Project vs product funding: what changes in your IT cost model

In brief

Moving from project to product funding changes the cost object, allocation, capex evidence, budget cadence and how success is measured. What to change.

Steve Elliott

Steve Elliott · · 7 min read


On this page
  1. What changes, in one table
  2. Which cost object should carry the money?
  3. How does cost allocation change?
  4. What happens to capitalization?
  5. How often should persistent teams be funded?
  6. How do you measure success without on time and on budget?
  7. Where to start
  8. Sources

When an organization moves from project funding to persistent product or value-stream funding, the main cost object changes from a temporary project to a long-lived team. That one change reaches allocation, capitalization, the budget calendar and the definition of success, and most IT cost models were built for the old cost object.

In our view, the money itself changes less than the model around it. What changes is the unit finance approves, the record that supports capex and the question each review asks.

What changes, in one table

How we see the shift:

Project fundingProduct or value-stream funding
Primary cost objectProject, with a start and end datePersistent team, product or value stream
What gets approvedScope, schedule and budget in a business caseTeam capacity for a period, against outcomes
Approval cadenceAnnual cycle, with change requests in betweenAnnual envelope, revisited at shorter intervals
Labor in the modelTimesheets coded to project codesTeam cost charged directly to its product
Capex evidencePhase gates on a project planFunding commitments and delivery records (see below)
Variance questionIs the project over budget?Is the team's spend producing the result it was funded for?
Success measureOn time, on budget, on scopeProduct outcomes and cost trend over time
TBM reportingProjects as a cost objectConsumer layer: products, value streams, portfolios

Which cost object should carry the money?

Under project funding, labor, contractors and capex all hang from the project code, which closes when the work is done. A persistent team doesn't close. Its cost recurs every month, so the team or product becomes the natural cost object.

The TBM Council has moved the taxonomy in the same direction. In TBM Taxonomy 5.0, the Business layer was renamed the Consumer layer, which "now includes Value Streams, Products, Portfolios, and Digital Services," according to the Council's migration guide. The guide's last step advises organizations to plan for future enhancements such as agile or product-centric costing using Consumer layer value stream dimensions. The taxonomy page describes the Consumer layer as the organizational units or roles that consume technology services. TBM Taxonomy 5.0.1 is the current version, released July 18, 2025.

In our view, projects don't disappear in a product model. They become an attribute. An epic, a regulatory change or a migration can still carry a code so that finance can report on it and support capitalization. But the code describes work the team is doing, not a separate budget the team draws from.

How does cost allocation change?

Cost allocation gets simpler for labor and harder for everything around it.

When a team is funded as a unit, its people are charged to its product directly, with no need to split each person's hours across project codes. Take an illustrative team of nine people at a fully burdened $180,000 each. That team costs about $1.62 million a year, or about $405,000 a quarter, whatever it works on. The product's labor cost is known before the quarter starts.

The harder part is shared cost. Hosting, data services, security and shared engineering teams serve many products. Product owners are likely to ask what they are paying for, so those costs need a defensible path to each product. The taxonomy allocates Solution costs directly to Consumers using usage, cost-driver or attribution models, and it allows headcount or even-spread methods too.

We think three choices matter most:

  • Which teams are products and which are shared services. A team that serves only other teams is usually better treated as a solution whose cost flows on, not as a product with its own outcomes.
  • Which drivers allocate shared cost. Consumption drivers, such as compute hours or API calls, hold up better in product reviews than headcount splits.
  • Where unallocated cost goes. Some shared cost will not have a good driver. Showing it as a visible line is better than spreading it thinly and calling it allocated.

What happens to capitalization?

Capitalization is where the two models clash most visibly, in our view. The Scaled Agile Framework (SAFe) CapEx and OpEx article says that historically, capitalization was applied in a waterfall or phase-gate process, where requirements and design milestones often served as the gates that started it. In agile development, it says, "these are not relevant phase gates." It suggests flagging features with capex potential, then estimating the capex share from story hours, completed story points or story counts. It says hours shouldn't be the default because of the overhead.

ASU 2025-06 removes all references to project stages from ASC 350-40. FASB said the old guidance was hard to apply, particularly in iterative development environments such as agile. Under the amended guidance, capitalization starts when management has authorized and committed to funding the software and completion is probable, and completion isn't probable while there is significant development uncertainty. The standard is effective for annual periods beginning after December 15, 2027. Our ASU 2025-06 post covers the detail.

In our view, the evidence moves for a product-funded organization. The phase gate stops being the trigger. Funding decisions and delivery records, such as a capacity approval, a requirements baseline or test results showing an unproven feature now works, can carry the weight. That can make the record of each funding decision part of the accounting evidence, not just the portfolio file.

Two practical points follow, in our view. First, a quarterly capacity approval for a persistent team may serve as evidence of a funding commitment, so record it with a date, an owner and the capabilities it covers. Second, the software capitalization record needs to link to the team's backlog at the capability level, because the team-level cost object is too coarse to show which work met the threshold. This is not accounting advice; confirm how the standard applies to your organization with your accounting advisers and auditor.

How often should persistent teams be funded?

Project funding follows the annual cycle: business cases compete, winners get budgets, and anything new waits for change control. Product funding keeps an annual envelope but makes allocation decisions more often. The sources differ on how often:

  • SAFe. The current Lean Budgets article says SAFe "funds value streams instead of projects," with each portfolio operating within an approved budget. SAFe 5's Participatory Budgeting article (an earlier version of the framework) says value stream budgets are typically updated twice a year, about every two program increments.
  • GAO. In GAO-25-107130, GAO found that leading companies update portfolios and reassess business cases at least every 6 months, and reserve their largest allocations until they have validated a design for the minimum viable product.
  • Bain. Bain describes leading organizations keeping a flexible annual allocation but adding check-ins through the year that can release more money or shift it between products, with distributions small at first and larger ones based on demonstrated results.
  • Beyond Budgeting. The Beyond Budgeting principles call for making resources available as needed rather than through detailed annual allocations, and organizing management processes around business rhythms rather than only the calendar year.

In our view, quarterly or semiannual reallocation within an annual envelope fits most enterprises better than abandoning the annual operating plan. The plan is approved as capacity for a set of products, and the mix inside it can move without a full re-budget.

This also changes variance analysis. A persistent team's spend is close to fixed, so the useful variance is between what the team was funded to achieve and what it delivered.

How do you measure success without on time and on budget?

On time, on budget and on scope measure delivery against a plan fixed at approval. A persistent team has no end date. GAO, in a 2026 report on defense acquisition, said success there is "all too often measured by activity (money spent), not by outcomes." That was about weapon systems, but the risk applies to any model that reports spend without results.

Bain describes a practical structure: senior leaders choose a handful of the most important business metrics, and product teams choose metrics they can influence that support them. Business, IT and finance leaders then use measured results, rather than annual business cases, to decide where funds go.

For the cost model, we think three measures do most of the work:

  1. Cost per product over time. Fully loaded, including allocated shared cost, trended by quarter.
  2. Cost against the product's outcome metric. Spend next to the measure the team was funded to move, so a funding review can see both at once.
  3. Run versus change inside each team. Persistent teams absorb support and maintenance. If change capacity shrinks quarter after quarter, the product is getting more expensive to keep even if its budget is flat.

None needs a new taxonomy. Each needs the team, product and funding decision in the same model as the cost.

Where to start

  1. Name the cost objects. Decide which teams are products, which are shared services and which value streams they roll up to, using the Consumer layer.
  2. Charge team labor directly. Move persistent teams off project time codes and onto their product.
  3. Keep project codes as attributes. Use them for epics, regulatory work and capitalization.
  4. Fix shared-cost drivers before the first product review. Product owners are likely to test them.
  5. Record funding decisions as evidence. Date, owner and capabilities covered, linked to the capitalization record.
  6. Set the review interval. Pick quarterly or semiannual reallocation and put it on the finance calendar.

An operating model earns its keep here, in our view: when the funding decision, the team, the work and its cost live in one connected model, a quarterly reallocation is a decision with evidence behind it rather than a spreadsheet exercise.

Sources

Frequently asked questions

What is the difference between project funding and product funding?

Project funding approves a scope, schedule and budget for temporary work, usually once a year. Product funding approves capacity for a persistent team or value stream and revisits the allocation at shorter intervals. In the cost model, the main cost object moves from the project to the team or product, and projects become attributes used for reporting and capitalization.

Does product funding make software capitalization harder?

It could under the old stage model, whose phase gates SAFe says don't fit agile work. ASU 2025-06 removes the project stages, effective for annual periods beginning after December 15, 2027. Capitalization then rests on funding commitments, probable completion and resolved development uncertainty. Confirm the treatment for your organization with your accounting advisers and auditor.

How often should product teams be funded?

There is no single standard. SAFe 5, an earlier version of the framework, says value stream budgets are typically updated twice a year. GAO found leading companies reassess business cases and update portfolios at least every six months. In our view, quarterly or semiannual reallocation within an approved annual envelope suits most enterprises, keeping the annual plan while letting the mix move.

Does TBM Taxonomy 5.0 support product-based costing?

Partly. TBM Taxonomy 5.0 renamed the Business layer the Consumer layer and added value streams, products, portfolios and digital services to it, so cost can be reported against products. The TBM Council's migration guide lists agile or product-centric costing using value stream dimensions among future enhancements to plan for, so, in our view, organizations still design that detail themselves.

Free playbooks

In-depth guides for product and operations leaders.

Browse the playbooks
Steve Elliott

Steve Elliott

Published November 22, 2026

More on TBM & IT Financial Management

Cost pools, towers, showback, unit cost.