Achieving Synchronicity - Why We Still Process User Research Insights by Hand

fieldnotes

7 Mar

In this new age of Generative AI, there’s a significant opportunity to speed up the process of research synthesis by using automated tools to group and theme insights.

If you haven’t encountered these tools, their operation is simple to explain. Sticky notes, transcripts and even raw video can be fed into tools that automatically group related quotes, observations and statistics into themes and over-arching insights. These tools work by matching phrases, implied sentiments and behavioural characteristics.

We can debate whether they produce accurate insights, but they clearly work quickly. As such, they seem to offer a tremendous performance improvement to the sometimes paced process of research synthesis.

So why are we at Cognitive Ink still avoiding significant use of auto-insight tools?

It might seem like a contentious choice, but there’s a logic behind our choice.

And it all comes down to getting in sync with what people share.

Let me explain.

We’re not making the argument that these new automated synthesis tools can’t pick out themes. Even before this brave new era where Generative AI / Large Language Models have been deployed in synthesising user research, there have always been various software based algorithms that purport to pull out themes from masses of data.

Instead, we argue that, as researchers and design mapmakers, we’ll be psychologically worse off for outsourcing and automating this crucial moment of discovery. If we automate synthesis, we’re concerned we’ll miss the change in our own emotional and cognitive states that comes from getting in sync with what someone has shared.

Think about this for a moment.

If you’ve never experienced it, there’s a moment in user research when you’ve sat with, soaked in, and lived with someone’s stories. Sometimes for weeks. It’s like climbing to the top of a hill of feedback, and then slipping down the other side, with an ever-increasing speed of insights. You become entrenched in their world-view. A moment of emotional breakthrough. Not so enmeshed that you lose perspective, but enough that you stop living solely in your own world.

A key point; this psychological transformation is hard. It’s hard for us to step outside our world-view, to understand situations from someone else’s point of view.

Despite these risks, this moment is one of the key reasons to conduct human centred user research; to understand what people experience, even if they may not put words to it.

It’s this psychological shift, in ourselves and the clients we support, that we believe are essential to carrying the right insights and framing into the act of design.

It’s how we speak with confidence and power about what people need. Because we’ve reached some level of synchronicity with the people that use the things we make.

But don’t mistake this synchronicity for naïve question-asking, note-taking and reporting. We don’t just adopt a point of view without some critical consideration of both our and their biases, contexts, purposes of sharing and other power differentials.

That said, so many products, services and experiences (and the technology that powers them) have been designed from very narrow points of view. These narrow points of view end up missing, misunderstanding or distorting their needs, contexts and constraints of the people that will use, or be affected by the things we make.

Imagine then, doing all the interviews, observations, surveys or shadowing studies. Then feeding all of that insight into a machine, pressing a button and then watching insight after insight spit out. Some of them might resonate more than others, based on in situ theories of behaviour we’re developing.

But we’ll have fast-tracked getting in sync and therefore won’t have the fine-grained ability to really discriminate what matters.

As the anthropologist Henry Glassie says, “I stand with people.”

I can’t imagine it’s possible to ‘stand with people’ while sitting, staring at a swirling screen, as a machine churns out my synthesis.

As the tools change, we’ll constantly re-evauate our use. But for now, we’re still understanding our insights at the workface, quote-by-quote.