The budget that disappears
A $5 billion asset regional bank budgeted $20M for "digital innovation" in 2023.
Their roadmap included:
- Mobile treasury enhancements
- Real-time payments
- API banking for commercial clients
- Enhanced fraud detection
- Digital account opening
12 months later, here's what actually got built:
- Mobile treasury enhancements: Partially completed (6 months late)
- Real-time payments: Delayed to 2024
- API banking: Cancelled due to architecture constraints
- Enhanced fraud detection: Scaled back to "pilot"
- Digital account opening: Requirements gathering phase only
Where did the $20M go?
Post-project analysis revealed:
- $7M: New capabilities that actually launched
- $5M: Integration work to connect new features to legacy core
- 3M: Maintenance and patches for existing systems (unplanned)
- $2.5M: Technical debt service (fixing problems created by past projects)
- $1.5M: Abandoned projects (API banking, scaled-back pilots)
- $1M: Compliance and regulatory updates to legacy systems
Only 35% of the "innovation budget" built innovation.
The other 65% kept existing systems running.
The four budget vampires
Where does innovation budget actually go? Four categories of hidden costs consume 40-60% of what banks think is innovation spending.
1. The integration tax
What it is: Every new digital feature requires custom integration with legacy core systems.
Why it's expensive: Legacy systems weren't built for integration. They have no APIs, no standard data formats, and no real-time capabilities.
Real example:
A bank wanted to add Zelle to their mobile app (a feature competitors launched years earlier).
Estimated cost: $500K
Actual cost: $1.8M
Why the overrun:
- Core system doesn't support real-time balance checks (Zelle requires instant verification)
- Built middleware layer to poll core every 30 seconds and cache results
- Middleware created data sync issues (transactions processed but not reflected in balances)
- Additional work required to reconcile Zelle transactions with batch-processed core updates
- Three months of debugging edge cases where middleware and core disagreed
The integration work cost 3× more than the feature itself.
Every new feature that touches the core system incurs this tax. For a typical regional bank, 20-30% of "innovation budget" is actually integration tax.
2. Unplanned maintenance
What it is: Emergency fixes, patches, and support work that wasn't budgeted but must be done.
Why it's expensive: Legacy systems break unpredictably. When they break, everything stops until they're fixed.
Real example:
A regional bank's online banking platform went down for 8 hours on a Monday morning.
Root cause: A security patch applied to the core system over the weekend created a conflict with the middleware layer built two years earlier for mobile banking.
Cost impact:
- $200K in direct fix costs (emergency contractor rates, all-hands debugging)
- $50K in credits to affected business clients
- $150K in delayed project work (entire engineering team pulled off projects to solve the crisis)
Total unplanned cost: $400K
This was one incident. The bank had 6-8 similar incidents that year.
Banks typically experience 10-15% of their IT budget consumed by unplanned maintenance—work that comes out of the "innovation" budget because there's nowhere else to take it from.
3. Technical debt service
What it is: Ongoing costs to maintain past "temporary" solutions that became permanent.
Why it's expensive: Every workaround built in the past requires ongoing maintenance, creates new bugs, and slows down future projects.
Real example:
In 2018, a bank built a "temporary" middleware layer to enable mobile balance checks while they planned to replace their core system "within 3 years."
Five years later:
- The core system is still there (replacement delayed indefinitely)
- The "temporary" middleware is now business-critical infrastructure
- Two full-time engineers maintain it
- It's been patched 47 times
- It's the source of 15-20 support tickets per month
- Every new digital project must account for its quirks
Annual cost to maintain this one "temporary" solution: $450K+
Most banks have 5-10 of these inherited "temporary" solutions.
Technical debt service typically consumes 10-15% of IT budgets—money that could build new features but instead services old decisions.
4. Compliance and regulatory drag
What it is: Mandatory updates to legacy systems for regulatory compliance.
Why it's expensive: When regulations change, banks must update all systems—including legacy cores that are expensive and slow to modify.
Real example:
When beneficial ownership reporting rules changed in 2024, a bank had to update:
- Core banking system (vendor charged $200K for the update)
- Account opening workflow (custom build, $150K)
- Customer data warehouse (another $100K)
- Compliance reporting system ($75K)
- Training and implementation ($75K)
Total cost: $600K for a regulatory requirement that added zero competitive value.
Regulatory updates to legacy systems consume 5-10% of IT budgets annually.
The real budget math
Here's what a typical $20M "innovation budget" actually funds:
Your $20M innovation budget is actually a $7M innovation budget.
The other $13M keeps the lights on.
Why CFOs don't see this
When CFOs review IT budgets, they see line items like:
- "Commercial Banking Digital Transformation: $20M"
- "Mobile Banking Enhancements: $3M"
- "API Banking Initiative: $5M"
These look like innovation investments.
What they don't see:
- 60% of "Digital Transformation" budget consumed by legacy integration
- "Mobile Banking Enhancements" delayed 9 months due to core system constraints
- "API Banking Initiative" cancelled after $2M spent because architecture couldn't support it
The budget categories hide the reality: most "innovation" spending maintains the past, not builds the future.
The compounding problem
The innovation budget illusion isn't static—it gets worse over time.
Year 1: 35% of innovation budget builds new capabilities
Year 3: 25% builds new capabilities (more integration tax, more technical debt service)
Year 5: 15% builds new capabilities (maintenance consuming most resources)
Why this happens:
Each new digital feature built on legacy infrastructure:
- Adds to integration complexity (more systems to maintain)
- Creates new technical debt (more workarounds to service)
- Increases unplanned maintenance (more ways things can break)
- Slows future projects (more constraints to work around)
The more you build on legacy, the less innovation budget you have for future building.
This is why many regional banks launched 5-7 new digital features in 2020 but only 1-2 features in 2024. The innovation capacity is declining even as budgets stay flat or grow.
The competitor advantage
Digital-first banks and fintechs don't have this problem.
Their $20M innovation budget breaks down like this:
They spend 70% of budget on innovation because they have 10% maintenance costs, not 50%.
This is the structural advantage of modern architecture.
When you build on modern platforms:
- Integration is cheap (APIs are built-in)
- Unplanned maintenance is rare (systems are stable)
- Technical debt is low (architecture is modular)
- Compliance updates are fast (one platform to update, not five)
Real-world innovation gap
Let's compare two banks with identical $20M innovation budgets:
Legacy Bank:
- Effective innovation budget: $7M (35%)
- New features launched per year: 2-3
- Time to market: 18-24 months
- Innovation rate declining
Digital-First Bank:
- Effective innovation budget: $14M (70%)
- New features launched per year: 8-12
- Time to market: 3-6 months
- Innovation rate accelerating
Over 5 years:
- Legacy Bank launches 10-15 features
- Digital-First Bank launches 40-60 features
By year 5, there's no comparison.
The client experience gap is so wide that competition becomes impossible. Digital-First Bank has launched features Legacy Bank is still planning.
The talent drain effect
The innovation budget illusion has a second-order effect: it drives away top talent.
What talented engineers want:
- To build new things
- To work with modern technology
- To see their work launch quickly
What they get at legacy-dependent banks:
- 50-60% of time maintaining old systems
- Wrestling with 1980s architecture
- Projects delayed or cancelled due to technical constraints
Real exit interview from a senior engineer:
"I joined to build the future of commercial banking. Instead, I spent 18 months building middleware to make a core system from 1987 talk to a mobile app. Every project was 'connect this modern thing to that ancient thing.' I never built anything truly new. I left for a fintech where I can actually innovate."
When top talent leaves, innovation capacity drops even further.
Three questions for banking leaders
1. What percentage of our "innovation budget" actually builds new capabilities?
Do the math:
- How much went to integration work?
- How much to unplanned maintenance?
- How much to technical debt service?
- How much to compliance updates?
If the answer is below 40%, you have an innovation budget illusion.
2. Is our innovation capacity increasing or decreasing?
Count features launched:
- 3 years ago: [X] features
- 2 years ago: [Y] features
- Last year: [Z] features
If Z < X, your innovation capacity is declining even if budgets are flat or growing.
3. How does our effective innovation budget compare to competitors?
If digital-first competitors have 70% innovation capacity and you have 35%, they can outpace you 2:1 even with identical budgets.
Over 5 years, that gap becomes insurmountable.
The bottom line
Most regional banks think they're spending 20-25% of IT budgets on innovation.
The reality is 8-12%.
The rest is consumed by:
- Integration tax (connecting to legacy)
- Unplanned maintenance (keeping legacy running)
- Technical debt service (maintaining past workarounds)
- Compliance drag (updating legacy for regulations)
Your innovation budget is smaller than you think.
And it's shrinking every year as:
- More digital features add integration complexity
- More technical debt accumulates
- More systems require maintenance
- More talented engineers leave
Eventually, the innovation budget shrinks to near zero—not because funding was cut, but because legacy maintenance consumed everything.
That's the trajectory for banks that defer modernization.
The only way to restore innovation capacity is to change the underlying architecture—not add more digital wrappers on top of it.
