Organisation design: spans, layers, and central or local
How companies arrange their people, how to count managers, and what a flatter or more central setup saves and risks.
Key takeaways
- An organisation design is who reports to whom and where decisions are made.
- The strong answer separates work that is the same everywhere from work that depends on the customer, and quantifies the trade-off.
- Common traps: Counting every removed manager role as a saving, when some people move into other jobs.
Key idea
An organisation design is who reports to whom and where decisions are made. Two numbers describe most of it: the span of control (how many people each manager leads) and the number of layers (how many levels sit between the front line and the top). Wider spans mean fewer managers and lower cost, but less time per person.
Cases about cost cutting, mergers and slow decisions often turn into organisation questions. "Why are we slower than our rivals?" can mean too many layers: every decision climbs up and down a long chain. "Why are our overheads high?" can mean narrow spans: many managers each looking after only a few people. The front line (staff who serve customers or make the product) is usually fixed by the work. The managers above them are where design choices show up in cost.
What sets the right span
- How similar the work is. Twelve claims handlers doing the same task can share one team leader; six specialists doing different, complex work often need a narrower span.
- How experienced the team is. New staff need more coaching, so spans start narrower.
- How much the manager also does hands-on work. A player-coach who sells as well as leads can manage fewer people.
- How good the tools are. Clear dashboards and standard processes let one manager oversee more people.
| Level | Span today | Roles today | Span proposed | Roles proposed |
|---|---|---|---|---|
| Claims handlers (front line) | not a manager | 960 | not a manager | 960 |
| Team leaders | 8 | 120 | 12 | 80 |
| Claims managers | 5 | 24 | 8 | 10 |
| Heads of claims | 4 | 6 | layer removed | 0 |
| Director | 6 | 1 | 10 | 1 |
| All manager roles | total | 151 | total | 91 |
So-what
Widening two spans and removing one layer cuts manager roles from 151 to 91 while the 960 front-line staff stay the same.
Worked case
What a flatter claims unit saves
The prompt
Using the table above: the fictional insurer pays each manager about EUR 90,000 a year and each claims handler about EUR 50,000, both including social charges (illustrative numbers). How many manager roles does the new design remove, what does it save each year, and what share of the unit's people cost is that?
The structure
- Saving = manager roles removed x cost per manager
- Roles today: add each management layer (front line divided by span, layer by layer)
- Roles proposed: same method with the new spans
- Key: Saving as a share of total people cost
Working it through
1. Team leaders today
960 handlers, 8 per team leader.
Team leaders today:960 ÷ 8 = 1202. Managers today
120 team leaders, 5 per manager.
Claims managers today:120 ÷ 5 = 243. All manager roles today
120 team leaders, 24 managers, 6 heads and 1 director.
Manager roles today:120 + 24 + 6 + 1 = 1514. All manager roles proposed
960 divided by 12 is 80 team leaders, 80 divided by 8 is 10 managers, the heads layer goes, and the director leads the 10 managers.
Manager roles proposed:960 ÷ 12 + 960 ÷ 12 ÷ 8 + 1 = 915. Roles removed
151 minus 91.
Manager roles removed:151 - 91 = 606. Yearly saving
60 roles at EUR 90,000.
Yearly saving (EUR million):60 × 90,000 ÷ 1,000,000 = 5.47. Share of people cost
Handlers cost 960 x 50,000 = EUR 48 million; managers today cost 151 x 90,000 = EUR 13.59 million.
Saving as a share of people cost (percent):5.4 ÷ (960 × 0.05 + 151 × 0.09) × 100 = 8.778. Realistic saving
Not every removed role leaves the payroll. If half of the 60 people move into specialist roles (for example complex claims experts) at similar pay, only 30 roles of cost go.
Realistic yearly saving (EUR million):60 × 0.5 × 90,000 ÷ 1,000,000 = 2.7
The recommendation
The flatter design removes 60 manager roles and one layer, so decisions travel through 3 manager levels instead of 4. The headline saving is EUR 5.4 million a year, close to 9 percent of the unit's people cost; if half the people move into specialist roles, the realistic saving is about EUR 2.7 million. Before recommending it, check that a team leader can really coach 12 handlers: if claims are complex or staff are new, quality and customer complaints may suffer. Phase it in team by team and watch claim errors and handling time.
Risks: Quality falls if team leaders can no longer review complex claims; Good people leave if the change feels like a cut rather than a better job; The saving is lower if severance (payments to staff who leave) is large in the first year.
Next steps: Compare error rates of teams that already run at a span of 10 or more; Plan which roles become specialist experts instead of managers.
Central or local
The second big design choice is where work and decisions sit. Central means one team serves the whole company, for example one finance hub, one buying team or one IT platform. Local means each country, region or business unit runs its own. Central gains economies of scale (a lower cost per unit as volume grows), one standard way of working, and stronger buying power. Local gains speed, knowledge of local customers and rules, and a sense of ownership. Many companies mix the two: central for back-office work that is the same everywhere (payroll, accounting, purchasing of common items), local for work that faces the customer (sales, pricing, marketing).
A consumer goods company in five Southeast Asian markets asks: should we centralise marketing?
Weaker answer
Yes, centralise everything, because central is always cheaper.
Stronger answer
Split marketing into its parts. Media buying and brand guidelines are the same everywhere, so a central team can buy at scale and keep one standard. Campaign ideas, language and promotions depend on each market, so they stay local. Then put a number on the saving from central media buying and test the risk to local sales.
Why the stronger answer wins: The strong answer separates work that is the same everywhere from work that depends on the customer, and quantifies the trade-off. The weak answer ignores the revenue side.
A fictional company runs a finance team of 12 people in each of 5 Southeast Asian countries, at an average cost of USD 40,000 a person a year. A shared finance hub in Kuala Lumpur would need 36 people at USD 35,000 each, plus USD 240,000 a year for systems and travel. What is the yearly saving, in USD million?
Counting every removed manager role as a saving, when some people move into other jobs. Counting only salaries: a manager also costs office space, systems and travel. Assuming one span fits every team. Forgetting one-off costs such as severance and moving work, which can eat most of the first year's saving. Treating "central" as all or nothing, when most companies centralise some tasks and keep others local.
A unit has 600 front-line staff and team leaders with a span of 6. How many team leaders does it have?
Which task is usually the best candidate to centralise?
Sources for this lesson (1)
- Recognized public explanations of case-interview concepts and terms
My notes on this lesson
0 of 5,000 characters. Saves automatically.
Try the 3 remaining checks and drills above to complete this lesson (0 of 3 done).
Spotted something wrong or out of date? Report a mistake. We check every report and correct the page.