What Analysts Are Really Asking CFOs? Performance Engineering Is the Answer Boards Are Demanding.
In earnings calls for publicly traded banks this January, analysts had plenty of questions about AI. But they had 2X as many about something else entirely.
According to a new PwC analysis of publicly traded bank earnings calls, 30% of analyst questions centered on the quality and resilience of revenue growth. Another 25% focused on expense flexibility and operating leverage. Just 15% were about AI investment and execution.
Read that again. Performance questions outnumbered AI questions by more than 3 to 1.
This is not a signal you can afford to misread. Analysts are not scoring CFOs on how boldly they are transforming. They are scoring them on whether their revenue can hold up across cycles. Whether their cost base can flex under pressure. Whether their operating model is built to perform, not just to change.
1. The Investment Is Not The Problem
To be fair to the banks in the room, the investment is real. Banks are pouring capital into exactly the kind of platform work that should matter. General ledger upgrades. Data architecture. ERP modernization. Multi-year programs reshaping reporting, controls, and operating models from the ground up.
PwC is right to highlight this. These are serious investments that deserve serious returns.
But here is the question PwC raises without fully answering: how do you ensure those investments translate into measurable ROI and durable efficiency gains?
That is not a technology question. It is a business performance engineering question. And most transformation programs are not designed to answer it.
They are designed to deliver. Scope is defined. Milestones are hit. Systems go live. And somewhere between the go-live and the earnings call, the performance gains that justified the investment fail to materialize in the way the market expects.
The reason is structural. Most programs start from technology and work toward the business. They spend months in discovery, learning how the institution actually operates, mapping processes, negotiating requirements, building from zero. By the time the system is live, the market has already moved.
McKinsey puts a number on how often this plays out badly. 70% of digital transformation efforts fail. 60 to 80% of technology capacity across financial institutions is locked into maintenance rather than performance. The investment goes in. The performance does not come out.
Analysts are not asking about transformation because they have already seen what transformation delivers.
2. Business Performance Has To Be Engineered. Not Promised.
This is where the conversation needs to shift.
The institutions that will answer the analyst question convincingly aren't the ones with the boldest transformation roadmaps. They are the ones that have engineered performance into the architecture of their business from the start.
That means beginning not from technology, but from business value drivers. Where does this institution actually create revenue? Where does cost accumulate? Where does risk concentrate? What are the specific levers that connect operational decisions to P/L impact?
When those questions are answered first, the software architecture that follows is not a transformation program. It is a performance engine. Every component is designed around a business outcome. Every decision point is connected to a measurable result. The system does not just operate. It performs.
This is what business performance engineering means in practice. Strategy mapped to architecture. Architecture built to deliver P/L impact. Not someday. From day one.
3. This Is What The Analyst Questions Are Really Asking For On The Earnings Call
When analysts press CFOs on revenue resilience, they are not asking about the transformation program. They are asking whether the business has a performance engine underneath it.
Can the revenue hold up across cycles? That depends on whether decisions, execution, and controls are connected to the value drivers that actually generate it.
Can the cost base flex under pressure? That depends on whether operational architecture was engineered for efficiency from the start, not bolted on afterward.
Can the operating model scale without losing control? That depends on whether the system was built around how the business actually runs, or around how a vendor thought it should run.
These are not technology questions. They are performance engineering questions. And they have a specific answer.
The market is already asking for this. The question is whether your intelligent performance operating system is built to answer.