UX Research isn’t a bottleneck, it’s a decanter
I’ve noticed a concerning trend in UX in the past few months. Earlier this year, a friend sent me a post citing the shift at Uber from UX research to A/B tests as proof that researchers were getting in the way of decision making. At a recent virtual UXR conference, I read a number of exasperated Slack messages asking for help dealing with the pressures of researching under shorter and shorter timelines. Even in my own work, I received a quasi-compliment from a colleague for doing my job “without being a bottleneck.”
Sound familiar?
As researchers, we face a lot of pressure to reduce the nuance and complexity of insights, quantify findings, and not get in the way of launching a new product. And when we push back against these pressures, we run the risk of being labeled a “bottleneck.”
UX Research should not be seen as a bottleneck.
A bottleneck is a term that refers to something that obstructs flow and has been used for many decades to refer to manufacturing problems. It’s related to the Theory of Constraints: to maximize profit, you must find and eliminate the biggest constraint to production.
The premise of research being a bottleneck misapplies an industrial efficiency concept to new product development. It assumes UX research is one more “step” in the assembly line for new products and services. If this “step” happens to take longer than the amount of time allocated, it creates a “bottleneck” in the operation and lowers the efficiency of the plant.
But the purpose of UX research isn’t to increase the efficiency of a process; it’s to increase the creativity and quality of solutions to complex human problems. We’ve moved on from the industrial revolution. The most successful companies aren’t building the most widgets in the least amount of time; they’re figuring out how to bring the most value to as many people as possible.
Because stakeholders often see UX research as another step in the assembly line, they put a lot of attention on speed. Don’t get me wrong, speed is super important when you’re trying to invent something new. The problem a lot of people see with UX research is that to do it well, it takes some time. You have to carefully craft a plan, do the research, analyze and synthesize the data, and then share it. Impatient stakeholders have a hard time waiting for the research process to finish, and when they compare it to the time it takes to get an A/B test live, they might argue for skipping the research in favor of getting a test in the market ASAP.
This argument is short-sighted. For example, if you skip research in favor of A/B tests, what is your hypothesis based on? Intuition? Opinions? Without the research insights to inform an A/B hypothesis, you could conceivably run dozens of tests before you arrive at something effective. It’s counterintuitive to the industrial mindset, but research can help you reduce the number of A/B tests because you will start with a narrower set of hypotheses to test.
A/B tests are also popular because when you run one, you get to feel like you built a thing, even if it’s a small experiment. I always feel slightly insecure about my job security because I’m not a designer or an engineer building something. While researchers can thrive by showing how we improve the things designers and engineers build, we’re still not builders. We don’t have code or pixels to put in our portfolios. And if researchers fail to show the impact, it becomes easier to see our function as blocking, not building.
Jared Spool wrote brilliantly about the importance of showing the impact of UX because “a great design won’t look like it took a lot of work.” This is especially true for research. When a stakeholder looks at the role of research from a distance, it’s easier to see it as a bottleneck. Spool’s advice? Show the work. When you involve the team in the research process, not just the end result, you expose more people first-hand to the work, which gives the team a greater appreciation for the value research adds. One of my go-to tactics for showing the work is a video highlight reel. Pulling together a few short clips of the best moments from your research into a 2-minute video helps everyone see the work behind the insights.
If UX researchers want to shed this perception of being a bottleneck, we still have to do good research. A few years ago, GV’s Michael Margolis gave a talk titled “Start at the End” where he lays out a key principle for good research: research your team first. That way, you’ll know who the decider is and how you can make their job easier. It’s a great reminder that knowing your team is part of doing good research. Good research fits into the timelines of a project and makes everyone aware of the tradeoffs of moving faster (or slower). Research done for the sake of doing research is not good research. If the research doesn’t move the team towards a more confident, well-informed decision, then it’s not good research. For example, if your team has a hard deadline in three weeks but you insist on conducting a month-long diary study, that’s not an example of research being a bottleneck, it’s an example of poor collaboration. Don’t call bad research a bottleneck, call it bad research.
So if research isn’t a bottleneck, then what’s a better metaphor? I think a wine decanter does a much better job representing the contributions of research.
A wine decanter is a clear glass vessel with an open top so you can see and smell the wine much more than you could when it was in the bottle. It’s job is an important one; it lets you use more of your senses to evaluate wine before you serve it. In other words, it adds data to the wine by placing it in a different environment. Pouring the wine from a bottle into a decanter changes your perspective on whether it would be a good wine to serve or not. Does it smell good? Does it look appetizing? In a decanter you can study the wine so you have more data to work with. It’s just like research!
A decanter also lets the wine breathe. By exposing it to air you introduce oxidation and evaporation, which help any funky smells blow off and mellow out the tannins. In other words, decanting wine makes it taste better. With research, when you take your work and study how other people react to it, you learn how to make it better. You’re letting your work breathe. When it breathes, there’s a better chance it will succeed.
Research isn’t a bottleneck, it’s a decanter. It doesn’t restrict the flow of wine, it helps it taste better.
Thanks for reading! What are your thoughts?
