Swiss Cheese Mistakes: How Data-Driven Decision-Making Can Help Fill the Holes
Connecting the Dots: Navigating Mistakes and Organizational Factors in Data-Driven Decision-Making
Recently, I've been invited to give a presentation about data-driven decision-making. The customer was interested in the tools, systems, and workflow that their employee could learn from and implement. This is a topic that hits very close to my heart. Over the years, I have accumulated hundreds of pages of notes on it, following my experience with different companies, including the culprits, behaviors, personals, and actual methodical systems that companies can implement now.
As I reviewed my notes, I got a flashback of a previous experience where the plane hit the ground hard. That was a mistake where there needed to be more data, and one very opinionated, strong-headed person completely defocused the group. I marked it as the Swiss cheese model failure.
While Swiss cheese is delightfully delicious, it has many holes. We can call these holes mistakes. Everyone makes mistakes. You can either learn from it or repeat it.
Let's explore how it is connected to data-driven-decision.
Mistakes happen when:
- Wrong processes
- Blackouts during the process – we don't fully execute everything; we take a shortcut, skip or forget. (This is different from the optimization of a system).
- Good process is being executed wrongly.
3 domains where mistakes happen:
- Skills ability.
- Legal/ adhering to the system steps and procedures.
- Knowledge.
When thinking about data-driven decisions, many people focus on gathering the data, yet they often miss on the skills it takes to turn them into knowledge; for that, we need to carefully examine what that decision entails and who are the people involved:
People involved: manager, team, peers, friends.
Technical conditions: equipment, work environment.
Organizational factors: timing, the definition of tasks, and availability of tools.
External implications: time, financial climate, people/industry mood, weather.
Data should be collected at all those levels and built into the organization's planning to make data-driven decisions. If you think about a business, it should plan for:
Demand
Cost
Workflow
Skills
Finance
CapEx
OpEx
Each one of the 7 layers of business planning is critical, and none should be taken as obvious. This is what makes data-driven business decisions hard. The culprit is in connecting the dots; how can we plan for skills, given the demand from the market? Which skills are missing? Who is capable of growing into them? These are more soft metrics than analytical ones, based on careful observation of the organization, where many people make mistakes.
Back to the Swiss cheese, people make mistakes and defocus.
This model describes how organizational factors lead to mistakes; it compares the various levels where mistakes occur. Think about very thin slices; there are many holes in each one. And they are in different places. Those are mistakes. A mistake will only appear or be relevant if it exists in multiple slices. But it can become a disaster if all the slices have holes in similar places that connect to each other. (this is the cross-organizational analogy). This mistake would go over all the holes and become a big organizational failure.
The reason why we need to carefully examine each hole and its impact lays in the company focus. It is easy to focus on one hole in one piece and try to fix it. But it would deteriorate from the larger picture and blind the organization from a potential plan crush.
When working with data, one of the big challenges is the context needing to be recovered. When using multiple systems and moving them around, the true meaning of it is lost, which can easily distract us from seeing the full picture and taking a data-driven approach. This is when opinionated, strong-headed, shrewd salespeople can distract a whole business. Take that into account when you start planning your business data-driven journey and processes around it.