Are you in love with your data? You’re probably in a state of denial about its integrity

Are you in love with your data? You’re probably in a state of denial about its integrity
Photo by Mika Baumeister / Unsplash

I can’t think of any tech leaders who don’t declare that their company is data-driven. Even those who formed their companies without data as a pillar for their decision-making are pushing their companies is this direction, and rightfully so.

Making decisions based on data is what enables companies to scale their operations, and improve with each iteration. Although data can’t replace the human aspect of good decision-making, biases often confuse fact and opinion.

Building processes for making data-driven decisions takes time, because, like any feature, it’s an iterative process of trial and error, learning what to measure and how to interpret the data in an actionable and correct way.

Numbers trump feelings

A problem that companies frequently face when aiming to become data-driven is that numbers trump feelings, even if the numbers are completely wrong. We all know what it feels like when attending a discussion during which a participant brings numbers to the table. The data becomes the reference point, regardless of their accuracy or statistical significance.

In fact, having wrong or incomplete data is much more dangerous than having no data at all. It’s amazing how many strategical, pivotal decisions are made based on inaccurate and incomplete data.

Once you’ve reached the point of converting data to conclusions, questioning the data’s integrity becomes very hard. Not only you are fighting the biases mentioned above, but fixing data-integrity problems can take time and effort, as missing data points need to be added and missing information has to be re-collected. As a result, more often than not, you are liable to enter a state of denial, especially if the data fits their pre-conceived theories.

Why is this happening in the first place?

The best way to tackle this problem is to avoid it. We discovered that one of the main reasons companies find themselves in this predicament is the employees’ understanding of the importance of data to the organization.

We found that, on average, close to 98% of tech companies’ leadership members place an extremely high value on data-collection However, more than a quarter (27%) of employees, across R&D, sales, and customer-success teams severely undervalue the importance of data-collection for the future of their company.

Is this really the source of the problem?

When a company’s engineers don’t view data-collection as extremely important, they will not take it into consideration in their designs, and might manipulate the numbers to optimize other parameters (i.e. storage costs, speed, privacy, etc.) They might also push data out of the scope if time is running out and customers are waiting. This type of problem has the potential to wreak serious damage over time, hurting a company’s ability to iterate and progress in the right direction.

Similarly, sales and customer-success team members might fail to keep their CRM tools with all the data in a structured way, making it difficult to connect the dots between what worked well and is worth doubling down on, or to abandon that which didn’t work at all.

Once you’re aware of your weakness for numbers, you should make sure that the data on which you’re basing your strategy is complete. You can avoid a significant pitfall by ensuring that the people collecting the data are aligned on its importance. This will help you make better decisions and save your company from damage.