Every user survey I've conducted has suggested significant changes to proposed functionality –
and this is so much cheaper than developing first, user-testing after.
Please tick one of (Y)es / (M)aybe / (N)o: |
Y |
M |
N |
1. Is it true that "normal people", your market, like and do roughly the same kind of things as we do? |
|
||||
Any comments? Which percentile are you, in terms of intelligence? Or income? Exposure to technology? |
2. A user survey, asking 20-40 representative people about your proposed features, is a good idea, early on? |
|
||||
Any comments? It's the single simplest, most cost-efficient way to improve your final product. |
Please tick one of (Y)es / (M)aybe / (N)o: |
Y |
M |
N |
3. If you perform a user survey, are you asking your users to design your product? |
|
||||
Any comments? No, we're not asking "What should we do", but "What do you think of these ideas". And it's not absolute — you do still interpret results. |
4. Lots of firms provide online multiple choice surveys, so it's simple and quick? |
|
||||
Any comments? Allow open comments for every question — you'll get so much great feedback. And be careful its online nature doesn't self-select those who are more into technology. |
Please tick one of (Y)es / (M)aybe / (N)o: |
Y |
M |
N |
5. A survey can provide you with enormous amounts of potentially interesting information? |
|
||||
Any comments? Actually, keep surveys fairly short, or people tire and provide less good feedback. Don't ever ask a question if its results can't actually change what you do. |
6. If they can save wasted development, should I get a survey out there today? |
|
||||
Any comments? If there's any way to misunderstand a question, your results will be harmed. So have your best wordsmith re-word it until you can see no possible misinterpretations, then test your questions out, two or three times, before you launch. |