Some implications of "digital" for scholarly writing and publishing

How not to create a website user survey

Reading Time: 4 minutes

I can’t resist surveys. I always sign up for website user experience surveys, because I feel that since I am in the business of digital information, I ought to be prepared to participate to help UX people design better sites. After all, we are told endlessly that listening to the customer is the key to a good interactive design. Sadly, despite my good intentions, I unfortunately almost never manage to complete them, because, well, the user experience is so poor – for the questionnaire, quite apart from the website! I usually abandon the survey without completing it, which I fully accept is a waste of my time and that of the compilers.

This time I have tried to make the exercise a little more valuable. Instead of condemning the survey writers to the lowest pit of hell, I’m providing what I hope is some constructive feedback. I can’t do this on the site itself because survey designers almost never allow you to provide this kind of feedback (“your survey is rubbish, because…”). This is because they have no idea how dreadful their survey is.

Questionnaires should follow the same usability rules as a website. The key principle is that the user’s time is limited and so the design of the survey, like the design of the website should maximise effectiveness and minimise redundancy.

The specific example was a questionnaire from Tesco, who used a third party to create a questionnaire about Tesco online shopping. But to be honest the standard of such surveys is very low, so I could have shown several other examples. Perhaps on balance this one is worse than most.

You are asked to complete a questionnaire about the website after using it to buy groceries. Your experience is fresh, so you feel you can contribute at that moment.

I’ve summarised my views as a set of recommendations.

Don’t waste the user’s time

You are provided with no fewer than 28 questions, of which 27 are mandatory – who has got the time for all of these? Does it really take 28 questions to state my views on any website? It would not take long to whittle down these 28 questions so, say, five or at most ten.

Don’t ask the same question more than once

User surveys of this kind cannot deal in subtle shades of meaning. Yet in this survey many of the questions are very similar. For example, question 1 is “Please rate the options available for navigating this site.”. Question 2 is “Please rate how well the site layout helps you find what you need”. Are these not rather similar? Could you not replace them with one question, “How easy was it to find your way around the site?”

There is similar redundancy in questions 3 and 4: “Please rate the consistency of speed from page to page on this site” and “Please rate how completely the page content loads on this site”. In this case, the questions are too specific as well as being overlapping. There might be delays in loading pages – is this is a problem?

Don’t expect too much sophistication on the part of the user

  • “Please rate the ability to narrow choices to find the product(s) you are looking for on this site.” Which average user will understand that?

Don’t ask questions that only a usability expert could answer

  • “Please rate the balance of graphics and text on this site”
  • “Please rate the readability of the pages on this site”
  • “What is your overall satisfaction with this site?”

Don’t ask philosophical questions“How well does this site meet your expectations?”

“How does this site compare to your idea of an ideal website?”

Plato might enjoy mulling over the second of these questions, but for the rest of us, the site simply works or it doesn’t.

Don’t expect too many shades of meaning

  • “How likely are you to purchase your next grocery shopping from a Tesco store?”
  • “Please rate your level of confidence in buying from Tesco grocery website.
  • “How likely are you to use Tesco as your primary grocery supplier?”

Any one of these questions might be enough, but to ask all three is excessive. The questions about shopping in-store and online are subtly different, and I don’t see why. Rather than ask if the user is likely to use Tesco as primary grocery supplier, why not simply ask “Do you buy most of your groceries from Tesco”?

Don’t ask impossible questions

  • “How would you describe your overall experience with the Tesco groceries website? (Select one)














This question takes the biscuit. How wacky is it to have to choose just one of 13 adjectives? Does that mean none of the others apply? How can I summarise my experience of the website in just one term? The site was convenient, otherwise I would not have used it, but it was also slow. But was it more convenient than slow? Who knows? Usability experts must think that users have infinite time and knowledge to be able to articulate their responses to a website. They must expect remarkable powers of judgement on the part of respondents.

Allow space for users to make other comments

As it happens, I had one comment I wanted to make about the website. Because of coronavirus, I assume, all shoppers are limited to a maximum of 95 items in their basket. Since each onion, each baking potato, each pepper, is counted as one item, it’s very easy to go over your limit. You are therefore obliged to choose prepacked items to reduce the total number of items in your basket, because one bag of six onions counts as one item. How crazy is that? I am forced to choose more packaging, simply because of an arbitrary rule. But I would have had to answer 27 mandatory, and useless, questions, to state my one open-ended question. To be honest, I couldn’t be bothered. Tesco will be none the wiser, and my dissatisfaction will not be resolved.


How to make research more successful


What’s wrong with a payola model for Spotify?

1 Comment

  1. Very sensible. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by WordPress & Theme by Anders Norén