I spent 20 years at Intel. Absolutely everything is measured and monitored there — manufacturing processes, product development, self-executed performance evaluations, everything. If there isn’t a metric for quality or quantity, one is invented. “High-medium-low” is sometimes the best that can be done. Some metric is always better than no metric, the thinking goes, because you can’t improve what you can’t measure. The name of the game is continuous improvement. It’s been that way from the beginning, which makes Intel one of the earliest truly data-driven decision-making organizations.
What it Means to be Data-Driven
The degree to which decisions are data driven varies greatly from company to company, and sometimes from group to group within a company. When a company has strong metrics in place for sales and production, but not for HR or R&D, it is probably not a data-driven decision maker. Companies that think they are data driven but really aren’t tend to make decisions based on intuition, then collect data to support those decisions. Truly data-driven companies have tangible indicators:
How Organizations Become Data-Driven
Many AI projects rely on real-time, increasingly granular data. Decisions are made at a correspondingly granular level. Integrating Enterprise AI (EAI) throughout an organization requires that employees further down the organization chart be empowered to make decisions historically made by upper management. For example, technicians need to be granted authority to take a machine out of production when a signal predicts imminent failure. Customer service reps need autonomy to offer a promotion to a customer when they sense a risk that there will be a switch to competitors.
Leadership by the highest-paid-person’s opinion is the opposite of being data-driven. Senior leaders who are used to making decisions based on their experience and opinions may struggle to relinquish control, but change only comes when leaders let data override intuition. Good data can provide valuable, verifiable insights that leaders have to learn to trust. Managers should model the same behavior, and educate themselves about the origin, accuracy and quality of the data they collect and use.
When faced with a new problem, data-driven decision makers ask, “What do we know?” before they ask, “What do we think?” The human part of decision making doesn’t go away in data-driven organizations, it just becomes better informed.