Weeknote 5: hula hooping and machine learning
This week is going down as one of the most stressful in a really long time. It’s the school holidays and my childcare plans have fallen through. Memories of homeschooling during lockdown flooded back. My heart has been in my throat, my stomach has been tied in knots and trying to keep calm has been a real struggle. I’ve been context switching and unable to dedicate focused attention to anything. This makes me mad. So mad that I turned to a bit of retail therapy: I bought a Powerhoop. What’s a Powerhoop, you ask? It’s the latest exercise fad that I’m jumping on board with. It’s hula hooping for grownups! What is not to love about that?! When I close this laptop tonight, I’ll be in my garden swinging my hips to some funky funky music. Goodbye, stress!
Exercise is one of the best things we can do to combat stress. And if it can make us laugh too – what a WIN!
One of my tasks this week has been to analyse a large piece of client research. A cross-section of clients has been interviewed with regard to their experience of Aptitude products & services. I have a portion of the transcripts to analyse and extract insight from. One of the hardest things about analysis is putting aside our own bias. It would be so easy for me to look at the transcripts and pull out all references to user experience and wax lyrical that this should be our main focus. To ensure I am approaching the research unbiased I’m coding it; this involves pulling out keywords from each text snippet. Sounds like a long job, right? It is. But it brings value by enabling us to uncover the real key themes, not just the themes I want to hear.
Another part of the coding process has involved sentiment analysis. Sentiment analysis is the use of natural language processing to systematically identify affective states and subjective information. It helps me see which statements were objective and which were subjective, which were positive, and which were negative. It gives me another way to ‘cut the data’. The more ways we can cut it, the richer the insights can be. I LOVE DATA ANALYSIS!
I went one step further with this piece of work. As I was looking into tools to help with sentiment analysis, I looked at other machine learning options on the market. I started thinking about all the Customer Advisory Board transcripts we have, about all the support tickets we have, and how we could tell a story if only we could crawl through that data. Historically we have not done user research at Aptitude, and none of our applications has analytics for quantitative behavioural tracking. I’ve found myself relying heavily on my own experience, and on internal insight - not on actual data. There is a risk associated with this. The risk of building a patchwork of opinions. The risk of misunderstanding client needs. The risk of only seeing part of a problem.
It would take years for me to code up all our support tickets. That’s obviously a very poor use of anybody’s time. However, machine learning would enable me to leverage what we already have to gain more insight into our users. The more we can learn about our users, the better products we design. And the data is sat there begging us to hear its story!
I have a user research head in the budget for 2022, but until they arrive, I’m very curious about what value we might get from machine learning. Even when I have said researcher on board, the ‘machine’ could automate the time-consuming coding, freeing them up to undertake more value-add research tasks.
The highlight of my week was looking through applications for the Graduate UI Designer role. The quality of the candidates is fantastic and I’m really excited to start the first interviews next week. Over the coming month, I’ll be preparing an onboarding plan. This is where we face one of the challenges of working remotely; how to thoughtfully and personally welcome somebody who is at the beginning of their career, with potentially no commercial experience, and who needs that little extra support. I will share more about how I approach this in the coming weeknotes.
The biggest frustration of my week was not completing the UX maturity analysis that I set myself last week. It’s unlikely I’ll have time next week due to attending 4 days of training, so I’ll forgive myself and put it aside until the following week.
My shout out this week is MASSIVE - Marcin Szulzyki, our Senior UX Designer, led the very first design sprint at Aptitude. The response so far has been phenomenally positive. An extremely well planned and executed event. The wealth of insight we have gained is incredible. You were on fire!
My mantra this week is: You can’t always do it all. Don’t sweat it. Leave the sweating for the hula hooping, not the desk.