This is Part 3 in a series about Collective Data Empowerment. If you want to get the whole series and accompanying tools in an ebook, go here.
Think about a challenge your company faces. Consider: who among your colleagues knows the most about it? Who has the most experience with it? Who feels its weight every day? They probably aren’t all analysts and data scientists.
Now recall a data analysis that was focused on the challenge. Were the people you just pictured in the room, or on the thread, from the beginning? Were they consulted and considered throughout? If not, you might have a homogeneity problem (pattern #5).
These are 16 patterns we see in companies that are struggling to evolve into data-driven cultures. They share a root cause:
Most companies have invested more in Selective Data Empowerment than Collective Data Empowerment.
In the first post of this series, we distinguished between the two:
On to the patterns…
1. Unused data
People don’t know what data your company has, and can’t find data when they need it. How can you use something if you don’t know it exists?
2. Debilitating delays
Data and analysis arrive too late to capitalize on opportunities. People struggle to understand it, and their questions aren’t answered fast enough.
3. Wasted resources
High-paid data people spend most of the day making data more usable, answering the same questions, and refreshing reports. They can’t focus on the advanced tasks that only they can do.
Data scientists, according to interviews and expert estimates, spend from 50 percent to 80 percent of their time mired in this more mundane labor of collecting and preparing unruly digital data, before it can be explored for useful nuggets.
— Steve Lohr, The New York Times
4. Lost knowledge
Teams lose the knowledge they create if it falls outside of narrowly-defined project goals. If it doesn’t have immediate value, and there’s no obvious home for it, it falls through the cracks.
5. Homogeneity
Project outcomes reflect the knowledge, abilities, and biases of a small, homogeneous group. These advanced practitioners convey insights that seem cryptic to most of their colleagues.
“Very few companies expect only professional writers to know how to write. So why ask only professional data scientists to understand and analyze data, at least at a basic level?”
— Jonathan Cornelissen, Harvard Business Review
6. Needless repetition
When people can’t see what others are working on, they waste time and money on duplicative work.
7. Puzzling barriers
Access controls and other restrictions persist without regular reconsideration and changes. It’s the classic battle between legacy and logic.
8. Lethargic data literacy
Data and insights do little to elevate data literacy if people can’t trace the path from question to answer. Fancy tools can boost the productivity of data elites, but they don’t lift the general baseline by themselves.
9. Competing definitions
You are entitled to your own opinions, but not your own facts. What about your own definitions? When Team A defines something one way, and Team B defines it another way, an innocent disconnect can cause serious problems.
10. Dominating HiPPOs
Data-driven cultures do not put the Highest Paid Person’s Opinions above the data. Carl Anderson, who wrote the book on data-driven cultures, describes the HiPPO’s destructive tendencies:
These are the know-it-alls who, almost on principle, ignore the data, ignore the evidence and recommendations, and do what they want because they know best (after all, their paycheck proves it!).
— Carl Anderson, oreilly.com
11. Reproducing nothing
The reproducibility crisis in the scientific community should make businesses interrogate their own data practices. When people can’t reproduce an analysis, they can’t check it for accuracy or reuse it.
12. Leaps of faith
People make choices based on insufficient, inaccurate, or outdated data and analysis — when they use data at all.
13. Dark data
Too much data work and analysis takes place without transparency, audibility, or accountability. Signs include rogue databases, unsanctioned software, and too many emailed spreadsheets.
14. Separate lanes for data, analysis, and context
Getting the full picture of an analysis or dataset requires looking for puzzle pieces in different places. The context people need so they can use data gets stuck within email threads, one-off DMs, and hallway conversations.
15. Analysis paralysis
People wait for the “perfect” data before deciding, even though the data they can access should be enough. A mix of factors drives this behavior, ranging from low trust to the steep learning curves typical of data tools.
16. Low trust
It’s hard to trust data if you don’t know things like who has touched it, where it comes from, and how it has changed over time. No wonder 56 percent of CEOs worry about the integrity of the data they use for decision-making.
Ready to get data-driven? Get right to it and download the complete guide to building a data-driven culture through Collective Data Empowerment.
Check out the rest of the Collective Data Empowerment series!
(Editor’s note: This post was updated on 9/19/2018 to add new resources and reflect the completion of the initial series of posts.)
