How to run a website satisfaction survey

What do you want to find out?

Think carefully about what you want to know. You may be looking for data on:

• overall satisfaction
• views on a particular section
• the background of your users and why they come to your site
• views on your new design.

Focus on one area of research and don't try and do too much.

If you are relaunching your website or a section of it, run the same survey before and after launch to assess views.

If you are using a survey to assess use of your website, then it makes sense to run the survey online. However if you are gathering data about your supporters or views about the information they want, then you may want to run the survey using a mix of on and offline (ie web, email, paper, phone calls etc).

Keep it simple

Get the questions right so people feel inspired to join in.

Questions should be short and specific. Ask straightforward questions with simple options. COI have produced a useful guide to measuring website quality which has some sample questions. Do include some open questions where appropriate so people can share ideas or views as well as ratings.

Think carefully about wording so there is no confusion about what you are asking and you can present analysis clearly. Questions on their own may not give you useful data so build in other questions which give you some context. For example on it's own, asking about overall satisfaction doesn't really give you useful data. This subjective rating could be affected by lots of things such as familiarity with the site, IT-literacy, the colours you use etc. In this case, it may be useful to get some basic information about whether they have visited the site before (and how frequently) and whether they are a supporter (have given money / volunteered / campaigned etc). Then analyse the questions together (eg 90% of supporters rated the website as excellent, 65% of those visiting the site for the first time thought it was excellent).

Questions about tasks are less subjective. For example, ask whether they found what they were looking for or whether they would recommend the site to others.

Do you really need name, email address, postcode etc? Requiring contact details may put people off responding.

Keep it short

Short surveys are good (1-3 questions) but may not give you enough information. Try to limit the number of questions to no more than 10.

If you have to run a long survey, clearly state how many questions or pages there are, indicate how long it will take to complete and once they've started show how far they are through the process.

Promote the survey

Whether the survey is on your site, on an external site (such as Survey Monkey) or offline you'll need to promote it in order to get enough people to fill it in to make the data useful. How you promote it will depend on what your survey is about.

For example if you are looking for data about who is visiting your site during a particular week / month, you may not want to promote the survey to your supporters as they may rush to visit the site and skew your data. In this case, simply add a prominent link on your homepage (or the section you are reviewing) and hope that people who are visiting your site will fill it in. However if you want views on a new design (for example) then use all your channels to invite people to take your survey.

Share your results

When your survey is over and you've analysed the results, share them. Either publish on your site and / or email respondents (if you've gathered their addresses) or your supporters with the key findings and tell them what action (if any) you are planning to take as a result.

Run the survey again

Use the same survey again so you can compare results over a period of time.