ve found that AI tools can misinterpret tone, mistake sarcasm for praise, or miss subtleties. A real, empathetic person is still needed to fact-check and validate results.

Looking ahead, AI holds exciting potential for research. Yet, while the AI world makes big claims, it’s currently best suited to tasks involving information-gathering and summarising. For now, our emphasis isn’t on using AI at any opportunity, but on mastering how to leverage its strengths.