The hidden cost of manual processes
Every manual process that survives in your organisation is a silent source of inefficiency. We're not just talking about the time an employee spends entering data from one spreadsheet to another: we're talking about errors that propagate, information that arrives too late for decision-making, and qualified talent trapped in repetitive tasks that generate no value.
In our 25+ years of process consulting experience, we have found that companies systematically underestimate the real cost of their manual processes. When we carry out a detailed analysis, the numbers are revealing: a mid-sized company with 200 employees may be dedicating between 3,000 and 5,000 annual hours to tasks that could be automated, equivalent to 2 to 3 full-time positions dedicated exclusively to moving information from one place to another.
But the direct economic cost is just the tip of the iceberg. Manual processes generate variability (each person executes the process slightly differently), opacity (nobody has complete visibility of a process in progress) and fragility (the absence of a key person can paralyse an entire operation). Recognising these hidden costs is the essential first step towards transformation.
In our 25+ years of process consulting experience, we have found that companies systematically underestimate the real cost of their manual processes. When we carry out a detailed analysis, the numbers are revealing: a mid-sized company with 200 employees may be dedicating between 3,000 and 5,000 annual hours to tasks that could be automated, equivalent to 2 to 3 full-time positions dedicated exclusively to moving information from one place to another.
But the direct economic cost is just the tip of the iceberg. Manual processes generate variability (each person executes the process slightly differently), opacity (nobody has complete visibility of a process in progress) and fragility (the absence of a key person can paralyse an entire operation). Recognising these hidden costs is the essential first step towards transformation.
Identifying bottlenecks: where it really hurts
The most common mistake when approaching process automation is trying to automate everything at once. At SEITON we apply a clear principle: don't automate a bad process; first improve it, then automate it. And to improve it, you need to precisely identify where the real bottlenecks are.
Our diagnostic methodology is based on four warning signals that indicate a process candidate for transformation:
We recommend starting with a pain point identification workshop with area managers, where the 10 most critical processes are mapped and scored against these four dimensions.
Our diagnostic methodology is based on four warning signals that indicate a process candidate for transformation:
- Excessive cycle time: the process takes significantly longer than it should. For example, a budget approval that takes 5 days when the actual analysis takes 2 hours.
- High error or rework rate: if more than 10% of process executions require manual corrections, there is a structural problem.
- Dependency on specific people: if a process stops when someone is on holiday or sick leave, it is a fragile and undocumented process.
- Volume of exceptions: when exceptions to the standard process exceed 25% of cases, the process no longer reflects operational reality.
We recommend starting with a pain point identification workshop with area managers, where the 10 most critical processes are mapped and scored against these four dimensions.
Process mapping: the foundation of all transformation
You cannot improve what you don't understand. That is why process mapping is the cornerstone of every transformation project we undertake at SEITON.
Our mapping methodology follows a three-layer approach:
Layer 1 - Current process (AS-IS): we document how the process is actually executed, not how the quality manual says it should be. This involves direct observation, interviews with executors and measurement of real times. It is common to discover that the real process differs enormously from the documented one.
Layer 2 - Value analysis: we classify each process step into three categories: activities that add value for the customer, necessary support activities (don't add value but are essential) and pure waste activities (waits, redundant information transfers, unnecessary approvals). In a typical process, between 40% and 60% of activities fall into the third category.
Layer 3 - Target process (TO-BE): we design the optimised process by eliminating waste, simplifying flows and identifying the points where automation delivers the greatest return. This design includes specific KPIs that will allow improvement to be measured objectively.
Our mapping methodology follows a three-layer approach:
Layer 1 - Current process (AS-IS): we document how the process is actually executed, not how the quality manual says it should be. This involves direct observation, interviews with executors and measurement of real times. It is common to discover that the real process differs enormously from the documented one.
Layer 2 - Value analysis: we classify each process step into three categories: activities that add value for the customer, necessary support activities (don't add value but are essential) and pure waste activities (waits, redundant information transfers, unnecessary approvals). In a typical process, between 40% and 60% of activities fall into the third category.
Layer 3 - Target process (TO-BE): we design the optimised process by eliminating waste, simplifying flows and identifying the points where automation delivers the greatest return. This design includes specific KPIs that will allow improvement to be measured objectively.
Automation criteria: what to automate and what not
Once processes are mapped, the key question is: what deserves to be automated? Not every manual process should become an automated one. Automation has implementation, maintenance and evolution costs, and only makes sense when the return justifies the investment.
At SEITON we evaluate each candidate process with a five-criteria matrix:
Processes that score high in at least three of these five criteria move to the technology selection phase. Those that don't are optimised through flow redesign and standardisation.
At SEITON we evaluate each candidate process with a five-criteria matrix:
- Volume: is it executed frequently enough to justify the investment?
- Standardisation: does it follow clear and repeatable rules? Processes with high variability are poor candidates for full automation.
- Error propensity: do manual errors generate significant economic, legal or reputational impact?
- Systems integration: does the process involve moving data between two or more systems? These transfers are ideal candidates.
- Customer impact: do slowness or errors directly affect the end customer experience?
Processes that score high in at least three of these five criteria move to the technology selection phase. Those that don't are optimised through flow redesign and standardisation.
Technology selection and implementation
Technology is an enabler, not an end in itself. At SEITON we follow a minimum necessary technology principle: the simplest solution that solves the problem is always the best.
The range of options is structured in increasing levels of complexity:
Level 1 - Basic automation: standardised templates, digital forms, approval workflows in collaborative tools (Microsoft 365, Google Workspace). Low cost, implementation in days.
Level 2 - Automated workflows: low-code/no-code platforms like Power Automate, Make or n8n that connect systems and automate flows without complex programming. Moderate cost, implementation in weeks.
Level 3 - BPM and advanced automation: Business Process Management platforms for complex processes with multiple participants, sophisticated business rules and audit requirements.
Level 4 - Intelligent automation: combination of automation with artificial intelligence capabilities to manage exceptions, classify documents or make decisions based on historical patterns.
Implementation should follow a phased roadmap, starting with a pilot on a contained process that allows the technology to be validated, results measured and confidence built before scaling.
The range of options is structured in increasing levels of complexity:
Level 1 - Basic automation: standardised templates, digital forms, approval workflows in collaborative tools (Microsoft 365, Google Workspace). Low cost, implementation in days.
Level 2 - Automated workflows: low-code/no-code platforms like Power Automate, Make or n8n that connect systems and automate flows without complex programming. Moderate cost, implementation in weeks.
Level 3 - BPM and advanced automation: Business Process Management platforms for complex processes with multiple participants, sophisticated business rules and audit requirements.
Level 4 - Intelligent automation: combination of automation with artificial intelligence capabilities to manage exceptions, classify documents or make decisions based on historical patterns.
Implementation should follow a phased roadmap, starting with a pilot on a contained process that allows the technology to be validated, results measured and confidence built before scaling.
Change management and results measurement
The number one reason automation projects fail is not technological: it's human. People need to understand why things are changing, how it affects them and what is expected of them.
Our approach is structured around four principles: involve from day one (the teams that execute the processes participate in diagnosis and design), communicate the why (automation eliminates low-value tasks, not jobs), train practically and progressively, and celebrate early results.
Every project must demonstrate its value with objective data. Before implementation, we define baseline KPIs:
To illustrate: an industrial engineering company with a technical quotation process that took 4 days on average with a 22% error rate, after mapping, redesigning and automating, reduced the time to 6 hours, errors to 3% and tripled the proposals issued per month. ROI was achieved in less than 4 months.
Our approach is structured around four principles: involve from day one (the teams that execute the processes participate in diagnosis and design), communicate the why (automation eliminates low-value tasks, not jobs), train practically and progressively, and celebrate early results.
Every project must demonstrate its value with objective data. Before implementation, we define baseline KPIs:
- Cycle time: total process duration
- Error rate: percentage of executions requiring correction
- Cost per transaction: total cost divided by executions
- Freed capacity: hours redirected to higher-value activities
To illustrate: an industrial engineering company with a technical quotation process that took 4 days on average with a 22% error rate, after mapping, redesigning and automating, reduced the time to 6 hours, errors to 3% and tripled the proposals issued per month. ROI was achieved in less than 4 months.
