(12) Systematic Management: Specialisation, Centralisation and Data
- Jon Smart
- May 29
- 9 min read
This is the 12th blog post in the Organising for Outcomes series. It is helpful to understand where we’ve come from, how today’s ways of working have evolved, and the context that those ways of working evolved in. This helps us to understand why we’re working the way we’re working and what we might want to change in today’s context, which is significantly different compared to previous technology-led revolutions.
In the previous post we took a closer look at the 3rd of three key systems that were developed as part of Systematic Management: Wage Systems. Many of the management innovations introduced in the time period from 1870 to 1900 are still in use today, sub-optimally, in a very different context.
In this post we'll take a deeper dive into three broad themes of Systematic Management, introduced between 1870 to 1900: Specialisation, Centralisation and Data.
Specialisation
Systematic Management was concerned with bringing order to the growing scale and complexity of industry in the 3rd Industrial Revolution. With scale and complexity no longer fitting in people’s heads, there was a need for new behaviours and mechanisms. It is no coincidence that the early 1900s saw a significant rise of the use of the word ‘organisation’ to refer to firms, as growth necessitated bringing order to chaos.
Increasingly narrow and repetitive roles
As the steel industry grew, fuelled by demand from the previous technology revolution, there was a significant increase in specialisation and functional silos (many with their own role-based union). There were narrower roles and there were more roles, with a goal of ‘more for less’ in the context of manual, repetitive, knowable, physical work.
Specialisation is of course not new. As per blog post 2, as societies went from being simple to complex, such as Mesopotamia around 4,000 BCE, we see the beginning of division of labour. And, as we looked at in blog post 3, the 1st Industrial Revolution took it to another level, with even more narrow specialisation, such that some roles were done by children. It was also at a scale not seen before, with the introduction of the factory system resulting in 1,000 people working in one building, as cogs in a machine, a move away from the craftsmanship of the past. These learnings were now being applied to Steel and Heavy Engineering.
And so, with the diffusion of learning from textile mills into the ‘manufactories’, there was a significant increase in the number of roles with narrower responsibilities. For example:
Superintendent
Assistant superintendent
Planner
Estimating Clerk
Rate Fixer
Draughtsman
Cost accountant
Purchasing agent
Job Order Clerk
Bill of Materials Clerk
Supply clerk
Route clerk
Instruction card clerk
Industrial Engineer
Timekeeper
Storekeeper
Office manager
Ledger Clerk
Foreman
Gang boss
Disciplinarian
Speed boss
Repair boss
Inspector
Machinist
Toolmaker
Mechanic
Labourer
As a result of this, the role of the foreman in particular, was significantly reduced. Before Systematic Management the foreman was at the centre of shop floor operations, with many responsibilities including supervising, planning, training, disciplining, timekeeping, inspecting, hiring, firing, managing, checking inventory and ordering materials.
Post Systematic Management almost all of these responsibilities had been moved into separate roles. As Taylor wrote in Shop Management (1903):
“All possible brain work should be removed from the shop and centered in the planning or laying-out department, leaving for the foremen and gang bosses [...] to see that the operations planned and directed from the planning room are promptly carried out in the shop.
The shop, and indeed the whole works, should be managed, not by the manager, superintendent, or foreman, but by the planning department.”
The foreman of old was the boss of the shop floor, with a high degree of autonomy and empowerment. However everything was done by the seat of the pants and with a lot of reliance on the foreman’s knowledge and relationships, which does not scale.
The foreman of 1890 onwards was an enforcer with less autonomy, making sure that the prescriptive and detailed per-worker job orders from the planning department were carried out.
There was a separation of:
planning and execution
thinking and doing
The behavioural norm was order-giver, order-taker, with work passing between roles. Brain work was removed from the shop, you did what you were told to do.
Foremen and clerks could find themselves focussing primarily on accurately operating the system rather than focussing on value and the overall company goals. People developed a more limited view of their job in the firm.
These behaviours, this way of working, is still alive and kicking in organisations today, in the Age of Digital & AI. The behavioural norm is order-giver and order-taker with work thrown over walls between role-based silos, with few people having a view of the consumer and value.
For example, once a six month long annual planning process has completed, ‘the business’ pass an order to Product, who do some work and pass orders to Engineering, who do some work and pass orders to QA, who do some work and pass orders to IT Ops, who do a thing and everyone hopes that it adds value, but no one actually goes back to check. Just so long as the role-based milestones are met, which reflect predetermined output decided at the point of having learnt the least. When there are issues, the incentive and hence the behaviour is ‘it’s not my problem, the hole is on your side of the boat’.
Centralisation
With increased scale, came a need for increased coordination. As organisations expanded it became less feasible to manage by knowledge and relationships alone, with complexity exceeding cognitive capacity. The growth in scale came from both existing companies growing as markets opened up, in addition to federations, mergers and acquisitions. Increasingly in the last twenty years of the 19th century there were amalgamation of loosely coupled smaller companies, with duplication of activities.
In addition, with workers doing ever smaller parts of the value chain in more narrow and repetitive roles there were even fewer people who had a view over the end-to-end flow of value.
In response to these challenges we see the growth of centralised ‘staff’ functions, as part of a ‘line and staff’ organisational structure, pioneered in industry by the railroad companies in the previous Industrial Revolution, due to the long distance of the railroads. In this 3rd Industrial Revolution the motivation for centralised staff functions was for coordination, standardisation, efficiency and more-for-less. The language used at the time was improved administration.
Each major function, such as marketing, procurement, design, planning, production, sales, accounting and so on came to be managed by a separate department. No one department had priority over the others, as did a railroad’s transportation department. Coordination of departmental activities was harder in an industrial enterprise than on a railroad and the coordination of many centralised staff functions with disparate activities proved to be non-trivial, just as coordinating the work had been previously.
This in turn led to the common adoption of an Executive Committee, a term which is ubiquitous today. The Executive Committee (ExCo) formed as a group of officers delegated with day-to-day decision making authority on behalf of the Board of Directors, with oversight across the centralised staff functions and the production lines. The ExCo became a standard feature of large firms in this time period, as part of the move to formalise and centralise authority.
By the 1920s most large industrial organisations had a centralised, functionally departmentalised structure.
These centralised staff functions have evolved over time to become the Corporate Headquarters, or Group Centre Functions which we recognise today. We start to see the beginnings of today's organisations. For example:
End of 19th Century ‘staff’ functions | Today’s traditional organisations |
Accounts Office | Corporate Finance Department Treasury Tax |
Auditor | Internal Audit |
Planning Department | Project Management Office |
Drawing Office (Engineering) | Enterprise Architecture |
Legal Counsel | Legal Department |
Personnel Department | HR Department |
Inventory Office | Supply Chain Inventory Management |
Purchasing Office | Procurement Department |
Sales Office | Sales Department Distribution Department |
Marketing Office | Marketing Department |
Real Estate | Real Estate Department |
Factory Owners Factory Manager President Vice President General Superintendent General Manager Central Records Office Typing Pool | Strategy Department Risk Department Compliance Corporate Communication Investor Relations |
Usually to be fully at one end of a spectrum of centralised to decentralised is sub-optimal, at least to stay there for a protracted period.
For example, fully decentralised can be advantageous in that there is an ability to be nimble, to grow fast, but it comes at the cost of a lack of coordination which can be detrimental to the larger organisation, with duplication, reinventing the wheel, divisions competing against each other, not being aware what others in the same organisation are doing, a lack of a consolidated view of the customer, higher costs, longer time to value and a fragmented customer experience.
To have fully centralised functions can help to address these downsides, with a view across the org, ensuring that trade-offs are balanced for the greater good, reducing duplication, ensuring more reuse, having better data, and a better customer experience.
However, there was a significant weakness as the pendulum swung to fully centralised functions, which is that very few people were entrusted with a greater number of complex decisions. As is often still the case today, decisions were bubbled up to a few in authority who were furthest away from the shop floor, the customer or the information. A handful of people would be making all the decisions.
Superiors made the decisions, while the subordinates did the doing, with the separation of thinking and doing.
Rather than fully centralised or fully decentralised, there is an optimal middle point (depending on context), which aims for the best of both, a federated approach. Overall principles, standards, sharing of learning and innovation, minimising expensive duplication, and providing support come from the centre, whilst Business Unit (BU) aligned satellite functions allow for optimising to the unique context of each BU, feeding innovation and learning back to the centre, to be shared. More on this later.
Data
A system is a set of interconnected parts that work together as a coherent whole to achieve a common purpose.
Data enables a system to be a system. In the absence of data it would not be possible for component parts to work together as a coherent whole and it would not be possible for the whole or the parts to achieve a purpose, due to the absence of a feedback loop. Data enables learning, action and improvement, in order to optimise for desired outcomes.
The larger the scale, the more complex the environment is, the more important it is to have data feedback loops to ‘see’ what is happening in the system of systems. For example cost accounting, production control and wage systems would not be feasible in the absence of data.
Therefore, a fundamental part of Systematic Management was the gathering of data, to turn it into information, to enable improvement. Information mostly moved upwards, in daily, weekly and monthly reports, enabling senior management to review plant performance and take action. Information also enabled comparison between plants and subdivisions to make it easier to identify and remediate inefficiencies. Starting in the 1850s in the railroads, and amplified in the last two decades of the 19th century, data enabled continuous improvement in how we worked.
Data also enabled a better horizontal flow of information, triggering what work to do, by who and by when. A customer order triggered production, a job order triggered individual work. It also enabled senior management to hold middle management more accountable. Management could spend less time directing work and more time optimising for outcomes.
Data enabled the following:
Transparency on the firm’s operations to inform vertically and horizontally enabling the system of systems to optimise for the whole
Cost efficiency: Control of expenses enabling more accurate pricing
System-of-work efficiency in the coordination of production across departments. For example, inventory available in the right place, at the right time, in the right quantity.
Worker efficiency: Motivation for worker effort
Comparison: Measures of overall organisational efficiency enabling the comparing of plants and competitiveness
Improvement: Information for continuous improvement
Systematic Management’s increased focus on data to optimise the whole was a precursor to and enabler of Scientific Management that followed.
Why we work the way we do today
We can see that ways of working from the 1880s still heavily influence how we work today. Many large organisations today still have heavily centralised functions, as per the table above, rather than federated (a fractal pattern which scales and allows for optimising for outcomes in local context, whilst having coherence for the whole).
The majority of organisations today also still have work passing by role based silo, with incentives in role based silos, and a lack of shared goals, end-to-end.
And, a lesson to re-learn from the past is the importance of a data feedback loop on how we work. For example, measures of quality, value, time to value, safety (controls, Minimal Viable Compliance, agile not fragile) and happiness (engagement and advocacy). Using that data feedback loop for improvement in how we generate value is also a lesson that needs to be re-learnt today at many organisations.
In most organisations today, people are busy being busy with little to no data on the system of work, with no incentive to continuously improve and with little attention paid to ‘how we do, what we do’.
Learnings for today:
(1) Multidisciplinary teams over work passing by role silos: moving up and down the pitch together, with a shared goal or hypothesis or problem to solve, rather than passing the baton in a relay race. Empowered teams over autocracy. In the context of unique change, it reduces time to value, increases flow efficiency, results in higher quality, with lower risk and more engaged colleagues & happier consumers.
(2) Federated over fully centralised or decentralised. It is infinitely scaleable, is fractal and enables coherence (also see Viable Systems Model, Stafford Beer)
(3) Data feedback loops used for continuous improvement: there are data feedback loops on the How and the data is used to continuously improve.
______________________________________
Also see:
______________________________________
Learning resources:
If you found this article useful, you might be interested in additional Sooner Safer Happier learning resources to enable you to lead with these behaviours:
If you want to explore your learning journey book a call with our team!
