Updates to Quirkos Web and Transcribe - September 2025
It's been a busy summer of updates to the browser based version of Quirkos, making it better for all your qualitative analysis needs!
The main focus has been on performance improvements, in particular in projects with large numbers of codes/Quirks. Previously, when you had more than 100 Quirks on the canvas, loading the overview or adding highlights would lag, but this has been vastly improved. There are still other areas to optimise in large projects, especially around deleting codes, which we hope to address later in the year.
However, we still stand behind our design philosophy: Quirkos is a tool for small qualitative projects, and having hundreds of codes isn't a good thing for most types of analysis. By the time you have more than a 100 codes or so, it's time to review and merge or delete ones you don't need: you'll quickly get to a stage where you are creating duplicate codes (when you can't remember if you created one for Anxiety, so created extra ones for Anxious). Meaningful comparisons between codes and sources are difficult when you have hundreds of codes with just one or two highlights in them: keep it focused!
Quirkos is also not a tool for very large numbers of sources: they can be long for sure, but again we are aimed at the typical small qualitative project with a couple dozen interviews or focus groups, not hundreds or thousands of small sources. These data sets lend themselves to more quantitative approaches (eg word frequency analysis) and aren't the focus for Quirkos.
Talking about sources though, we have also just rolled out a major update to the amazingly popular automated transcription service! This is an optional add-on for Quirkos that lets you transcribe up to 50 hours a month for just US$12 a month. The most requested feature we've had is to add speaker identification, especially for focus groups, so that you'll automatically get a 'Speaker_1', 'Speaker_2' tag for each participant. We've recently added this, as well as upgraded the server systems that in our testing roughly halves the time to do transcription! So you should now see a 1 hour interview come back in around 5 minutes! The one 'feature' we've removed is the automatic language identification: this was often not very good at guessing, so you now need to specify what language your audio file is in.
Our transcription service is accurate, as well as much faster and cheaper than any of our competitors. However, the most important thing is that it is fully end-to-end encrypted: audio is encrypted on your own computer before it's uploaded, and transcribed on our own server that isn't shared and doesn't send data back to any third parties. The transcript is encrypted to sent back to you, and this ensures maximum data protection and security. As an added bonus, like our other servers it runs entirely on green, zero carbon energy, and we've highly optimised it for efficiency.
We are so proud of this system, we give everyone a 100 minute free trial to see how good it is, and are increasingly seeing people signing up for Quirkos just to use the transcription service!
Additionally, we've also added Optical Character Recognition (OCR) support to our PDF import, so if you have an older or scanned document, you can extract the text from it. Obviously, the accuracy of this depends on the quality of the image, but it really helps with older journal articles, and is another free addition to Quirkos Cloud. We have also improved Word import, and adding ODT import.
There's a lot on the radar for further improvements in the rest of the year! We will also provide a free update the offline version of Quirkos with a lot of these performance and import improvements later in the year. However, we are still not planning to add any 'AI' coding or analysis tools to Quirkos. We still believe these tools are too inaccurate, run by amoral companies, exploitative to people and planet, and take away from the core aims of qualitative research: finding new and complex understandings to rich lived experience and social issues. Plenty of other tools have tacked on 'AI' features to keep up with the hype, and you are welcome to them.
As ever, drop us an email (support @ quirkos.com) if you have any suggestions for other new features or improvements you'd like to see in Quirkos!