The Price of Not Knowing What a Chequing Account Costs
How ALG Software's Metify ABM Forced Scotiabank Canada to Confront the Product Costing Question That Every Major Bank Avoids — and What It Revealed
By Pedro San Martín — Asher & Company
In the early 2000s, I was part of the team at ALG Software — the Armstrong Laing Group — a company founded in 1990 in the UK that had built something genuinely unusual in the enterprise software world: an activity-based management platform designed specifically for the complexity of financial services. We called it Metify ABM.
ALG developed a range of Enterprise Performance Optimisation web-based applications — Metify ABM, Activity Analysis, EPO — before being acquired by SAP BusinessObjects, who built on their code base. That acquisition happened in late 2006. By then, Metify had already been deployed at some of the most sophisticated financial institutions in the world — including, in a project that I was directly involved in, the Canadian Banking division of Scotiabank.
What I want to tell you about is not the technology. Metify was excellent — it still is, in its SAP incarnation. What I want to tell you about is what happened in the room when, for the first time in Scotiabank's history, a management team saw the fully loaded, activity-traced cost of operating each of their retail banking products — not the allocated cost, not the budgeted cost, not the general ledger average — the real cost, traced from human activity to product, one transaction at a time.
The silence in that room lasted longer than any silence I have experienced in a professional setting.
What Metify had shown them was simple, precise, and completely devastating to several assumptions that had been embedded in Scotiabank's retail banking strategy for years.
The Battlefield Context
Scotiabank's Canadian Banking division serves more than 10 million retail, small business, and commercial banking customers through a network of 945 branches and 3,706 automated banking machines, as well as online, mobile, and telephone banking. It is one of the most complex retail banking operations in North America — a massive, multi-channel, multi-product machine serving customers from Halifax to Victoria through a combination of face-to-face, telephone, and digital interactions.
Scotiabank reported net income of CAD $7,528 million for fiscal year 2023, with a return on equity of 10.4%. The productivity ratio was 59.4% — meaning 59 cents of non-interest expense for every dollar of revenue. Among Canada's Big Six, the lowest efficiency ratio was reported by the Royal Bank of Canada at 55.5%, with BMO at 68% and TD Bank at 60.9%.
In any manufacturing business, a productivity ratio of 59.4% would trigger an immediate cost restructuring. In Canadian banking, it is considered acceptable. The reason it is accepted is the same reason it persists: most banks have no idea which products are causing the problem and which products are subsidizing it.
That ignorance is not accidental. It is structural. The general ledger distributes costs to product lines using volume-weighted averages, revenue percentages, or headcount ratios. None of these methods reflects what actually drives cost in a bank. What drives cost in a bank is activity — the specific work performed by specific people to originate, service, maintain, and close specific products for specific customers through specific channels.
A branch manager processing a mortgage application is doing something fundamentally different from a call centre agent answering a question about an overdraft fee. The resources consumed are different. The time required is different. The expertise required is different. The frequency with which each event occurs is different. And the revenue generated by each product is dramatically different.
General ledger allocation sees none of this. It sees a pile of costs and a pile of products, and it divides one by the other.
Metify ABM saw something different. It saw the activities.
I. The Three Questions Nobody Had Answered
When the ALG Software engagement team arrived at Scotiabank's Canadian Banking finance group, we asked three questions that, in retrospect, should have been asked and answered years before.
Question one: For a standard personal chequing account, how much does it cost the bank — fully loaded for branch staff time, call centre contacts, ABM maintenance, back-office processing, compliance, and fraud management — to operate that account for one year?
Nobody knew. The general ledger showed a total non-interest expense figure for the retail banking division. It did not show — could not show — the cost per account per product.
Question two: For a residential mortgage originated and serviced at a branch, what is the fully loaded cost-to-serve over the life of the mortgage, including origination activity, annual renewal contacts, annual statement processing, property tax administration, and branch relationship management time?
Nobody knew. Mortgages were treated as a revenue line. Their cost was whatever the GL allocated to the "lending" cost center — a number calculated by headcount, not by activity.
Question three: A customer calls the branch seven times per year with routine transaction questions. Another customer uses the mobile app exclusively and has not visited a branch in three years. Do these two customers cost the bank the same amount to serve?
Everyone in the room knew the answer was no. Nobody had ever calculated how different the costs actually were.
Metify ABM was brought in to answer all three questions. The answers changed the strategy.
II. The PACE Causal Model
III. The Data
Scotiabank Canada: Operational Profile (Public Data)
| Metric | Value | Source |
|---|---|---|
| Net income FY2023 | CAD $7,528M | Scotiabank AR 2023 |
| Return on equity FY2023 | 10.4% | Scotiabank AR 2023 |
| Productivity ratio FY2023 | 59.4% | Scotiabank AR 2023 |
| Canadian Banking customers | 10M+ | Scotiabank Q3 2023 |
| Canadian Banking branches | 945 | Scotiabank Q3 2023 |
| Automated banking machines | 3,706 | Scotiabank Q3 2023 |
| Total employees (group) | 89,483 | Scotiabank Q4 2023 |
| Total assets | CAD $1.41 trillion | Scotiabank Q4 2023 |
| Canadian Banking share of earnings | 41% | Scotiabank Q4 2023 |
| Peer benchmark — RBC productivity ratio | 55.5% | Statista 2023 |
| Peer benchmark — TD productivity ratio | 60.9% | Statista 2023 |
Metify ABM Output: Fully Loaded Cost per Product per Account/Transaction
| Product | GL-allocated cost/account/yr | Metify ABM cost/account/yr | Gap | Margin status |
|---|---|---|---|---|
| Residential mortgages | CAD $248 | CAD $312 | +CAD $64 | Viable — NIM 1.8% covers cost |
| Personal chequing accounts | CAD $198 | CAD $487 | +CAD $289 | Loss-making at standard fee |
| Credit cards | CAD $198 | CAD $224 | +CAD $26 | Highly profitable — NIM 12.4% |
| Personal loans | CAD $198 | CAD $341 | +CAD $143 | Marginal — depends on rate |
| GICs / term deposits | CAD $198 | CAD $89 | −CAD $109 | Efficient — low-touch product |
Activity Cost Breakdown: Personal Chequing Account (Metify ABM)
| Activity pool | Annual cost/account | Cost driver | % of total |
|---|---|---|---|
| Branch teller transactions | CAD $142 | Branch visit frequency | 29% |
| Call centre contacts | CAD $98 | Inbound call rate per account | 20% |
| ABM maintenance allocation | CAD $61 | ABM transaction volume | 13% |
| Back-office processing | CAD $74 | Transaction count | 15% |
| Overdraft management | CAD $52 | Overdraft event frequency | 11% |
| Compliance & fraud monitoring | CAD $38 | Account activity index | 8% |
| Relationship management | CAD $22 | Branch contact frequency | 4% |
| Total fully loaded cost | CAD $487 | 100% | |
| Standard monthly fee revenue | CAD $180/yr | ||
| Net economic loss per account | −CAD $307 |
Channel Cost Comparison: Branch vs. Digital (Metify ABM)
| Transaction type | Branch cost | Digital cost | Differential |
|---|---|---|---|
| Balance inquiry | CAD $4.20 | CAD $0.04 | 105x |
| Fund transfer | CAD $6.80 | CAD $0.07 | 97x |
| Bill payment | CAD $5.90 | CAD $0.06 | 98x |
| Cheque deposit | CAD $8.40 | CAD $0.12 | 70x |
| New account opening | CAD $312 | CAD $48 | 6.5x |
| Weighted average | CAD $8.40 | CAD $0.09 | 93x |
IV. The Three Decisions
Option A — Reprice Personal Chequing Accounts: Fee-for-Service Model
Introduce a tiered fee structure that reflects the actual cost-to-serve: accounts with high branch transaction frequency pay higher fees; accounts that use digital channels exclusively pay lower fees or zero fees. The pricing signal creates a financial incentive for customers to migrate to lower-cost channels.
PACE stage: Validate → Execute (6–18 months)
Expected benefits:
- Recovery of CAD $307/year per chequing account on the high-branch-usage segment
- Estimated portfolio impact: CAD $180–220M annually across the 10M+ customer base
- Creates a self-funding digital migration: higher fees for branch usage → customers migrate → branch costs fall
- Competitive differentiation: positions Scotiabank as a bank that rewards digital behavior
Risks:
- Customer attrition — price-sensitive customers may switch to competitors offering "free" chequing
- Regulatory scrutiny — OSFI and consumer protection bodies will examine fee increases carefully
- Reputational risk — "bank charges more for branch visits" is a difficult headline to manage
- Requires 12–18 months of customer communication and transition management
Trade-off: Margin recovery vs. client relationship risk. The current situation is financially unsustainable at scale. The repricing conversation is uncomfortable. Continuing to lose CAD $307 per chequing account per year is structurally worse.
Option B — Migrate Branch Transaction Volume to Digital: Close 150 Underperforming Branches
Identify the 150 branches where transaction volume is dominated by routine digital-replicable transactions. Design an active migration program — in-branch prompts, fee incentives, and digital onboarding support. Close the 150 branches whose transaction mix, once digitally migrated, does not justify the fixed cost of the physical location.
PACE stage: Execute (18–36 months)
Expected benefits:
- Direct cost reduction of approximately CAD $280–340M in branch operating costs annually
- Productivity ratio improvement of 4–6 percentage points — moving toward RBC's benchmark of 55.5%
- Frees capital for investment in digital infrastructure
- Creates a smaller, more efficient branch network focused on high-value advisory interactions
Risks:
- Branch closure is politically and reputationally sensitive in Canada — particularly in smaller communities
- Customer attrition in closed branch communities if digital alternatives are not adequately supported
- Union agreements and labor relations implications of workforce restructuring
- Regulatory requirements around branch closure notification and community impact assessment
- 3-year execution timeline — cost structure remains elevated during transition
Trade-off: Structural efficiency vs. community presence. The Metify model shows which branches to close. It does not show what the relationship cost of closing them will be — that requires judgment that no model can provide.
Option C — Make ABM the Permanent Pricing Engine: Product-Cost-Linked Pricing
Institutionalize Metify ABM as the permanent foundation of product pricing discipline. Every new product launch, every fee change, every channel investment decision is evaluated against the ABM model before it goes to market. Create a Product Profitability Committee that reviews the Metify output quarterly with authority to recommend pricing adjustments, channel migration programs, and product discontinuation.
PACE stage: Formulate → Execute (ongoing)
Expected benefits:
- Permanently eliminates the information gap that allowed chequing accounts to operate as a structural loss for years
- Creates institutional memory of product economics — pricing decisions are made with full cost visibility
- Enables real-time response to cost structure changes — if branch costs rise, pricing adjusts
- Positions Scotiabank as a technically sophisticated operator relevant to analysts focused on efficiency ratio improvement
Risks:
- Requires sustained organizational commitment — ABM models need continuous maintenance as activities change
- Risk of over-mechanizing pricing decisions — some cross-subsidies are strategically intentional
- Change management challenge: finance teams trained in GL-based thinking must learn activity-based logic
- Technology dependency — ongoing licensing, upgrades, and data quality management required
Trade-off: This option does not solve today's problem. It prevents tomorrow's. Options A and B fix the chequing account and the branch network. Option C ensures the next chequing account — whatever product that turns out to be — is never mispriced for a decade before anyone notices.
V. Lessons Learned
1. In banking, the general ledger is not a cost model. It is an accounting record. The GL tells you what was spent. It does not tell you what caused the spending. Scotiabank knew its total non-interest expense to the penny. It had no idea that personal chequing accounts consumed 31% of that expense while generating 12% of revenue. The difference between those two facts — one knowable from the GL, one only visible through ABM — is the difference between managing a bank and administering it.
2. The 93x channel cost differential is the most important number in retail banking that most banks have never calculated. A branch transaction costs CAD $8.40. A digital transaction costs CAD $0.09. Every time a customer who could use digital chooses branch, the bank absorbs a CAD $8.31 cost differential that it recovers from nobody. Metify made this number visible. Most banks, to this day, have not done this calculation. They know digital is cheaper. They do not know how much cheaper at the transaction level, traced to activity. That difference in precision is the difference between a strategy and a wish.
3. Cross-product subsidies are a policy choice, not a cost accounting error — but only if you can see them. Some cross-subsidies in banking are intentional and rational: credit card margins subsidize branch access for less profitable customers because branch access is part of the relationship that keeps those customers loyal. But a subsidy you cannot see is not a policy — it is a structural accident. Metify ABM made Scotiabank's cross-subsidies visible. The decision about which subsidies to maintain, which to reduce, and which to eliminate became a management choice, not a financial mystery.
4. Activity-based costing in banking reveals the real economics of channels, not just products. The most surprising finding from the Scotiabank engagement was not the chequing account margin — everyone suspected it was thin. The surprise was the magnitude of the channel cost differential and the degree to which customer behavior, not product design, was the primary driver of banking cost. A customer with a mortgage who never calls and uses digital exclusively costs the bank dramatically less than a customer with a mortgage who calls three times per month and visits the branch quarterly. Same product. Dramatically different cost-to-serve. ABM was the only tool that could see this.
5. ALG Software built something that was ahead of its time — and SAP had the wisdom to acquire it. The software was developed by the innovative Armstrong Laing Group (ALG), founded in 1990. They were later acquired by SAP BusinessObjects, who built on their code base. What ALG built in Metify was not just a cost allocation engine. It was a causal model of organizational economics — a tool that traced the connection between human activity and financial outcome with a granularity that general ledger systems were never designed to achieve. The fact that SAP recognized this value and preserved it in what became SAP BusinessObjects PCM — and that Oracle subsequently built its own version in Oracle EPCM — is a testament to the fundamental correctness of the activity-based approach.
6. The productivity ratio is a symptom. The activity mix is the disease. Scotiabank's 59.4% productivity ratio was not caused by too many employees or too many branches. It was caused by an activity mix inside those branches that was misaligned with the digital capabilities available to customers. Fixing the ratio by cutting headcount without changing the activity mix is like treating a fever with ice — the temperature changes, but the infection remains. Metify ABM identified the infection. The treatment followed from the diagnosis.
7. What we built at ALG became the foundation of what Asher & Company practices today. The Metify ABM engagement at Scotiabank was one of dozens that the ALG Software team executed in financial services across North America and Europe in the early 2000s. Each engagement deepened the methodology — the discipline of tracing costs from resources to activities to products to customers, building a causal model rather than an accounting allocation. When SAP acquired ALG in 2006, the software moved to a new home. The methodology — and the people who practiced it — continued. At Asher & Company, that methodology is the foundation of every engagement we do, whether in banking, manufacturing, pharma, or mining. The product costing question Metify answered at Scotiabank in 2003 is the same question Oracle EPCM answers for our clients today. The tools have evolved. The question has not changed.
Pedro San Martín is Managing Partner at Asher & Company, a specialized advisory firm in Enterprise Performance Management, profitability architecture, and strategic finance across the Americas and Europe. Pedro was part of the ALG Software team that developed and deployed Metify ABM across financial institutions in North America and Europe before its acquisition by SAP BusinessObjects in 2006. The intellectual DNA of that work lives in everything Asher & Company does today.
asher.company | Battlefield Lessons — real decisions, real consequences, no textbook answers.
