Throw away prototypes
Start from High Level
I am near finishing Udacity course on Rapid Prototyping. It was just as great, casual, and useful as I have come to expect from other courses on Udacity.
The whole purpose of prototyping is to fail fast. It sucks when your site is even just in beta version and you get email or some feedback, somoene saying “it’d be so great if i could just do this”, and then you realize that it’s a very logical complementary feature to have. But now building it into the app requires you change code on at least 3 levels (db, backend, css/js)…reevaluate design aesthetics and reconsider functionality impact of this new feature as a whole. So, even though, this feature is very useful and should be no brainer to be present in the app, and it’s not technically hard to implement…beside all of this it is likely to be put off for todo list of next version x.
Bug discovered in development/production costs many times more to fix than in prototype stage
Another reason, for rapid prototyping is that you don’t need to spend long time shooting in dark and trying to find meaning in life to come up with what user would want, or if this is good. When you can just ask people. You will get richer insight and get it fast, without exhausting your brain stamina. In Rapid prototyping Quantity made of iteration, may they be dumb, is favored over Quality. It’s simply easier to see what’s working and working when you see it than to sit and try to visualize few scenarios.
Stages of Prototype
Low Fidelity -> Medium Fidelity -> High Fidelity -> Beta -> V1.
In a sense goal behind prototyping is testing User experience, and is a long continuous process. A/B testing is one good example of it that is used in apps in production However, in early stages of prototype the user feedback you look for is more geared toward critical aspects of app functionality and hardly about font or color, and in later stages feedback is fine tuned to things like animation, font and so on.
Low Fidelity: Test critical points of interaction, layout, message, less about aesthetics, use wireframes if needed instead full photoshop mockups.
Medium Fidelity: Testing userflow, interaction between pages and ui elements, navigation, design should help in navigation, but dont need to be eye candy and balanced. Interactive mockups are required.
Circles of Feedback:
There are about four circles of feedback that you can get feedback from, each is good for certain type of feedback.
- Friends & coworkers
- Experts, UX designers, Professor, etc
- Customers (who are paying)
- Customers’s customers (who use the product)
Not all people are good for advice, though they can give good advice. Some people are more likely to give good advice. People who are too close like your dad is likely to say yes *put anything you made* looks awesome. Similarly, people who have worked on the idea/app will be experts in what the app and will not notice unintuitive parts or want parts that they have grown comfortable to not having.
This is further extended to people who are too technical, in Low fidelity stage, you want to get feedback on basic structure of app and see if elements are making sense, Technical people are more likely to jump in to specific details such as design. Which is not what you are testing right now, even if advice is right. But you haven’t decided on the design yet, and other factors may influence it, and those factors may have not even come up and certainly the user is not aware of them yet.
Remember do not let suggestions lead the design blindly. It’s important to get questions answered on what you are testing for.
Interacting with the User
The video of user being recorded while getting is better than audio, Audio is better than notes, notes is better than just conversation.
Video of user getting excited while using your app is great for investors and presentation purposes as well.
You want users to think aloud, one way to get them started is tell them to practice by telling you how many windows are there in their house?
Now tell ask them how did they come up with this number, can they lead you through the thought process. Now do the same thing with the app testing.
It’s very important to let user know that you are not testing them, they can’t make a wrong moved, being confused is completely ok. What you are testing is your app. If it makes you confuse, it’s great, since this is exactly what you wanted to know that what parts are confusing.
Guide user by asking to do certain task from certain screen.
Help out user if they get stuck on some parts and move them to testing other parts, other parts need testing as well.
Ask questions..qualitative for low fidelity
- about completing a task
- purpose of ui elements
- suggestions on critical things they had confusion on
- give their thought on other ideas you have on ui elements
- other interactions with apps
- can user easily learn this app?
Once we have feedback for low fidelity stage prototype we can move onto next stage if answers to both questions is yes:
- Did we get all questions answered?
- Did feedback say we are moving into right direction?
Many apps go through 5 or more iteration process.
Feedback by 5 users cover 75% of problems.