Value Realization Isn't a Quarterly Reconciliation. It's a Readout of the Work.

Look closely at how a modern strategic-portfolio, funding, or capex module actually produces its numbers. It holds a plan — allocations, business cases, expected value by initiative — and then it reconciles. Actuals are exported from the systems where the work runs, pulled in on a cadence, and compared against the forecast. The enterprise-agile and portfolio vendors roll this up as "cost of work" and forecast-vs-actual burn; the connected-planning vendors model capex and value realization and, to their credit, market it explicitly as protecting against leakage. The output is a clean variance line on a review slide.

But notice what that line is. It is an inference assembled after the period closed, from data polled out of other systems and fitted back to a plan. The value target lives in the model. The work that delivers the value lives somewhere else. The number you present is a re-enactment of a quarter that has already happened — a document about the investment, maintained beside the process that spends it.

Be fair about what these approaches get right, because over-claiming here is the fastest way to lose a finance reader. Forecast-vs-actual modelling imposes real discipline; it forces trade-offs to be named and defended, and a portfolio without it is worse in every respect. Scenario analysis — reallocating capital across a book of bets under different assumptions and seeing where the value moves — is a genuinely hard, genuinely useful capability. And the newer tools that pull actuals through live metric integrations really do cut staleness versus a spreadsheet someone updates by hand once a quarter. None of this is a straw man.

So the critique is not that the model is crude. It is that the model is a separate copy. Interrogate the verb every one of these systems shares: they poll. However continuous the integration, a scheduled read of an external ledger is still a read of an external ledger, pulled into an artifact that sits to the side of the work and has to be reconciled back to it.

This is the whole distinction, and it survives any roadmap. A real-time read of a separate system of record is not the same thing as a value figure that is a query over the same running process that produces the value. In the first case there are two objects — the plan and the work — and a reconciliation forever chasing the gap between them. In the second there is one object, observed at the instant it changes.

The consequence is the difference between detecting leakage and preventing it. A faster integration lets the model notice a starved or off-track bet sooner. It still notices after — after the capital kept flowing to a diverging initiative, after the value quietly stopped accruing, after the variance grew large enough to clear the noise floor of a quarterly read. The sharpest vendors in this category have even conceded the shape of the problem by shipping tighter and tighter linkage layers to shorten the lag between a decision and the actuals. Shrinking the lag is worth doing. It does not change the kind of thing you are looking at: a detection, arriving after the money was already committed.

The alternative is not a faster model. It is not a model at all. Entroid runs every enterprise process as a composition of five primitives — Deterministic Workflows, Intelligence Orchestration, Atomic Agents, Functions, and Connectors — on one Semantic Ontology, in a single runtime. Because the work actually executes on that fabric, its live state is the data. There is no separate copy to poll.

On top of that same running fabric sit the Business Portfolios and Business Planning modules — investments, business cases, value realization. Here, capital is not allocated to a line item in a model that watches from the side. It is bound to the executing primitives — the very Workflows, Agents and Functions that deliver the outcome. Realized value becomes a Function: a query over the state of the process instances that produce it, computed continuously from delivery rather than entered and then reconciled. Progress is a byproduct of the work, not an attestation about it.

Consider a market-expansion bet whose value target is on-time site activations. In a portfolio model, that target is a forecast line, and the truth about it arrives when someone exports activation data from the delivery systems and fits it back to the plan at the next review. On the fabric, the value figure is a query over the very workflows and agents activating the sites. Realized value is emitted by the work. A site that slips does not wait for a review to register as leakage — the number moves the moment the workflow diverges, while the capital can still be redirected. This is an architectural property of computing value from live state, not a claimed result from any deployment.

The same shift changes what distributed planning even means. Business Portfolios realizes value through distributed planning and business proposals — the mechanism by which parts of the enterprise put forward where capital and effort should go. Beside the runtime, a proposal is a spreadsheet allocation: a number in a plan that someone must later prove out against actuals. On the fabric, a proposal is a live claim on real capacity — it draws on the same primitives and the same running capacity the enterprise is actually operating.

• Contention is visible immediately — when two proposals reach for the same capacity, the conflict shows at the moment of the claim, not when the actuals collide a quarter later.
• Funding is enforceable, not aspirational — because the money is bound to the executing primitives, a bet that is starved or drifting off its value target surfaces at divergence, where prevention is still possible, rather than at reconciliation, where only accounting is.

The board question underneath all of this was never "what was this bet worth last quarter." It is "what is it worth now, and can I trust the number well enough to move capital today." A reconciliation answers the first and hopes nobody presses the second. A readout answers both, because the number and the work are the same object.

Two honest caveats, because a CFO will supply them if you don't. First, this is not a claim that Entroid floats free of your estate. It reaches your ERP, ledger, capex and treasury systems through governed Connectors — the one primitive licensed to touch the outside world, with authentication, authorization and audit. The distinction from the connected-planning and portfolio platforms is not integration versus none. It is that the money and the value target execute on the same fabric as the work, rather than being modelled beside it and reconciled to it.

Second, this is not a claim that the other tools are static or that they "cannot" read actuals in real time — the best of them do, and it genuinely narrows the staleness gap. The structural difference is simpler and more durable than any feature comparison: a fast read of a separate ledger is still a read of a separate ledger, and it still leaves a human to close the gap between the plan and the doing. A value figure computed as a query over the same running process has no gap to close, because there is no second copy to reconcile.