I find educating users is getting more challenging lately. Even if you build the perfect interoperability platform with every feature a user may need, if they don’t know what’s possible, where to find it and how to use it. It depends on many factors like culture, habits, and even age. Some users prefer reading and others expect everything to be super intuitive, some search a KB and others simply ping a colleague, or ask ChatGPT. I wonder what have you seen to work best from (or a combo):
- Having an extensive (written) documentation - explain all capabilities, how-tos and use cases, I feel fewer people are reading these days and focus time has dropped, so there needs to be a balance in how much content you provide - dozens of pages are probably too much. But, if you hook it with an AI assistant e.g. have it on SharePoint for Copilot or on Confluence for Rovo to consult with, you can have an interactive chatbot helping users.
- Training sessions (video/slides) - demos can be easier for Desktop/Browser UX, since users see what’s happening instead of having to figure out where, e.g. a button is. There are two issues IMO - how many people will watch these videos and maintainability. If you have to re-record videos all the time, that’s a nightmare. Some tools help make a demo while just clicking through an app (I’ve tried only browser-based ones), but I don’t feel they save much time.
- Intuitive UX - that’s probably the best solution, but UX is an art, and when you have very complex capabilities or technical limitations isn’t always enough to rely on.
- In app tooltips and hints - a very elegant solution, but adds some complexity. It’s good to have the option to track how many of the tips users see before clicking dismiss
- Chat UI (or voice) - this is what everyone is trying to implement now. You tell the app what you want to do, and it figures it out. The docs from point 1, can be a good foundation here. Problems - hallucinations, having to provide relevant context, typing and coming up with an effective prompt can sometimes take more time than clicking around, cost, security concerns (if using external providers), hard to achieve with complex scenarios.
Perhaps I’m missing something…