The PDCA continuous improvement cycle

At a food facility in Amman, the loading area logged six slip incidents in a single quarter, one of them ending in a two-week sick leave. Management demanded a "final fix," so the maintenance officer bought slip-resistant boots, handed them out, and closed the file. Three months later: five new incidents, in the same spot.

What happened wasn't a lack of effort — it was a lack of order: a jump straight from problem to action, with no analysis before and no measurement after. Every ISO standard is built on a different sequence called PDCA: Plan, Do, Check, Act — the Deming Cycle, the backbone of every management system from quality to safety to environment.

The cycle on the facility floor

Go back to the slip incidents and apply the four steps properly:

1. Plan. Start with the data, not the solutions: all six incidents happened in the loading area, four of them in the morning during floor washing, where water pools at the ramp. Set a measurable goal: cut slip incidents in the loading area from 6 to a maximum of 2 next quarter. Then set the plan: move the washing schedule outside peak traffic, install drainage at the ramp, slip-resistant boots, signage — each action with an owner and a due date.
2. Do. Execute the plan as written, and document the execution: when the drain was installed, who received the boots, when the washing schedule changed. This documentation isn't bureaucracy — it's what makes the Check step possible at all.
3. Check. At the end of the quarter: three incidents. The target was two, so it wasn't fully met — but a good check doesn't stop at the number: all three incidents happened on the night shift, which turned out to never have adopted the new washing schedule at all. Measurement exposed a gap nobody could see.
4. Act. Roll the new schedule out to the night shift, lock in what worked as a permanent standing procedure rather than a temporary trial, then open a new cycle with a sharper goal. This step is the difference between a one-off improvement and one that compounds.

Why a cycle, not a project that ends

The difference between a company that "solved the slip problem" and one that manages safety is that the first closed a file and the second keeps turning: each cycle starts from the level the previous one left it at. A system that stops turning ages silently — equipment, people, and risks all change while it stays exactly where it was.

Because PDCA is a structure, not just a tool, you'll find it in the very architecture of ISO standards: the context and planning clauses map to Plan, the support and operation clauses map to Do, performance evaluation maps to Check, and improvement maps to Act. So when an auditor asks you about "continual improvement," they're not asking for a slogan; they want to see at least one complete cycle: a measurable goal, documented execution, an honest check, and an adjustment built on it.

The cycle works at every scale: across the whole system in the annual management review, across a single department, and across a single problem like the slip incidents. The skill is the same; only the scope changes.

The cycle's memory is its records: written Plan goals, Do evidence, Check measurements, and Act decisions are what let the next cycle pick up where the last one left off instead of rediscovering everything from scratch. A team whose members rotate but keeps its records improves; a team that stays fixed but keeps no records repeats its mistakes with new faces.

Common mistakes

1. Stopping at Do. Plans that get executed and never checked. Test yourself with one question: "The action we took last year — did it work?" If you don't have a measurement to answer that, you're active, not improving.
2. Plan without a baseline. "Improve safety" isn't a goal; "from 6 incidents to 2" is a goal. Without a number before execution, no number after it means anything, and the Check step turns into impressions.
3. Act as punishment. When a measurement deviates, a mature system asks: "What in the way we work allowed this?" — not "Who do we blame?" Punishing people pushes everyone to hide data, and the Check step loses its raw material.

In goiso

goiso is built on this same cycle. Plan: set up your sites, assets, and team, activate the standard, then let the seed wizard turn clauses into inspections and tasks distributed across sites and team — see how the seed wizard works. Do: execute from the inspections and tasks boards. Check: the KPI board measures performance against defined thresholds and plots the trend from periodic snapshots — see how to read the KPI board. Act: whatever deviates opens a corrective action tracked on its board through to closure. The whole cycle lives in one place instead of being scattered across notebooks and spreadsheets.

Summary

• PDCA = Plan, Do, Check, Act: the Deming Cycle, the backbone of every management system standard.
• No Plan without a measurable goal and a baseline, and no value in execution that will never be checked.
• Check reveals what the eye can't see — like the night-shift gap — and Act locks in what worked and adjusts the system, not the people.
• The cycle never ends: every Act opens a new Plan from a higher level.

And sooner or later, in the Check step, you'll run into a gap between what the standard requires and what's actually happening. That gap has a name and a structured way to handle it: nonconformity, the subject of the next lesson.