Feedback from your customers and website visitors is excellent way of understanding what you’re doing right and perhaps more importantly what you’re not doing well.
We use it when auditing clients site alongside analytical data to get a complete picture of how people move around a site
Often the most useful information can be gleaned from open ended questions such as ‘What is the purpose of your visit to our website today?’
If you ask someone to describe things in their own words, you’ll get the real reason why they came to the site, plus an insight into the language that they use.
Contrast that to asking them to pick one choice out of a list (that you wrote in your own words) and you can see why it’s so useful.
Of course, it’s not just simple as putting a survey up there and assuming that everyone will fill it out.
The Case Study.
The BBC is testing a new version of their home page and invites you to try it out and leave your feedback.
By using the word beta, you’re more likely to get more IT literate people who understand what a beta test is and will accept that there may be a few issues. They’re also like to give their opinion.
Notice that they are actively asking for feedback and
After looking around the site, we wanted to leave a comment about setting the location. We wanted to suggest that perhaps if entered a Swansea postcode, the site should detect this and offer Swansea as an option, in the same way that Google does.
Considering they want feedback they don’t make it easy to do.
Rather than wait a couple of minutes and then popup a window, the visitor needs to
1 .Click on link virtually identical to the previous one
2. Go to another page and locate a panel that says leave us some feedback
The the fun really starts
Bearing in mind we wanted to leave some feedback, we then get asked around 15 questions before we leave the feedback we wanted.
This includes a set of mandatory questions on
- Job Type
Followed by more mandatory questions on what I thought about it the site
So Why Is This So Bad?
The first rule of surveys (as with forms) is that the longer you make the greater the chance you have of people not completing them.
This is especially true if you decide to make every question mandatory and don’t provide any reason for it.
While it’s useful to have demographic information, is it really vital to know I’m a white british male from wales?
Or would it have been more useful to ask me for my feedback right at the beginning
How Can You Avoid Making The Same Mistakes
- Don’t do what the BBC has done and simply take an existing survey and republish it
- Understand your audience and their motives for filling out the survey
- Only ask directly relevant questions
- Test the survey on a small sample
- We would on the principle that you never get it right first time
- So we make our mistakes on the small sample and then refine