In user experience there is often a heavy emphasis on doing research. Often this involves doing user research studies. Here you might extensively interview. Or you might watch focus groups directly use your product or app.
Although this is a fantastic approach and one I advocate myself, what are you to do if you can’t afford much research yet?
I have talked before about making educated guesses in UX and this is no exception.
Even with extensive user experience research, you are still making an educated guess.
Here are a few approaches that I have used. Many of these were used on the award-winning SnowflakeStories.com. Specifically to improve our user experience without a heavy investment in research up front.
- Apply common patterns and best practices. GoodUI and Pattern Tap are a good source for this.
- Place yourself in the perspective of your ideal customer. Be open minded. Pretend and make believe.
- Use products and devices that your end users will use. Don’t use a 5k iMac (my favorite) use a cheap laptop or cell phone that is an average of what your audience uses.
- Keep an eye on actual analytics (I suggest Google Analytics). What actual browsers are people using. Which pages are entry points?
- Review user sessions of key pages by hand with a tool like MouseFlow. Look for things that hang people up.
- Guerrilla user testing. Just observe don’t sell or describe your product.
- Review significantly competing sites on lower end devices. Ask yourself, how can this be improved for my ideal customer?
- Don’t loose mind of conversion. Ask yourself for every page, what is the one action I want my customer to take on this page. Make it dead simple and obvious.
These are just a few ideas to get you started. But remember, the advantage of taking this UX Lite workflow is a faster time to market. As soon as you can review actual user feedback and actual users (see #4 and #5) you should prepare a list of UX issues to resolve.
This is one advantage we had with SnowflakeStories.com. Much like a design system itself, we could quickly resolve and iterate our interface based on actual user feedback.