In our digital age, the line between enlightening insights and misleading narratives from data has never been more crucial. As professionals in UX design and data visualization, the choices we make have profound implications. Which truth will your data tell?
A dark pattern is a common UX term that describes experiences that were designed to trick users into acting against their interest. A popular example is making a subscription cancellation process more challenging than necessary. Another example is making a bad deal seem more appealing than it otherwise would be if presented transparently. Instead of solving a problem or helping a user achieve their primary goal, a dark pattern uses deception to trick users into helping a team or company achieve a business goal.
Although a more familiar concept with UX designers, dark patterns exist within the data fields too. The common thread found with dark patterns between the UX and data fields is the intent to deceive rather than inform. Ethical UX and data is about being transparent to users and audiences. You should always be aware of the fine line that exists between guiding and misguiding users. Dark patterns often go far deeper than simply deceptive designs. It reveals the soul of a culture.
For example, the Wells Fargo Cross-Selling Scandal is a good example of this. The key metrics Wells Fargo used encouraged employees to act against their customers' interests. Tellers, feeling the pressure, opened new accounts without customers being in the loop, all to meet impossible quotas. Reflect on how bad the culture must have been there. An honest conversation about fixing an obviously flawed performance metric couldn't happen before the situation got out of hand. This also shows how data encouraged poor governance, which resulted in huge fines and incalculable brand damage. And this isn't an isolated situation. From the likes of Enron to Bernie Madoff, many of the largest corporate scandals making headlines had data manipulation at their fraudulent heart.
Another data controversy can be found in academia. Data Coloda, an analytical blog started by three obscure behavioral scientists who closely examine evidence presented in academic research, recently gained attention after publishing a series of blog posts casting doubt about the validity of several studies associated with a well-known "honesty" expert due to alleged data manipulation. Instead of having an open conversation in response to these important questions, those who brought attention to these concerns are being sued. This demonstrates what kinds of real consequences pursuing data transparency can get you.
Data can be manipulated in a way to create better "storytelling" when finding truth is no longer a constraint. Going down this road may allow you to show data in a way that impresses investors or get funding for your research, but these self-serving approaches are not solving real customer problems nor expanding real academic knowledge. It's a look at yourself in the mirror type of personal decision about what kind of data professional you are or going to be. Y
You cannot ethically “fake it until you make it” when working with data, and you should always act as if the truth will eventually come out. Once you start faking data, then the gap between reality and your narrative will grow wider each day from that point and will become harder and harder to hide as time goes on. You will be spending more and more time on keeping up the narrative rather than using this time to investigate real problems or make tangible improvements.
Fudging data may allow you to show more success or avoid tough conversations in the immediate short-term, but obscuring data forever is usually impossible. If you've ever created a realistic sample dataset, then you know the surprising challenge of making a trendline from fake data appear both realistic and compelling. Hand-waving these complexities away is pure hubris.
Data and insights often exist in a realm of ambiguity that can be unsettling. The inconvenient truths that data can sometimes reveal provides another kind of discomfort. Yet, you should never mask these issues with misleading analysis nor go further by altering the data itself when sharing insights with your audience.
The temptation will always exist to make data look more interesting than it actually is or to tip the scales to boost your business metrics without actually doing anything substantive. Yet, data professionals know playing loose with data is a self-defeating proposition and maintaining the integrity of data that people make critical decisions on is way more important than whatever short-term benefits you can gain by being unethical.
Saying "I don't know" when lacking evidence contributes to your credibility as much as sharing a compelling insight backed by solid evidence. The world is complex. Problems you are trying to solve are often messy. Not knowing an answer to every question is okay. Data cannot underpin every decision. You don’t want to set the expectation that you are a superhero to stakeholders, who can find gold in any dataset. This would only add to the pressure of finding a compelling story when one doesn't exist.
Furthermore, you will make mistakes. Yet, as long as you have a transparent approach with the right mindset, then you have every reason to find the confidence to explore the toughest data questions existing today. You should never allow internal or external pressure of getting a specific result influence your analysis. So, trust your process and trust your experience. The right people will respect you for not trying to BS them.
While possessing skills in storytelling and visualizing information is a part of data visualization, the field relies even more on an honest mind when constantly dealing with ambiguous questions and complex decisions. Especially when temptations exist to expand one's ethical tolerance for personal gain, the greater the number of us who embrace responsible data practices, then the more meaningful the results we can bring forward in our collective pursuit of truth.
