Numbers equal facts, right? False.
It’s not that the number itself is lying to you -- it can’t, it’s a number! But the story, the promise, the bias may skew your interpretation of the data insights you’re seeing.
Confirmation bias is a very real psychological tendency. It occurs when a person or group of people believe in something and then find data to back up that belief.
It is critical that you be aware of these “storytelling” biases that are so common in the world of marketing and business operations. While it may be entirely understandable that business owners or those reporting to business owners, CEOs, and the like, would want to present data insights that point to all things positive, it’s certainly not helpful. A business cannot grow and flourish atop false data insights.
But beyond confirmation bias there are other, perhaps even more manipulative ways, data insights can lie. Resources, processes, presentation, visualization, filtering -- each of these ways in which we interpret data may be inherently flawed to some degree. We’re here to help make you aware of some of the red flags surrounding these data insight falsehoods so you can be get closer to the truth of what your data is really telling you.
No, Not All Data Insights Are FalseWe are certainly not suggesting that all data insights are false or skewed in some way based on human interpretation. While human error can and should be taken into account, for example you may consider allowing a percentage allotment for human error based on natural bias, there are other more secure ways to debunk skewed storytelling when it comes to data insights.
If the data insights you are analyzing -- big or small -- are presented in a way that is appropriately categorized and benchmarked over precise periods of time, you can get much closer to the truth.
The Hard Truth About When Data Insights Lie...Don’t take it personally, but data insights do lie. And not just to you, but to everyone.
Knowing what common data insights tend to be skewed, or are otherwise flat out false, will help you and your team catch them quick and get to the truth. No one wants to fall into the trap of taking or changing the course of business action as a result of false data insights. This could quickly lead to the demise of the business or project. So, here are the more frequent data insight lies we recommend watching out for…
False Data Insight #1: Data Visualization TricksPerhaps one of the oldest data visualization tricks in the book, the truncated y-axis delivers a whopper of misleading information.
The lesson learned here, in addition to always looking closely at your x- and y-axis points, is that data insights can tell a very different story based on their visual presentation. Be wary of how any software or business dashboard (or executive) displays data in a way that may skew its meaning.
False Data Insight #2: Finding Patterns Where Patterns Don’t ExistIt is no understatement to suggest that the human mind is a complex system. When it comes to interpreting data insights, the brain can easily play tricks on itself. It’s pretty easy to do as we just learned from the simple truncated y-axis data visualization in example #1.
Pattern-making is not dissimilar in that the human mind sees what it wants to see. We tend to organize, label, and make patterns as a means of simplification and understanding. As a species this has served as well, but it’s not always the greatest asset when analyzing data.
Unlike a drastic increase or decrease that can be manipulated through charts or other methods of display, patterns that may or may not be real, can be formed in a variety of ways from a variety of data sources: numbers, code, and spreadsheets, as well as charts.
Data scientists are often specifically tasked to find patterns. This is no surprise as this is how we tend, for better or worse, to make sense of the world. But it’s important that we ask ourselves what kinds of patterns we are searching for, and to what end. And is a pattern really what we want? Let’s keep in mind that it is the pure data we’re searching for, and that may or may not lead to a pattern.
False Data Insight #3: The Right Place at the Wrong TimeTiming is everything. But really, timing does count for quite a lot when measuring and analyzing data insights.
Frequent errors occur when business owners or their staff look at a quick slice of data and automatically make assumptions without taking into account historical trends. A “slice” could be a month, quarter, or year, but there are times where it’s absolutely essential to view all of your historical data dating back to the beginning, or at least year-over-year, to get a clear picture of what’s happening. There will likely be factors such as seasonality, market, and economic trends to take into account that may influence the story you’re seeing. While one data insight may be interesting and may even tell the story you want your data to tell, it could be an outlier that isn’t really significant enough to make any kind of impactful difference.
One data point on one given day may be meaningful to some extent. There’s a time and a place for knowing and presenting obscure or truly stand-out factoids, but it’s the overall picture that tells the full story from beginning to end. This bigger piece is what should typically inform next steps for business growth. Flash-in-the-pan stats are meaningless when compared and contrasted to larger trends.
False Data Insight #4: Averaging the Wrong AveragesThink about it: if you’re calculating the average net worth of all the people in any given bar on any given night and Bill Gates walks in, the average net worth is suddenly going to be outrageously skewed upward. This is why averages may be guilty of telling the worst lies of all.
Data insights like this do tell a story and they may even trend data towards particular goals you’d love to see or report, but they ultimately say very little about reality. Minus the singular Bill Gates data point, you’re left with a very different picture of the average net worth in that room.
What you want to analyze are individual distribution data insights. There is a significant amount of learning you could do to truly understand averages, but the bottom line here is to be aware that averages can be exceedingly misleading.
How to Get to Data Insight TruthIt’s a sad, but true fact; numbers do lie. Our universal language of mathematics is not always as perfect as we may imagine it to be. Numbers can be used, averaged, and portrayed creatively and you don’t want to be on the wrong side of that manipulation.
One of the best ways to get to data insight truth is to be aware of the litany of falsehoods that are so often passed off as reality. Be smarter and know what to look for.
As a reminder, here’s a quick recap of what we discussed:
- Beware of presentation/visualization tricks
- Watch out for patterns that may not matter
- Consider historical data over time to understand the big picture
- Remember that averages can be misleading
Second is to utilize an all-in-one business dashboard that’s secure in its delivery of factual data. Appropriate filtering and visualization helps to greatly minimize your chances of interpreting false data insights that result in poor decision-making. Set up your dashboard so that you always have a historical view to reference. This will keep you anchored in reality without getting swept away in the immediate statistic-of-the-day.
Don’t waste time looking at inaccurate data insights. Getting comfortable around the right business intelligence analytics helps you and your team discover the truth faster, more accurately, and with a lot less fuss. This way you can be confident in the actions you decide to take that are informed by non-biased data insights.
If you’re looking for the truth, the whole truth, and nothing but the truth, get started with Cyfe today for free. No credit card. No obligation. Just insight! Grow your business by getting fast facts that result in action.