Cause-and-Effect (Fishbone) Diagram
A structured framework to systematically map and categorise the root causes of a failure before attempting to design solutions.
§1 Understanding the Framework
Developed by Kaoru Ishikawa, the Cause-and-Effect diagram forces teams out of immediate, often flawed, conclusion-jumping.
When a complex problem occurs, human nature frequently dictates fixing the most obvious symptom. The Fishbone diagram counteracts this by providing a visual map that branches out into categorical domains, guiding a team to explore all underlying contributors. The "head" of the fish represents the problem or effect, while the "bones" extending from the central spine represent the diverse categories of potential causes.
This templated approach guarantees that investigations do not stall in isolated silos. By providing exactly six entry vectors, analysts map environmental constraints next to technological hurdles, creating an unbiased, holistic view of the failure landscape.
Contents§2 Standard Categorisation Methods
While the six branches can be entirely custom, they are most frequently populated using established industry mnemonics. Select the grouping that closest matches your sector:
Manufacturing (The 6Ms)
Machine, Method, Material, Manpower, Measurement, and Mother Nature (Environment). Highly suited for physical production environments and hardware engineering.
Service Industries (The 6Ps)
People, Process, Policy, Plant (Technology/Facilities), Program, and Product. Customarily utilised in software delivery, consulting, and administrative processes.
§3 Worked Example: Website Performance
A breakdown from a real-world digital agency post-mortem outlining compounding issues.
Consider a scenario from Mary’s Consulting, where a newly deployed company website experienced severe performance degradation. The defined effect was quantifiable: Site performance fell below the LCP < 2.5s target on image-heavy pages. Instead of merely blaming the server size, the team broke down the contributing elements using a tailored set of categories.
| Category | Identified Root Causes |
|---|---|
| People | Limited frontend performance optimisation experience within the delivery team. |
| Technology | Excessively heavy JavaScript bundle; inefficient CMS rendering layers; absence of code splitting. |
| Tools | No performance checks configured on Pull Requests (PRs); absence of a Content Delivery Network (CDN) for static assets. |
| Process | Performance testing was isolated to the very end of the cycle; no enforced performance budgets during the build. |
| Materials / Content | Too many distinct font weights loaded simultaneously; vast, unoptimised imagery uploaded by content editors. |
| Environment | User latency magnified by mobile 4G network conditions; staging/production connected to a slow hosting tier. |
By visualising this spread, the team identified that the solution required a combination of upskilling (People), infrastructure changes (Tools/Environment), and stricter build pipelines (Process/Technology), rather than applying a single patch.
Contents§4 Implementation Guide
To deploy this framework effectively in your next retrospective or incident analysis, follow these four chronological phases:
- Define the Effect clearly. Draw the head of the fish. Write the problem statement. It must be specific, quantified where possible, and agreed upon by the entire group.
- Establish the Categories. Draw the spine and the six angled lines. Choose between the 6Ms, 6Ps, or a custom blend (e.g., People, Technology, Tools, Process, Materials, Environment).
- Populate the Causes. Conduct a brainstorming session. For every symptom or hurdle raised, categorise it onto the appropriate branch.
- Ask "Why?". For every cause attached to a branch, interrogate it with the Five Whys technique to ensure you have reached the systemic root, rather than a surface-level symptom.
1. Target Precision
Vague effects yield vague causes. Always use numerical thresholds (like LCP < 2.5s) to ground the analysis.
2. Team Diversity
Never populate a diagram in a vacuum. Bring in varied disciplines to avoid missing vital categories like Environment or Process.
3. Next Steps
A diagram solves nothing on its own. Translate identified root causes into a Pareto chart to prioritise the immediate corrective actions.