Let's take this conversion funnel as an example. What it tells us is that many people drop-off and leave at the homepage and the pricing page.
What it doesn't tell us is why these people leave. Here are some potential reasons:
- The benefits and the product value proposition are not clear enough
- The price is too high
- The visitor was just browsing around and there is nothing wrong with the product
- The product is trustworthy enough
The only way to know, is to ask these people. This can be done either through user testing or through other feedback mechanisms like feedback widgets (example).
In general when you build a product you have two types of data you can look into:
- Quantitative data that come from analytics (e.g. conversion rate, funnel drop-off)
- This data help you to understand the big image, spot areas that need improvements and define priorities
- Qualititative data that come from user feedback (e.g. user testing feedback, testimonials)
- This data help you to understand why these problems exist and what are their potential solutions (e.g. understand why people drop off during sign up)
Your goal is stay human and keep balance between quantitative and qualitative data.
A problem that I've seen is the industry is that companies focus way too much on quantitative data. They want to make everything an A/B experiment, they want to track all kinds of events and KPIs and start acting like robots, forgetting that they are building a product for humans with real emotions.
Personally I prefer to follow the human-oriented approach and use my common sense. Data are nice and helpful to look at the bigger image. To understand if I am doing well or not. In no case they can replace user feedback and user testing.
User testing doesn't only tell me why people are dropping off but also how they feel and what they think at that moment. My goal is not only to make them convert and get their job done but to also create something that is fun to use and brings a bit more beauty into their lives.