Get contextual feedback from your users
It's hard to speak to your users every week, especially in a B2B setting. B2B users are busy, and hence recruitment is difficult and expensive. Yet, even when you run lots of quantitative experiments, qualitative insights are highly valuable.
One method that has worked really well to gather lots of qualitative feedback without full-blown user interviews have been contextual feedbacks with rating and open question. So when the user completes or abandons a task (e.g., navigating 3 or more of your sites or clicking a sequence in your app), you ask for feedback.
Here are some learnings from gathering 5k+ per months for the past two years:
- Keep the question general, better use "Did we help you reach your goal?" rather than "Did our feature X help you?" — users don't understand your lingo and want to solve their problem, not use your feature
- Just use thumbs up/down, don't use star ratings or emoji feedback; every user has a different understanding of what 4 stars mean and more options lead to lower response rates
- Run the feedback only for a limited time but with high visibility—show it only for a couple of hours or day, but make it very prominent. This way, you gather lots of feedback but don't annoy returning users
- Create labels on-the-fly—don't overthink categorization by thinking about possible answers, start with a label such as "User likes navigation" and reuse if other feedbacks match exactly or add new labels such as "User dislikes color". You can still combine labels later, but you need to abstract text feedback into some categories first as this will help you to group them and find patterns
- Make both raw data and analysis available to everyone in the organization
- Proactively share the insights that you gather in every product and UX meeting, overshare massively to make teams actually use the insights.