In many companies, you often hear opposite opinions a few weeks apart. One month everyone is excited about a new management software that "changes everything"; the next month people go back to doing things the old way, because the software was never really adopted. The problem isn't the tool. It's the idea that improvement has to come from a single event.
Continuous improvement is the opposite approach: a series of small daily adjustments, made by the same people who do the work, which over time produce measurable effects. The reference philosophies — Kaizen in Japan, Lean in the United States — were born in large-scale industrial settings, but their principles can be translated into concrete practices for professional firms and smaller companies too [1][2].
This article covers the working definition, Kaizen and Lean tools adapted to everyday business, the rituals for involving people, the useful metrics and the most common mistakes during the start-up phase.
What continuous improvement is: the difference between changing and improving
How many times is the same procedure revised in a company after it was first written? In many smaller Italian companies, it rarely happens: once formalized, the procedure fossilizes — and that becomes the reason procedures don't get adopted.
The term "continuous improvement" is used interchangeably with "change," "innovation" and "optimization" — but the differences are concrete and bear directly on the business owner's operating choices. The definition introduced by Imai in 1986 [1] and the later Lean codification by Womack and Jones [2] help draw the boundaries, especially in a context like Italy's, where value added per employee in micro-enterprises stops at €36,400 against €56,600 for the average Italian company [5], and the OECD estimates Italian micro-enterprises to be about 30% less productive than their European peers [7].
Working definition. Continuous improvement is the collective habit of regularly reviewing — in small steps — processes, tools and work practices, with the goal of reducing waste, reducing errors and increasing effectiveness. It has no end date: it is a cycle, not a project.
Here are its boundaries with neighboring terms.
Continuous improvement vs organizational change. Change is episodic: it starts, it ends, it gets evaluated. Continuous improvement is habitual: it has no closing moment. Someone going through a company restructuring is managing change; someone who reviews one step of the delivery process every week is practicing continuous improvement. The cluster on organizational change management covers the former; this article covers the latter.
Continuous improvement vs innovation. Innovation introduces something new — a product, a technology, a business model. Continuous improvement refines something that already exists: the process already in use, the procedure already written, the ritual already practiced. The two approaches are not mutually exclusive, but they answer different questions.
Continuous improvement vs optimization. Optimization is a one-off technique: you optimize a process once, and you're done. Continuous improvement is a recurring philosophy: the same process is reviewed several times over time, each time starting from the result of the previous cycle.
Kaizen and Lean as specific methods. Kaizen and Lean are specific codifications of continuous improvement — not synonyms of the general concept. Kaizen [1] emphasizes small daily steps with everyone taking part; Lean [2] emphasizes eliminating waste along the value stream. "Continuous improvement" is the general concept that comes before both and can use either.
Start with a small step: the Kaizen operating principle for people with no time
Why should a company with 50 open operational problems start by solving just one? Because solving 50 in parallel doesn't solve them: it freezes them all.
The most common mistake when people discover Kaizen is trying to apply the entire framework in one go. Research by Bortolotti, Boscari and Danese on 317 plants in 9 countries [4] shows that the Lean implementations that succeed are those that start from "soft practices" (involving people, training, leadership) even before the technical tools. In a smaller company, this translates into a very concrete operating principle: the first improvement must be small enough to be done the same week.
The small-step micro-protocol. The path is made up of four simple phases, which can be applied in less than a week to a problem of limited scope.
-
Identify the small problem. Not a systemic problem ("customers aren't satisfied"), but a specific, measurable one ("the average response time to support emails is 48 hours against the 24 we promise"). Specificity is the criterion: if the problem isn't measurable, it isn't small enough yet.
-
Try a solution for one week. The experiment involves a single person or a single pilot process. You don't wait to find the perfect solution: you launch the first plausible solution with an explicit verification criterion.
-
Check the result. At the end of the week, you measure whether the problem has shrunk. If it has, you standardize the solution. If not, you analyze why and adjust the approach.
-
Standardize and move on to the next one. The solution that worked is documented (even a single line in an existing procedure), and you identify the next problem.
Examples for the three profiles.
For an independent professional with 1-2 team members: the routine for filing client documents (currently disorganized, everyone does it their own way) is standardized with a shared folder structure, tested for a week.
For a company with 7-15 employees: the handover ritual between shifts or between weeks (currently informal, based on memory) is structured with a half-page handover document, validated over three weeks.
For a company with 80-100 employees: the standard for the first day of onboarding (currently dependent on the availability of whoever welcomes the new hire) is formalized in a 10-point checklist, tested with the next three new hires.
The critical finding of the research [4] isn't which Lean tool works best, but that successful implementations start from people — not from tools. A "small step" that doesn't involve the person carrying it out isn't a Kaizen small step: it's a micro-change imposed from above.
The Kaizen and Lean tools that really work in a company (5S, PDCA, Kanban)
How many of the 7 wastes codified by Toyota are present, in a different form, in an Italian professional firm? All seven — only they're called "unanswered emails," "documents redone three times" and "meetings without decisions."
Kaizen and Lean were born in industrial settings — Toyota, from the 1950s to the 1970s — and it's natural to wonder whether the tools hold up in an eight-person accounting firm or a small machining shop. The answer, supported by classic Lean literature [2][3][6], is that the tools hold up as long as you understand which family of problems each one serves: physical and digital organization (5S), the experimentation cycle (PDCA), visual management of the workflow (Kanban).
5S — Organizing physical and digital space.
The 5S (Sort, Set in order, Shine, Standardize, Sustain) are a method for removing clutter from physical and digital workstations. The small-business adaptation applies directly:
- Sort: get rid of what isn't needed (obsolete files, unused tools, expired documents).
- Set in order: give every object/file/folder a fixed place, so that it can be found in less than 30 seconds.
- Shine: maintain order with a periodic review (5 minutes a day or 20 minutes a week).
- Standardize: codify the organization system so that anyone can maintain it without constant instructions.
- Sustain: make the system a habit, not an initiative.
When it makes sense in a smaller company: any time time is lost looking for documents, tools or information. It requires no purchases or software: it requires a clean-up session and an agreement on structure.
When it isn't worth it: in very small organizations (1-2 people), where the cost of standardization outweighs the benefit, or in environments that are highly chaotic by nature (e.g., construction sites, showrooms with seasonal products).
PDCA — The iterative improvement cycle.
PDCA (Plan-Do-Check-Act) is the systematic experimentation cycle on which all continuous improvement is based. It applies to an existing process: you plan a change (Plan), implement it on a small scale (Do), check the result (Check), and adopt or discard the change (Act). For company procedures, it is the engine of updates — as described in the guide to business procedures.
When it makes sense in a smaller company: on any process you want to improve without risking breaking what works. The ideal cycle length for a smaller company is 1 to 4 weeks.
When it isn't worth it: in emergencies, where there is no time for the planning and checking phases. In that case you act first, then apply PDCA after the fact to learn from the event.
Kanban — Visual management of the workflow.
Kanban is a board (physical or digital) that shows the progress of work: typical columns are "To do / In progress / Done." It eliminates the question "where are we at?" without requiring status meetings.
The 7 wastes (muda) identified by Ohno [6] — waiting, overproduction, unnecessary transportation, over-processing, excess inventory, motion, defects — also show up in service businesses, in different forms. Waiting shows up as emails that go unanswered. Overproduction as documents prepared before they're needed. Defects as documents redone three times. Kanban makes this waste visible before it becomes expensive. For the business process mapping that Kanban is applied to, the dedicated cluster is the starting point.
Involving people: the weekly rituals that keep continuous improvement alive
Why does the classic "employee suggestion" end up in a drawer in so many companies? Because there is no cycle that closes the loop — the suggestion goes out, but it never comes back as a decision.
The most relevant finding of the research on 317 plants [4] isn't which Lean tool works best, but the fact that successful implementations stand out for having practices that involve people — training, suggestion systems, hands-on leadership — even more than for the sophistication of their tools. Liker, describing principle 14 of the Toyota Way [3], puts it this way: "become a learning organization through relentless reflection and continuous improvement." In a smaller company, this translates into a small number of recurring rituals — weekly or monthly — that keep collective attention on improvement.
Ritual 1 — Weekly "what are we improving" meeting. Duration: 15 minutes. Frequency: end of the work week. Participants: the operational team or a subgroup of 3-5 people. Expected output: a single thing to try the following week, with an owner and a check date.
The structure is simple: what went well this week? What didn't? What will we try to change? The format never changes, because predictability is the condition for habit.
Ritual 2 — A suggestion system with a response cycle. Tool: a shared document or a physical board. Rule: every suggestion gets a response within 7 working days — even "we got it, we're thinking about it." The closed loop is the difference between a suggestion system that works and one that breeds cynicism. Without responses, nobody submits suggestions anymore after the second month.
Ritual 3 — Monthly retrospective on a specific process. Duration: 30-45 minutes. Frequency: once a month. Subject: a specific process chosen in advance (not "how are things going in general"). Expected output: 1-2 PDCA micro-experiments to launch the following month.
Ritual 4 — The "stop and fix" standard. When a problem happens for the second time, you stop and study it together before going back to work. Ohno [6] describes this principle as "stop and fix": prefer a temporary slowdown to a chronic problem. In a smaller company, it translates into an explicit rule: "if the problem comes back, we raise it right away — not next week."
The language of the rituals must be simple: avoid "reporting ritual," "quality circle," "huddle standup." Simple terminology lowers the barrier to adoption. On effective business meetings — of which improvement rituals are a subset — the dedicated cluster offers the operating structures.
Measuring continuous improvement without turning the company into a dashboard
What good is a dashboard of 47 indicators that nobody looks at? None — and it's the first thing continuous improvement should eliminate, even before adding new metrics.
Without measurement, continuous improvement turns into a subjective impression — "I think things are going better." With too much measurement, it becomes bureaucracy that smothers the very initiative it should support. The Lean principle of "perfection as a direction" [2] suggests a balance point: measure little, measure well, measure over time. The useful metrics aren't new — they build on the KPI system already in place — but they need to be selected with a criterion specific to continuous improvement: they measure the trend, not the absolute value.
Four families of metrics.
Flow metrics: lead time (time from receiving the request to delivering the output), throughput time (actual processing time), work-in-progress (number of activities open at the same time). The trend to watch: are they shrinking over time?
Quality metrics: rework rate (how many activities are redone after the first cycle), errors detected per period, customer complaints by recurring type. The trend to watch: is the frequency of the same errors going down?
Participation metrics: number of suggestions received in the suggestion system, percentage of suggestions closed within 30 days, number of PDCA micro-experiments launched per quarter. These metrics measure the health of the involvement system, not just operational performance.
Impact metrics: hours saved per improved process (estimated before and after each PDCA experiment), increase in productivity per employee in the areas where improvement has been most systematic.
The guiding principle. Every metric must have: a trend visible over time (monthly data at a minimum), an owner who reads it, a review cadence. A metric without an owner doesn't get read. A metric that gets read but offers no possibility of action gets abandoned.
For the overall KPI system of which continuous improvement metrics are a subset, the cluster on business KPIs offers the complete picture. For more structured performance measurement, the reference is the cluster on business performance measurement.
Common mistakes when starting continuous improvement (and how to spot them in time)
Why isn't a company that adopts the Kaizen vocabulary automatically adopting Kaizen? Because vocabulary costs nothing, practices cost time — and it's time, not vocabulary, that makes the difference.
Lean implementations rarely fail because of the wrong choice of tools. They fail — according to the data on 317 plants [4] — for recurring reasons: lack of leadership support, insufficient training, using the tools as slogans. In a smaller company, these mistakes take typical forms, recognizable in the first three to six months of the journey. Spotting them in time is the difference between correcting course and going back to "doing things the old way."
Mistake 1 — Starting without leadership support. Sign: the project stays confined to a functional team or an operations manager, with no involvement from senior management. Correct version: continuous improvement is a leadership choice, not something delegated downward. The first PDCA cycle must be visible to the company's leadership.
Mistake 2 — Imposing the tools from above as slogans. Sign: team members use the vocabulary (Kaizen, Lean, 5S) but the practices are superficial or simulated. Correct version: every tool is introduced with a hands-on training session (30-60 minutes) and a first guided experiment done together, not just announced by email.
Mistake 3 — Skipping the training phase. Sign: systematic errors in applying PDCA (Check and Act get skipped), 5S that stops at the first S (Sort), Kanban that gets abandoned after two weeks. Correct version: before introducing a tool, devote at least one session to "how it works," with an example applied to your specific context.
Mistake 4 — Measuring too much and measuring too soon. Sign: the time spent measuring exceeds the time spent improving; people experience the measurement system as control rather than as a tool. Correct version: in the first three months, measure a single metric per improvement project. Add metrics only once the review system has stabilized.
Mistake 5 — Confusing continuous improvement with a total reorganization. Sign: the project is suspended after 4-6 months because "we're changing too many things" or because "we don't have the expected results." Correct version: continuous improvement doesn't produce visible results in 6 months; it produces habits and small cumulative results. Anyone expecting a "transformation" started from the wrong premise.
For managing the deeper, episodic interventions that continuous improvement doesn't cover, the cluster on organizational change management is the right reference.
Limits and conditions of applicability
Industrial vs service settings. Kaizen and Lean tools were born in manufacturing settings and have been tested systematically on mostly industrial samples [4]. Service businesses can apply their principles, but adapting them requires translating the wastes (the 7 muda [6]) into cognitive and relational activities — an operation that isn't automatic and may require an initial period of experimentation.
Minimum size. In very small organizations (1-3 people), the cost of improvement rituals may outweigh the benefit. The practical rule: rituals make sense when there is at least a second person taking part, because continuous improvement is a collective process, not an individual one.
Sources [1], [2], [3], [6]: management books. The references cited are recognized management texts, not peer-reviewed studies. The claims about the benefits of Kaizen and Lean are supported by [4] (peer-reviewed), which documents their conditions for success on a large sample. The operational guidance comes from combining the two types of source.
ISTAT [5] and OECD [7] data. Value added per employee is a measure of apparent productivity for 2022 and varies greatly from sector to sector; the OECD's European comparison covers only micro-enterprises with up to 9 employees, not the whole range of small and medium-sized companies. Both describe the Italian production system as a whole, not the individual company.
FAQ — Frequently asked questions about continuous improvement
Are Kaizen and Lean the same thing? No. Kaizen [1] is a principle (the daily habit of small improvements), Lean [2] is a management system that includes specific tools for eliminating waste. Kaizen can be practiced without Lean; Lean incorporates Kaizen as its underlying philosophy. In a smaller company, it is often more useful to start with Kaizen principles (small steps, participation) before introducing the more structured Lean tools.
Where do you concretely start? With the smallest, most visible problem. Not with a reorganization project. In the first week, you identify a single recurring activity that is carried out inefficiently, try a minimal change, and check the result.
How long does it take before you see results? Micro-results are visible within 4-8 weeks if the journey is structured with documented PDCA experiments. Cumulative results on the company's overall efficiency become measurable after 6-12 months of consistent practice. Anyone expecting visible results in 2-3 weeks is confusing continuous improvement with one-off optimization.
Can smaller companies apply Lean? The research [4] — conducted on a sample that includes Italian plants — shows that success depends mainly on "soft practices" (involvement, training, leadership), not on sector or size characteristics. Smaller companies can apply it with the same constraints as any other company: the adoption curve is similar.
Operational summary
Continuous improvement is a collective habit built with small PDCA experiments, simple tools (5S for organization, Kanban for flow, PDCA for experimentation), weekly rituals that keep attention alive, and a few trend metrics. Its value isn't measured by the intensity of the interventions, but by their regularity over time.
The two factors that most clearly separate implementations that succeed from those that get abandoned are the involvement of the people who do the work (before the tools) and leadership that makes improvement visible (before the experiments). Without these two elements, the tools remain slogans.
Conclusion
Continuous improvement is not a project with an end date: it is a collective habit built with small daily steps, simple operational tools (5S, PDCA, Kanban), recurring rituals that keep attention alive, and a few trend metrics. The difference from organizational change is clear: change starts and ends, while continuous improvement requires ongoing maintenance. The difference from innovation is just as clear: innovation introduces something new, improvement refines something that already exists.
To learn more about how to structure the procedures continuous improvement works on, the guide to business procedures is worth a look. If you are facing a single transformation effort rather than a systemic habit, the reference is the cluster on organizational change management. If you haven't mapped your processes yet, process mapping is the natural starting point: without a map, PDCA spins its wheels.
The lessons that come out of a job just completed become maintenance only when they land in a procedure, an acceptance criterion or a start-up check: the path is described in the page on lessons learned.
Picture a company where every week, at the same time, the people who do the work stop for fifteen minutes to ask themselves what to improve — not in the abstract, but on a specific process. Picture that same company twelve months later: dozens of procedures revised, waste removed, people proposing solutions instead of raising problems. That is where continuous improvement leads. The gap the OECD measures between Italian micro-enterprises and their European peers — about 30% lower productivity [7] — isn't closed by an announced transformation: if anything, it narrows through the accumulation of small, repeated improvements.
Sources and references
[1] Imai, M., "Kaizen: The Key to Japan's Competitive Success", McGraw-Hill, 1986 (2nd ed. 2012).
[2] Womack, J. P. and Jones, D. T., "Lean Thinking: Banish Waste and Create Wealth in Your Corporation", Free Press, 1996 (2nd ed. 2003).
[3] Liker, J. K., "The Toyota Way: 14 Management Principles from the World's Greatest Manufacturer", McGraw-Hill, 2004.
[4] Bortolotti, T., Boscari, S. and Danese, P., "Successful lean implementation: Organizational culture and soft lean practices", International Journal of Production Economics, 160, 182–201, 2015.
[5] ISTAT, "Annuario statistico italiano 2025 — Capitolo 14: Imprese", ISTAT, Rome, 2025 (2022 data). Available at: https://www.istat.it/storage/ASI/2025/capitoli/C14.pdf
[6] Ohno, T., "Toyota Production System: Beyond Large-Scale Production", Productivity Press, 1988.
[7] OECD, "OECD Economic Surveys: Italy 2024", OECD Publishing, Paris, January 2024. Available at: https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/01/oecd-economic-surveys-italy-2024_18011b9d/78add673-en.pdf