The Top 3 Ways to Make Your Organization Truly Data-Driven

Most organizations consider themselves driven by data. What they mean is that their spreadsheets aggregate a few data sources into KPIs each day. If they're more sophisticated, they have scaled pipelines that do the same thing with a pretty reporting tool. But this is being data-driven like standing on a splash pad is being a freediver (diving deep without a scuba tank).

Being data-driven means that data makes you make decisions you wouldn't or couldn't otherwise make. Being data-driven is uncomfortable. It's eye-opening. It's, at times, sobering.

But being data-driven is also hugely empowering because data allows you to measure the impact of your decisions, learn from them, defend them with clarity and confidence, and reach new peaks of performance.

How far have you been? How far do you want to go? What will it take to make it?

Data tells you.

As a freediver who must make optimal decisions about his goals and limits, I've learned a lot about using data to achieve my best results. These lessons also apply to organizational performance.

Want to make your organization more data-driven?

Here are the top 3 ways.

1 - Measure change, not the past.

Data should compel change. If it doesn't, what is it for?

Start with this question:

What do I want to improve?

Is it customer retention? Your manufacturing plant's efficiency? Your engineering team's development speed?

Maybe you want to reduce the time you spend on production support tickets, or cut your cloud costs, or increase delivery route efficiency.

As a freediver, I want to improve the amount of time I can stay at a modest depth (bottom time) and my safety margin, not my max depth (though depth is the easiest metric for non-freedivers to understand). Many freedivers measure their progress in feet. I measure mine in seconds. Knowing my goals helps me gather and utilize the right data to achieve them.

Figure out what you want to improve, then use data to track your progress and help you decide how to improve it. Watch your metrics change as a result of your decisions, publish your progress, and adapt as needed.

2 - Deliver discomfort, not confirmation.

If you or your customers react to your data with, "That's fine; same as yesterday," then no one is using you're data to make decisions and you're not data-driven.

The two most telling responses to valuable data are:

"Oh $hit, what a great opportunity," and its fraternal twin, "Oh $hit, what a terrible mistake."

and

"Huh, I wonder what that means."

As a freediver, I don't need data to tell me I made it to 20 meters again. And again. And again. That's what a report does. I need data to tell me I spent 10 seconds at 20 meters on the first dive but could only spend 5 seconds on the next dive. I won't like it, but I'll use that to make different—better—decisions about my third dive.

Discomfort compels action. Confirmation—i.e., comfort—never does and never will. Deliver data that—in the words of my friend and Data Product Owner Adolfo Alvarez —gets people out of their seats to do something.

3 - Be not afraid, be not mad.

The biggest reason organizations are not data-driven is that they're afraid of what the data will say.

Maybe your development speed is embarrassingly slow. Maybe your customers are churning at an uncontrolled rate. Maybe your marketing spend isn't returning a positive ROI.

As long as the data doesn't show the problems, they don't exist, right? Or, at the very least, no one knows about them, so they can just stay under the rug where they belong.

The risk is that your organization will continue making the same decisions that produced those problems until they become too expensive to fix. Then, no one wins.

It's far better to confront bad numbers head-on and let them guide your decisions.

As a freediver, fear is truly the mind-killer, as Frank Herbert wrote in Dune. A freediver cannot be afraid or ashamed of what the data tells them. Some days, you won't even make it to your warmup depth. Heed the data from your computer and your body. Don't act recklessly. Adjust and find a way to perform better tomorrow.

Here are three stories from my career about the consequences of fearing data:

1. One of my stakeholders wanted data to support the claim that spending more money on the product yielded better results for the customer. The data I had did not back that up. This stakeholder did not want to share the data, even within their org, for fear of how it would be received. They continued making the same claims.
2. My team discovered that the formula our stakeholder used to measure market penetration was wrong. Our correction dropped their penetration number by 20 percentage points. They agreed the new formula was correct but resisted adopting it because it would make them look like their performance suddenly tanked. This story has a happy ending: We persuaded them to look at both formulas over 6 months until they got calibrated to the new number. They eventually replaced the wrong one with the right one.
3. In an effort to increase the value of my team's data product, I asked our senior product manager if we should measure the usage of our reports and data sets. They said, "What if the number is low?" We never measured it.

Be not afraid of what the data might say. Be excited. Use it to show everyone what you are doing well and what you need to improve. Then make decisions to improve your results and deliver true value.

None of these strategies requires new technology or more data, necessarily. They require understanding and accepting how data produces value and making the requisite mindset changes in your organization.

Unless your organization is data-driven, how will it know how far it should go, what its limits are, or how far it has come?

How much bottom time do you want?

Follow any of the above strategies, and you will be well on your way to becoming truly data-driven at work or in your own personal life.

What are your best strategies for creating a data-driven organization?

For those interested in how a freediver uses data to help them reach peak performance:

A freediver needs to know how deep they have been and how long they have held their breath in order to safely set their next goals. Without a baseline, new goals are dangerous. If one's real baseline is 5 meters, don't attempt 15 meters.

A freediver needs to know how much time has elapsed since their last dive (their surface interval) so they have enough time to catch their breath, process lactic acid, and offgas CO2/nitrogen build-up in their bloodstream before their next dive. Surface intervals increase as dive depth and time increase.

A freediver needs to know what depths and times they can achieve in different scenarios: diving with weights and fins, diving without weights and fins, floating still, or swimming horizontally in a pool. Without data on those different disciplines, they may set dangerous targets.

Equally important, a freediver's safety diver and/or coach needs this performance data to decide when they may need to intervene and rescue a stricken freediver before it's too late.

Freedivers use dive computers to gather this data so they can make better—safer—decisions than they otherwise could, and reach their peak performance.

To learn more about freediving and why I do it, check out my travel blog post: https://themanasas.exposure.co/why-i-freedive        

Editing help by Jai Dabolkar .

Thanks to Franziska Moenster for additional details about surface intervals.