Find the Bottleneck in Bank Operations
Banks pulling every lever gain 30% or more in efficiency. Bottleneck analysis shows which single step limits output and where to act first.
Alexander Kiel
Business Analyst · Banking & Wealth Management
22.08.2026 · 8 MIN READ · PROCESS OPTIMISATION
ON THIS PAGE
1. Map the Flow End to End
2. Measure Where Work Waits
3. Run a Bottleneck Analysis on Data
4. Exploit the Constraint First
5. Align Other Steps to the Bottleneck
6. Plan Capacity for Peaks and Growth
Sample case: N26
In short
FAQ
Every operations process has one step that sets its pace. Add staff anywhere else and output stays the same, while the queue in front of that step keeps growing. Bottleneck analysis finds that step and directs effort to it. McKinsey reports that banks pulling every lever achieve efficiency gains of 30 percent or more and that more than half of the transformation potential comes from the top 10 to 20 end-to-end processes.
Time is getting shorter as well. US securities moved from two-day to one-day (T+1) settlement on 28 May 2024, leaving post-trade teams a single business day to confirm, allocate and fund each trade. A step that used to absorb a backlog overnight now delays settlement.
Yet many banks still respond to backlogs by adding people to every team or by automating whichever task is easiest. Do you know which single step in your largest process decides how much work leaves it each day?
STEP 1
Map the Flow End to End
Start with one process, from the trigger to the outcome the client sees: an account opening, a payment investigation or a corporate loan drawdown. Draw every step, handoff, system and approval. Record who does the work and where cases wait between steps. A bottleneck stays invisible when each team only sees its own part.
Choose the process with care. At one large UK bank, McKinsey mapped about 50 end-to-end processes, of which 15 processes made up roughly 80 percent of the overall cost base. Start where volume and cost are highest. My article on as-is process mapping shows how to capture each step and handoff before you change anything.
✓
Pick one process from client trigger to final outcome
✓
Draw every step, handoff, system and approval on one page
✓
Mark the places where cases wait between teams
Which end-to-end process in your operations carries the highest volume and the highest cost?
Keep in mind: Map the process as it runs today, not as the procedure describes it. Workarounds often hide the real queue.
STEP 2
Measure Where Work Waits
Most of the time a case spends in a process is waiting, not being worked on. For each step, count the items in the queue at the start of each day, the items that arrive and the items completed. Record the age of the oldest open case. A queue that grows every day marks a step that cannot keep up with demand. In the operations I have worked in, work waited longest at compliance and KYC review.
Little's Law links these numbers. John Little published its proof in 1961. It states that the average number of items in a queuing system equals the average arrival rate multiplied by the average waiting time. If the queue at a step doubles while arrivals stay the same, the average wait at that step doubles too.
✓
Count queue, arrivals and completions for each step daily
✓
Record the age of the oldest open case per step
✓
Calculate average waiting time from queue size and arrivals
At which step in your process does the queue grow, even in a normal week?
Keep in mind: Waiting time is rarely in any report. Measure it directly, because it usually exceeds the time spent working.
STEP 3
Run a Bottleneck Analysis on Data
Compare capacity with demand at each step. Capacity is the number of cases a team can complete in a day with its current staff, skills and systems. The step with the lowest capacity relative to demand is the constraint. Check the result against the queue data: the largest and oldest queue usually sits just in front of it.
Watch utilisation closely. A Kellogg School of Management blog notes that for a single server, waiting times more than double at 80% utilisation and double again at 90%. The Theory of Constraints Institute adds that most systems have one single resource constraint. That step deserves attention first.
✓
Compare daily capacity with daily demand at every step
✓
Check the result against the largest and oldest queue
✓
Confirm the constraint with the team that runs the step
Which team in your process works at full utilisation while others wait for its output?
Keep in mind: A team that looks busy is not always the bottleneck. Follow the queue, not the overtime.
STEP 4
Exploit the Constraint First
Before you hire or buy, get more out of the bottleneck you already have. Make sure it never runs out of work and never works on incomplete cases. Move admin tasks, meetings and simple checks to other teams. Cover breaks and absences so that the constraint step keeps running through the whole day.
The potential is often large. The Theory of Constraints Institute states that the constraint of most organisations is not well utilised, often less than 50% on a round-the-clock basis. Applying the first three focusing steps typically exposes a minimum of 30% hidden capacity within the first few months, available free of cost, without any investment.
✓
Keep a buffer of complete cases ready for the constraint
✓
Move admin work and simple checks away from the bottleneck
✓
Cover breaks and absences at the constraint step
How many hours a day does your constraint team spend on work only it can do?
Keep in mind: An hour lost at the bottleneck is an hour lost for the whole process. It cannot be recovered elsewhere.
STEP 5
Align Other Steps to the Bottleneck
Every other step now serves the constraint. Release work into the process at the pace the bottleneck can absorb, not faster. Check cases for completeness before they reach it, so that missing documents are chased upstream and not by the scarcest team. Rework that returns to the constraint uses its capacity twice.
Improving a step that is not the constraint adds little. The Theory of Constraints Institute advises to maximize the utilization and productivity of the constraint, not of other steps. McKinsey describes a bank that achieved a 20 percent efficiency improvement in account closure, a process worth less than 1 percent of the total operations cost. My article on the four eyes principle shows how to place checks without adding a queue.
✓
Release new work at the pace of the constraint
✓
Check completeness before cases reach the bottleneck step
✓
Track rework that returns to the constraint team
How many cases reach your constraint team incomplete and have to be sent back?
Keep in mind: Local efficiency targets push teams to process more than the bottleneck can take. Align targets with total output.
STEP 6
Plan Capacity for Peaks and Growth
Once the constraint runs well, decide whether to elevate it with more staff, cross-trained cover, automation or a redesigned step. Then look ahead. Capacity management compares forecast demand, including month-end, campaigns and regulatory deadlines, with the capacity of the constraint. When you add capacity the bottleneck may move to another step, so repeat the analysis.
Deadlines are getting tighter. ESMA has proposed changes to facilitate the transition to a shorter settlement cycle (T+1) by 11 October 2027 in the EU. UK firms had to comply with the operational resilience rules by 31 March 2025, which includes testing severe but plausible scenarios for important business services.
✓
Forecast demand for month-end, campaigns and regulatory deadlines
✓
Compare the forecast with capacity at the constraint step
✓
Repeat the analysis whenever capacity is added or removed
When did your bank last check whether the bottleneck moved after a change?
Keep in mind: Elevate the constraint only after you have exploited it. Otherwise new capacity hides old waste.
The Resource Utilization Trap | Henrik Kniberg (Coach @ Crisp)
SAMPLE CASE
N26
N26, a German digital bank, grew quickly across Europe while its controls against money laundering failed to keep pace. In May 2021 BaFin ordered the bank to ensure adequate personnel, technical and organisational resources for its anti-money laundering obligations and appointed a special commissioner to monitor the work.
The capacity gap showed in the reporting queue. BaFin imposed an administrative fine of EUR 4.25 million for the delayed submission of a high number of suspicious transaction reports. In October 2021 it limited N26 to a maximum of 50,000 new customers per month, stating that the shortcomings in risk management were due to the bank's substantial growth.
The problem returned. In May 2024 BaFin fined N26 EUR 9.2 million after the bank systematically filed reports late in 2022. BaFin lifted the growth restriction effective 1 June 2024, after N26 had invested more than EUR 100 million in compliance, infrastructure and teams.
Key takeaway: Growth is a change in demand. When volume rises faster than the capacity of a control step, the queue grows quietly until a regulator or a client notices it first.
"We are pleased about the trust of our regulators and will continue our close exchange in the future."
Valentin Stalf (CEO @ N26)
In Short
Finding the bottleneck in bank operations rests on six habits: map the flow end to end, measure where work waits, confirm the constraint with data, exploit it before buying capacity, align every other step to its pace and plan capacity for peaks and growth.
The result is shorter waiting times, more output from the same teams, investment aimed at the one step that limits the process and early warning when demand outgrows capacity.
Frequently Asked Questions
What is bottleneck analysis?
Bottleneck analysis is a method to find the step in a process that limits its total output. It compares the capacity of each step with the demand placed on it and looks at where queues build up. The step with the least spare capacity is the constraint. Improving that step raises output, while improving other steps mostly moves more work into the queue.
What is the theory of constraints?
The theory of constraints is a management approach developed by Eliyahu Goldratt. It holds that every system has at least one constraint that limits its performance. Its five focusing steps are to identify the constraint, exploit it, subordinate everything else to it, elevate it and prevent inertia from becoming the new constraint once the old one has been broken.
How is capacity management different from bottleneck analysis?
Bottleneck analysis looks at the process as it runs today and finds the step that limits output now. Capacity management looks ahead. It compares forecast demand, such as month-end peaks, product launches or new regulation, with the capacity of each step. Together they show both where to act today and where the next bottleneck is likely to appear.
TAGS
ABOUT THE AUTHOR
Alexander Kiel
Business Analyst for process optimisation, data migration, business integration and project support in banking and wealth management. Lean Six Sigma Green Belt and Certified Scrum Master. Former private banking front office, with front-to-back experience from client advisory to data and systems.
KEEP READING
Related Articles
PROCESS OPTIMISATION
17 OCT 2026 · 7 MIN READ
Simplify the Process Before You Automate
Ford cut accounts payable head count by 75% by redesigning work. Process simplification gives banks a smaller and cheaper process to automate.
PROCESS OPTIMISATION
11 DEC 2025 · 8 MIN READ
Automate Reconciliation in Bank Operations
18% of accountants make errors every day. Reconciliation automation with clear matching rules and owned exceptions keeps bank books in line.
PROCESS OPTIMISATION
16 OCT 2025 · 8 MIN READ
Desktop Procedures That Survive a Handover
42% of institutional knowledge is unique to one person. Desktop procedures a newcomer can follow keep bank operations stable after a handover.