Two customers can earn you the same revenue and cost a hundred times more to serve.
In banking and insurance the product is information, and the cost of moving it varies enormously. A wholesale account settled electronically and a retail account processed by hand can bill the same and cost worlds apart. A blended cost rate hides the difference. Time-driven costing puts a number on every transaction, channel and customer, so you finally see which relationships fund the bank and which quietly drain it.
Cost and Profitability Consulting · 150+ models since 2010 · TDABC
In financial services the unit cost of serving a customer varies 5 to 10 times around the average depending on demand intensity, industry research shows. In one illustrative sector pattern, two divisions with identical sales had wildly different costs to collect, a low five-figure cost versus a seven-figure one, a more than hundredfold difference, because one served a hundred wholesale accounts electronically and the other thousands of retail accounts by hand. Conventional ABC collapsed at this scale, hundreds of staff surveyed monthly, reports taking over a month, which is exactly why TDABC was invented. TDABC assigns cost through a capacity cost rate and time equations, giving true profit per customer, channel and transaction without surveys.
SAME REVENUE, A HUNDREDFOLD DIFFERENCE IN COST
Illustrative sector pattern. Two divisions, the same sales. One served a hundred wholesale accounts electronically; the other thousands of retail accounts by hand. The cost to collect diverged more than a hundredfold.
The numbers are already there. They are just blended away.
Unit cost varies 5 to 10 times around the average by demand intensity, industry research shows. A single blended cost rate, the industry default, is wrong for almost every customer it touches. Identical revenue can carry wildly different cost: in an illustrative accounts-receivable pattern, two divisions booked the same sales, one serving a hundred wholesale clients electronically at a low five-figure cost a year, the other thousands of retail accounts manually at a seven-figure cost a year, more than a hundredfold apart on the same revenue.
Conventional ABC does not scale here. Surveying hundreds of employees every month and waiting over a month for a report is the origin story of TDABC: the method was created precisely because activity-based costing collapsed inside a large financial institution. Studies find a large share of financial-services firms actively use ABC and many more pilot it, so adoption is high; what is missing is a method that survives the transaction volumes.
Two parameters, no surveys.
A capacity cost rate per processing or service group, and time equations that describe how each transaction, channel and customer consumes time. The cost drivers that matter here are transaction channel, manual versus electronic, exception and dispute rate, call intensity, product complexity, and the risk and compliance load carried by the relationship. The 5 to 10 times spread in unit cost comes straight from these drivers.
Customer cost = 0.5 min per electronic transaction + 9 min per manual transaction + 4 min per inbound call + 15 min per dispute or exception + 2 min per statement x capacity cost rate of the servicing group
Illustrative account-servicing equation. A single manual transaction costs roughly eighteen electronic ones; the average charges them the same.
ONE TRANSACTION, FOUR CHANNELS, FOUR COSTS
Illustrative. The same service request costs very different amounts by channel. A blended rate spreads these evenly and prices every channel wrong.
The whale curve in banking is steep.
A minority of customers and segments carry the institution while a long tail destroys value once fully costed. The transversal pattern holds: the top 20 percent of customers can represent 150 to 300 percent of profit, and 30 percent or more of relationships are unprofitable once cost to serve is loaded. The sector patterns prove the size of the prize. A large retail bank used TDABC over a terabyte-scale data set processed overnight to cut several hundred million of cost and lift its market value by billions over eighteen months, while collapsing reporting from over a month to a few days. A global brokerage eliminated over half a billion of cost and saw its share price climb sharply after replacing a legacy system that took over a month to produce a single report. A regional bank booked a seven-figure annual improvement in year one. The margin was always there; it was hidden inside a blended rate and a month-long close.
CUSTOMERS, RANKED BY CUMULATIVE PROFIT
Illustrative. Cumulative profit climbs well past 100 percent on the best customers, then a long tail of unprofitable relationships drags it back once cost to serve is loaded.
Strong data, weak process design.
Financial services scores among the highest of any sector on Cost Allocation, Profitability Visibility and Tools & Governance, and its Data & Technology maturity is among the strongest, alongside IT. The weak point is TDABC Process Design: the data and the tools are there, but processes are rarely mapped into time equations that survive transaction volume. ABC adoption is high, yet many implementations stall, a European bank replaced a failed ABC model with TDABC in its securities and derivatives processing precisely on this point. The gap is method at scale, not ambition.
Illustrative profile across the 7 dimensions of cost maturity. Financial services is strong on allocation, visibility, tools and data, and weakest on TDABC process design at scale.
When AI runs the back office, where does the margin move?
AI changes the cost base of banking and insurance directly: transaction processing automates toward the electronic end of the cost curve, AI absorbs tiers of risk and compliance work, and agents handle servicing that used to fill a call centre. As manual cost falls, the spread between customers narrows in some places and widens in others, and only a per-customer, per-channel cost model can tell you where. The institutions that win are the ones that already price on cost to serve, not on a blended average.
Six ways into the sector's cost.
Frequently asked questions
- How do you calculate the true cost of serving a bank customer?
- Beyond a blended cost rate, use time equations that load each channel, transaction type, call and exception against a capacity cost rate. Unit cost in financial services varies 5 to 10 times around the average depending on demand intensity, so an average is wrong for almost every customer it touches.
- Why are some bank or insurance customers unprofitable?
- High manual-transaction intensity, frequent exceptions and disputes, heavy call volume and low-margin products. In one illustrative accounts-receivable pattern, two divisions with identical sales had a more than hundredfold difference in cost to serve, because one served a hundred wholesale accounts electronically and the other thousands of retail accounts by hand.
- What is TDABC for financial services?
- Time-driven activity-based costing assigns cost through a capacity cost rate and time equations, giving per-customer, per-channel and per-transaction margin without surveying staff. It was invented because conventional ABC collapsed at banking scale, where surveying hundreds of employees monthly and waiting over a month for a report did not survive transaction volume.
- Why did ABC fail in banks?
- Surveying hundreds of staff every month and waiting over a month for a report does not survive transaction volume. Banks have replaced failed ABC models with TDABC in securities and derivatives processing and cut reporting from over a month to a handful of days.
See which relationships fund the bank, and which drain it.
ProfitAudit 360 builds the per-customer, per-channel picture on a TDABC base. Or take the Profit Check first: five minutes, no data upload, and it points to where your blended rate is most likely wrong.