UX Researchers: Spice up your work with a chatbot project
Jul 20, 2017 · 2 min read

Have you hit a UX research rut? You know, where your research is entirely focused on validating if designs “make sense”.
If you want to spice up your work, consider working a chatbot. Here’s why chatbot projects are awesome for researchers:
- Chatbot testing requires a true hypotheses. The big question when designing chatbots is rarely “is it intuitive?”. Chatbots invite deeper questions like — “what’s an appropriate tone?”, “is it trustworthy?” These are meaty and fun to tackle!
- You need qual + quant data to design a great chatbot. As a researcher, it’s easy to stick to one type of data: whether is be small-scale “deep” research or large quant research. For chatbot, you absolutely need both. Unless you’re making something simple, you’ll need a large corpus of data to pick your use cases and plan your dialog. You’re also need observational research to get the details right. Fun!
- Chatbots live and die by primary research. With chatbot design, you don’t get any free points for a good layout or a subtly pleasing animation. A good portion of the success of your chatbot will be dependent on your primary research. You’re the hero! Woot!
- Chatbots need to be released before they’re perfect. You can’t fully tune a chatbot without it being used in the wild a bit. This gives you an opportunity to think about longer-term monitoring plans.
On top of this, chatbots provide a great opportunity to think beyond web patterns and consider broad principles like memory limitations, and attention considerations. This gives you strong transferable skills for new and novel platforms.
Maybe a chatbot is the challenge you’ve been craving.
