What’s the Problem With Proxy Users?
In this era of product management, I think we all agree that user research and feedback is important. We also know that sourcing users for this research can be difficult. So we often turn to proxy users in our research. Proxy users are people who are not actually your users, but are close enough that they know what your users would want… up to a point.
Common proxy customers are people like your sales team, the manager of your users, consultants who work with your users, people who used to fit the persona of your users but don’t any more. These people can be very valuable sources of information when used properly, but there is a very big danger in using them as your only source of research.
Why is it bad to use proxies? Well, the main reason is obvious: They’re not actually your users! The purpose of talking to users is to understand their needs, understand their behaviors, see how they interact with your product. And you can’t truly get at that by talking to a proxy.
You Only get the Highlights
Proxies are viewing the actual users from their own angle and through their own biases. I once worked on rebuilding an internal product for a client’s support team. The first person they gave us to interview was the manager of the team – a proxy. She had been managing the team for many years and had even done the job at one point. However, it turned out her views of the ideal workflow and the views of her team were different. In a product that was already very information-dense, she insisted her team could not do their job without additional data presented on the screen. But when we spoke with the actual team and shadowed them on their job, they didn’t look at this additional data at all.
Proxies also lose the nuance. They are the human equivalent of a revision guide – they can give you the highlights, but if you try to dig a level deeper, there is nothing there.
It turns out that this manager wished her team would check this data, but doing so wasn’t important to completing their tasks and actually slowed them down when their main goal was number of orders closed per day. If we had built to the manager’s specifications, we’d have built a product that was less usable, and with a potentially negative impact on its users. By speaking to the actual users, we were able to build a clean, simple workflow that allowed the team to process more orders than before.
Proxies also lose the nuance. They are the human equivalent of a revision guide – they can give you the highlights, but if you try to dig a level deeper, there is nothing there. Your users are very complex people with a lot of different motivations and feelings on a given day. Often, they aren’t even aware of small subtle things they may say or do that can drastically alter what we build. When we interview users, we often hear an answer, and immediately ask “can you tell me more about that?” or have the opportunity to get clarity on answers that could be interpreted multiple ways. If you’re only talking to proxies, you are going to miss out on the wealth of information that is sitting just below the surface.
How can Product Managers fix the Proxy Problem?
So why do we use proxies? Because, frankly, it’s easy. Particularly in enterprise organizations, there can be a lot of political churn and concern about talking to actual users. You will run into roadblock after roadblock trying to get to an end user, and everyone else in your organization will try to convince you that the proxy is really all you need to talk to. But if you fall into that trap, you can end up building product on very incomplete data. We need to be the ones to push our companies to think bigger and not get hung up on the things we think we know and the fear of what might happen. We need just to get out there and actually talk to our users. Our users want to tell us what they think. They want to be a part of shaping our products. We mustn’t diminish their voice by relying on proxies.