Believe it or not, in my opinion, survey and feedback poll design is one of the hardest forms of user research as the propensity for bias feedback increases exponentially. Rebecca Weiss, a PhD candidate at Stanford and team member at Mozilla, said it best: “if respondents don’t understand your question in the exact same way and can’t respond easily, you will get measurement error”
Together, we’ll understand your business and user objectives creating and supporting insights with relevant data. Where no data is available or further data is required, a survey or multiple feedback polls can be extremely useful. It’s important to write down your research goals before hand, and afterwards any statistical significance should always be further verified using other tools.
User Testing Tools
- Sending out multiple questions to a user base
Here, user data is collated and all users are sent a promotional email by the company containing a link to the survey. This can be seeded on social channels too. As users are answering anything up to 10 questions and spending a few minutes or more of their time, it is usually recommended that this runs in conjunction with a competition.
- Posting an individual question or poll
A feedback poll is a one-question, one-answer element that appears on the site, subtly, as the user is browsing the site. As this is only one question, there is a significant reduced level of commitment and over time, you can collate lots of qualitative and quantitative data using this method.
For Confetti.co.uk we wanted to understand the purpose and reasoning of users coming to the homepage. After several user interviews, it was determined that users were looking for inspiration for their wedding; of different sorts. I asked users “What are you looking for today?” via a feedback poll that popped up, subtly, in the users corner of their screen. Wanting to keep the question small and simple I didn’t believe it required a great amount of effort on the user’s behalf. The results were a mixture but the overall response was that users were just browsing which conflicted against my original results. Wanting to determine the cause of this, and to ensure it’s confidence value, I amended the question to “Are you looking for inspiration for your wedding?”. 84% of users said no. 3rd time lucky. I tried again, this time being even more specific and with a small introduction of friendliness – some colloquial language always puts people at ease. I also kept the word “today” in the poll to associate some form of urgency in the question. “Hello! Are you looking for some inspiration today for your wedding?” Again this proved inconclusive when trying to verify my results. After trying to understand the user’s mindset, it was understood that users answering such a question feel some form of commitment. Also, users weren’t directly associating this question to a poll, but instead assume that you will forcibly interrupt their journey.
The final iteration of the question validated the user interviews and was as such: “Hey please can you quickly help us out by answering this one quick question – why have you come to Confetti today?”.
Get in touch
If you’d like to learn more about a structured UX approach I’m more than happy to discuss this with you and your requirements. I’m certain with my track record in using this approach I can facilitate in a proven methodology that will get the best out of your team and the most efficient user-centric solution for your client.