All posts
7 min read

Forrester is right about the IT Finance reset. The architecture underneath is the harder problem.

Forrester's diagnosis of IT Finance is correct. The relaunch will not stick on top of the same architecture that produced the diagnosis. Two architectural realities — and three structural moves — define what the reset actually requires.

The diagnosis is in. In his May 2026 piece for Forrester, Principal Analyst Greg Zorella made the case that CIOs have to fundamentally reset the role of IT Finance. Cost visibility is not the same as value insight. ITFM and TBM investments are not delivering the business influence they promised. The mandate has to shift from record-keeping to value co-creation, or IT Finance will keep losing the executive table.

We agree with every line of it. Forrester named what most IT Finance leaders have been quietly admitting for two years. The work now is to describe what the reset actually requires underneath the mandate, because the relaunch will not stick on top of the same architecture that produced the diagnosis.

Two architectural realities underneath the four barriers

Forrester names four structural barriers stalling IT Finance maturity. Weak enterprise data foundations. Stuck at TCO and showback. Missing crossover skills between finance and technology. Frameworks that cost more than they deliver. Every one of them is real, named with the precision of someone who has watched the same pattern across hundreds of practices.

Underneath the four sit two architectural realities. They explain why the barriers persist no matter how many transparency initiatives an organization runs.

Failure one: cost models that are never comprehensive. Cloud gets modeled one way. SaaS another. AI another. Vendor contracts and PDF invoices sit outside the model entirely. There is no semantic common language across spend categories. Margins and unit economics cannot be built from data that is missing 30 to 40 percent of the IT estate. Forrester's first barrier (weak enterprise data foundations) is the visible surface of this. The fourth barrier (heavyweight frameworks that cost more than they deliver) is what it costs to keep papering over it.

Failure two: financial context never reaches the moment of decision. Spend gets reported in portals stakeholders do not visit. Decisions happen in Slack, sprint tools, developer environments, agentic workflows. Current platforms cannot push contextualized cost data into any of those surfaces. Forrester's second barrier (stuck at TCO and showback) and third barrier (missing crossover skills) are downstream of this. A team cannot operationalize cost data they never see at the moment they need it, no matter how skilled they become or how mature the framework is on paper.

AI is making the gap acute

Forrester notes that AI magnifies every weakness in IT Finance. The numbers back the diagnosis. IBM's May 2025 survey of 2,000 CEOs found that only 25% of AI investments are delivering the business value organizations expected. That gap is not a model quality problem. The models work. It is a financial infrastructure problem. The architecture organizations are trying to govern AI with was built for an era when humans made spending decisions at human speed, through a small number of central governance moments.

That era is over. Spend now happens at machine velocity. Agents in Claude Code, Cursor, Codex, and internally-built workflows are choosing model architectures, retrying failed prompts across fallback chains, and routing inference across providers without checking a dashboard. They cannot check one. They were not built to. Every action carries cost implications the IT Finance team is accountable for and has no architectural way to see in time.

The Monday morning is familiar to every IT Finance leader by now. The CFO asks why the AI line on the P&L moved 40%. The CIO cannot answer at the granularity the question requires. The IT Finance team needs two weeks to assemble the data. By the time the answer arrives, the next quarter's investment decisions are already being made in the same blind spot the last quarter's were made in.

This is the scene Forrester is describing when he writes that CIOs lose credibility at the executive table. It is not a credibility problem. It is an information arrival problem. The information the CIO needs to defend the AI investment exists somewhere in the stack. It just does not get assembled into a defensible answer fast enough to matter.

Three structural moves the relaunch actually requires

The relaunch Forrester prescribes is the right destination. Three structural moves get there. These are the three any IT Finance practice has to make to complete the reset, and they are also the three we have organized StitcherAI's architecture around:

1. A semantic foundation across every category of spend. Cloud, SaaS, AI, vendor contracts, PDF invoices, internal IT. One business model that maps technology cost to business outcomes in the same vocabulary finance, engineering, product, and the board can all read. Without this, every conversation starts with a translation meeting and ends in a reconciliation exercise.

2. Distribution at the point of decision. Cost and margin context arrives where decisions are actually made: Slack, sprint tools, developer environments, the agentic workflows choosing architectures in real time. No new dashboard. No login. The context shows up next to the decision. Without this, accountability stays centralized with IT Finance no matter what the mandate document says.

3. Finance as code. The cost model becomes software. Version-controlled, peer-reviewed, auditable. When someone asks how an allocation was calculated, the answer is a commit history, not a cell reference in a file nobody owns. Without this, the reasoning the new system has to do at machine velocity is not trustworthy enough for the board to rely on.

None of these are dashboards. None of them are a smarter interface on top of the same data layer. They are different infrastructure, designed for the operating model Forrester is asking IT Finance to step into.

Where StitcherAI fits

The Monday morning looks different when this architecture is in place. The CFO does not have to ask why the AI line moved 40 percent. The CIO already knows. The cost model that produced the number was assembled in the same semantic layer that delivered context to the engineering team the moment they chose the model architecture. The variance is not a question to answer on Monday. It is a conversation already underway, with a defensible answer attached.

StitcherAI is that architecture, built as one system:

The semantic foundation ingests every category of spend, normalized through the FOCUS open billing standard, the one the major cloud providers and thousands of companies have already adopted and that StitcherAI's founder helped create.
The reasoning layer activates the foundation in real time, embedded in the workflows where humans and agents actually make decisions instead of waiting for someone to log into a dashboard.

From record-keeping → value cocreation.
From dashboard era → embedded intelligence.

Value in weeks, not years.

Forrester has named where IT Finance has to go. The architecture that gets it there is now available to every organization, not only the Fortune 100 willing to spend two years and several million dollars building it themselves.