π PRAGMA · Enterprise Decision Operations Enterprise Decision Diagnostic · 2026
The rung nobody operates

The
DecisionVoid

Every enterprise runs one integrated decision — fusing demand, supply, finance and commercial. It should be continuous, and owned. It's neither. So value drains through the gap, across every lever of the P&L.

100 67 51 ~18 −33 −16 −33 LIST INVOICE POCKET $ KEPT on-invoice off-invoice cogs+serve
THE DRAIN — one lever, illustrated. The pricing decision, unoperated, per $100 of list: invoice runs ~33% below list, off-invoice deductions take ~16% more — you keep a sliver. The same drain runs on demand, operations and capital. This is what an unmade decision costs.
+1%
on any P&L lever → ~8–12%
of operating profit — the price
lever is just one of four
~$2T
drained worldwide each year in
the decision void — demand,
operations, pricing, capital
1
integrated decision that should
fuse all four — run episodically,
owned by no one
The decision layer

The answer exists. The decision doesn't.

A decade of analytics, planning tools and copilots produced the answer. Nothing turns it into a committed, owned, executed call. Strategy advises that rung from above, software feeds it from the side, copilots act beneath it — nobody operates it. So the call is left to a monthly meeting, and the value drains out the bottom.

Every other system stops at “task done.” PRAGMA operates the decision all the way through to “value proven.”

Why the void persists

The tools are bought. The decision is still unowned.

After a decade of analytics, planning platforms and copilots, the bottleneck has moved. It is no longer the answer — it's operating the answer into a committed, accountable decision. That is the void, and it sits on every lever of the P&L.

Every large enterprise runs one integrated decision process that is supposed to fuse demand, supply, finance and commercial into a single committed call. In CPG it's S&OP; in retail, open-to-buy; in banking, ALCO; in marketing, the mix model. Same process, same failure mode: episodic, siloed and unaccountable. It meets monthly while the market moves daily.

The cost surfaces as value left on the table — a volume miss, a price given away, a cost absorbed, capital parked in the wrong place. Each is a decision that was never operated: the signal arrived, an answer existed, and nothing turned it into an owned, executed, measured call. A single point on any of these levers is worth roughly 8–12% of operating profit.

This is structural, not occasional, and it is an operating problem before it is an analytics one. Functions run on conflicting incentives and fragmented data; no role owns the net outcome; exceptions become defaults. The strategy was sound on paper and dissolved on contact with a live market and real stakeholders.

This diagnostic places the void on the decision ladder, breaks it into its four levers, walks the evidence across eight sectors, sizes the prize from global to Brazil, then shows how PRAGMA operates the decision — and a checklist you can run on your own P&L.

What it costs

  • +1% on any lever → 8–12% of operating profit — bigger than most cost programs.
  • ~$2T a year drained worldwide across demand, operations, pricing and capital.
  • >40% of gross-to-net erodes in B2B the first time anyone reconstructs the decision.
  • Worse where it's hard to govern — the leak rate rises in LATAM and Brazil.

Why it persists

  • Episodic — the integrated decision is a monthly meeting, not a continuous operation.
  • Siloed — demand, supply, finance and commercial each optimize locally; the net is no one's job.
  • Unaccountable — no one proves the decision moved the P&L, so no one is on the hook for it.
  • Advised, fed, executed — not operated — consultants, analytics and copilots all stop short of the call.
The model

The void is a missing rung.

Every enterprise is a stack of four layers, ascending in leverage. Three are crowded. One — the decision — is empty. Consultants advise it from above, analytics feed it from the side, copilots execute beneath it. Nobody operates it. That empty rung is the void.

advised, not
operated
StrategyConsultants & the board — set direction, then leave. Advice, not a system.
the void
PRAGMA operates here
DecisionThe call that turns strategy into execution — committed, owned, proven. Nobody runs it.
owned, but
blind upward
ProcessRPA, workflow & planning — o9, UiPath, ServiceNow. Execute; can't decide.
crowded,
commoditizing
TaskCopilots & agents — fast, cheap, everywhere. Do the work, decide nothing.

The decision is literally the bridge between strategy and execution — the rung where a signal becomes a committed call someone is accountable for. It is the highest-leverage layer in the enterprise and the only one no software category owns. Decision-intelligence tools recommend from the side; they don't operate, execute, or carry the outcome.

So the call gets made the old way: in a monthly meeting, by whoever has the loudest function, with no counterfactual and no proof it moved the P&L. Commercial pulls for volume, operations for control, finance for margin — each locally rational, the net globally wrong. The decision falls into the gap between them, and the value drains out the bottom.

The tell is always the same — variance no one can explain: a margin band that widens by rep and season, a forecast that misses in both directions, capital that lands where last year's politics pointed. A void you can't see until someone reconstructs the decision that should have been made.

The evidence · eight sectors

Same void, eight dialects.

Every figure below is a decision that was never operated — a price given away, a promotion that didn't pay, a return no one priced in, a forecast that missed. Different sector, different name, identical mechanism: the integrated call falls into the gap and the value drains out. For each: the dominant leak, the ranked causes, and the headline number.

01

Consumer Packaged Goods

Food & beverage · personal care
home & pet care
Gross-to-net

Volume growth has flattened and price hikes have hit the shopper's tolerance ceiling, so the fight moved into trade spend — the largest and least-governed line on the P&L.

  • Promotions that don't payTrade spend tops 15% of revenue at most CPGs; a large share of events fail to drive incremental volume and cannibalize higher-margin SKUs in the same portfolio.
    >$500B/yr
    global promo
  • SKU proliferation & complexityLong tails of low-velocity SKUs add supply-chain and channel cost with no consumer-backed rationale; exclusives launched without supply-chain alignment raise logistics cost.
    P&G · Nestlé
    cutting 2025
  • Gross-to-net drift & unverified deductionsTrade-terms structures accumulate distortions; heavier discounting triggers more deductions, many never validated against contract.
    exception-
    heavy
  • Reporting that hides the business60% of CPG leaders say financial reporting doesn't match how the business is structured — obscuring visibility and delaying decisions.
    60%
Deloitte: disciplined revenue growth management recovers 3–5% of gross profit a year — most of it already inside the trade-spend line.
02

Retail

Apparel · specialty
omnichannel & grocery
Post-sale loss

The margin drains after the sale — through returns, shrink and late markdowns — and most of it is preventable loss that crosses systems and teams no single owner controls.

  • ReturnsUS retail returns reached ~$850B in 2025, with ~19% of online sales coming back; one return touches order, POS, inventory, finance and fraud systems that frequently disagree.
    ~$850B
    ~19% online
  • Shrink~$112B in 2024. Because shrink is measured on sales but margins are thin, a 2% shrink rate can erase ~40% of net profit; self-checkout leaks ~3.5% vs ~0.2% at staffed lanes.
    2% → −40%
    of net
  • Late, undisciplined markdownsUnsold seasonal stock forces deeper cuts later; delayed markdowns appear as unexplained variance, never as a governed decision.
    value
    decays fast
  • Fraud & cross-channel gapsFraudulent returns and claims ran ~$103B in 2024; buy-online-return-in-store fraud added ~$4B.
    ~$103B
    fraud
Appriss Retail: "retailers that continue to work in silos will continue to erode profits." Returns, fraud and shrink are one cross-functional problem.
03

Banking & Financial Services

Retail & digital banks
lending · payments
Spread & risk

The deal-vs-control tension is underwriting itself: the front office books volume and limits while credit, risk and pricing discipline lag — and the spread leaks through funding cost, provisions and cost-to-serve.

  • NIM compression & repricing asymmetryDeposit costs reprice faster than loan books (elevated deposit beta), squeezing the core spread even when headline rates look favorable.
    deposit
    beta
  • Underwriting discipline erosionWhen growth outruns control, loan-loss provisions climb; borrower-based limits exist precisely to guard against the erosion of underwriting standards.
    provisions
    ↑
  • Price realization & fee leakageNegotiated waivers, fee giveaways and pricing exceptions accumulate account by account with no central view.
    +1% price →
    ~10–12% profit
  • Cost-to-serve & operational dragFragmented systems, manual exceptions and fraud quietly raise the true cost of each customer relationship.
    fragmented
    stack
The math is brutal because the spread is thin: every basis point of leaked price or added provision is disproportionate to the bottom line.
04

Logistics & Supply Chain

Freight · 3PL · last-mile
distribution networks
Cost-to-serve

Strong rates still leak margin when shipment facts don't match billing — leakage forms across thousands of small, reasonable-looking exceptions that no single team owns end-to-end.

  • Accessorial creep & freight-audit gaps10–20% of freight spend is lost to rating errors, duplicate payments, accessorial abuse and contract non-compliance; 68% of invoice discrepancies are rating errors, yet under 15% of firms run systematic audits.
    10–20%
    of spend
  • Detention, empty miles & service failuresDetention alone drains ~$11.5B/yr in lost productivity; overrides create empty miles and missed consolidation; a miss costs twice — the failure and the recovery.
    ~$11.5B
    detention
  • Cost-to-serve never mapped to lane or customerDetention, demurrage and rework sit in separate ledgers, never tied back to a lane or SKU; no team owns total-cost accountability.
    no single
    owner
  • Stale pricing vs cost driftSurcharge tables updated infrequently; volume growth hides that carrier cost and accessorial frequency outran the customer rate.
    rate lag
The field signature: in logistics, margin leakage appears exactly where commercial intent and operational reality drift apart.
05

Healthcare, MedTech & Pharma

Providers · medical devices
pharmaceuticals
Net revenue

Revenue is earned clinically but lost administratively — denials, underpayments and gross-to-net rebates drain the P&L in the gap between care delivered and cash collected.

  • Denials & underpaymentsNet revenue leakage jumped from $38.6B to $48.4B as denials rose; hospitals spent ~$18B overturning denials in 2025 and ~$43B chasing owed payments; Medicare paid ~83¢ on the dollar.
    $48.4B
    leakage
  • Non-labor cost inflationDrug spend up ~9% and supplies climbing; McKinsey projects health systems could shed up to 13 points of margin over five years.
    up to
    −13 pts
  • MedTech & pharma gross-to-netRebates, GPO terms, 340B and chargebacks erode list-to-net; fragmented procurement spend is large — one global pharma recovered €45M by closing it.
    €45M
    recovered
  • Administrative friction as structural costPrior-auth appeals and audit exposure rose sharply, delaying cash and raising bad debt even when providers ultimately prevail.
    friction
    cost
~$48B of net revenue leakage now sits between care delivered and payment received — a denial-and-rebate waterfall by another name.
06

Industrial Manufacturing & Distribution

Components · machinery
distribution channels
Price lag

A "volume-is-king" reflex plus slow price response means input-cost inflation gets absorbed silently and channel discounts compound — until the average margin hides a tail of negative-margin deals.

  • Cost-pass-through lagOne maker took 14 months to notice a 12% input-cost rise because costs were bundled — then needed an 8% price jump to recover the lost ground.
    14 mo →
    +8% needed
  • Channel & distributor discount stackingDistribution pocket margins can run ~20 points below direct, often unnoticed; rebates and terms obscure realized price by route to market.
    ~20 pt
    gap
  • Volume-focused incentivesSales structures reward revenue over net economics, so reps discount to hit volume and price realization lags strategy.
    revenue ≠
    net
  • Off-invoice sprawlRush orders, freight, returns and "last price paid" service policies quietly accumulate beyond the invoice.
    +16% off-
    invoice
McKinsey: a 1% improvement in realized price lifts operating profit ~8% on a typical income statement — bigger than most cost programs.
07

Technology & SaaS

B2B software · platforms
AI-native products
Unit economics

Revenue can grow while profit shrinks — discounting, churn and unmanaged cloud cost erode unit economics beneath an expanding ARR line.

  • Discounting & concession erosionWeak pricing and end-of-quarter concessions suppress realized price; the deal desk is the SaaS pocket-price waterfall.
    deal-desk
    leak
  • Churn & weak net retentionMedian B2B churn sits ~3.5% and median net retention ~103% against a 111%+ target; below-median retention forces costly replacement just to stay flat.
    NRR 103%
    vs 111%
  • Cloud / COGS creep & support cost-to-serveA healthy blended margin can hide top accounts running near 40% on unmanaged cloud usage; payment processing alone is 2.5–3% of revenue.
    40% on
    top cohort
  • Services dragProfessional-services-heavy deployments erode margin as ARR expands — unless tightly productized and standardized.
    unproductized
    PS
The leak hides in plain sight: blended gross margin looks fine while specific customer cohorts quietly run negative.
+1

Professional Services

Consulting · agencies
delivery & SOW work
Scope leak

In services the off-invoice discount is unpriced labor delivered after the SOW is signed — scope creep, free revisions and payment terms quietly erode the realized fee.

  • Utilization below the healthy lineBenchmark billable utilization fell to ~68.9%, under the ~75% threshold considered healthy — overserved, unbilled hours are the hidden discount.
    68.9% vs
    75%
  • Scope creep52% of projects experienced uncontrolled change, up from 43% five years earlier — every absorbed "small addition" is functionally an off-invoice discount.
    52% of
    projects
  • The prestige-account trapThe largest, most prestigious clients often sit at the bottom of the pocket-price band, subsidized by smaller clients paying closer to list.
    15–25%
    quote→pocket
The fix is the same one product firms use: make the leakage visible per account, then govern it — rank clients by realized rate, not headline fee.
The framework

One void, four levers.

Pricing is the leak everyone names — but it's one of four. The same integrated decision sets demand, operations, pricing and capital, and each is a P&L lever where an unmade call drains value. Together they are the four corners of the C-suite. The void opens on all of them at once.

CCO · Demand

Demand

What will we sell, where, to whom — and have we committed to it?

Forecasts that miss in both directions, promotions that don't pay, assortment and channel calls made by gut. The volume decision, run monthly against a market that moves daily.

COO · Operations

Operations

At what cost and service level do we actually deliver it?

Cost-to-serve no one ties to a customer or lane, inventory in the wrong place, expedites and rework absorbed silently. The supply-and-service decision, decoupled from the demand it's meant to meet.

RGM · Pricing

Pricing

What do we charge — and what do we actually keep?

List-to-pocket erosion, off-invoice sprawl, a margin band that widens by rep and season. The price decision, ceded to thousands of daily concessions no one owns. The lever in the hero chart.

CFO · Capital

Capital

Where do we put the money — working capital, credit, investment?

Capital parked where last year's politics pointed, credit and provisions lagging risk, trade-offs settled by whoever argues hardest. The allocation decision, made annually for a position that shifts weekly.

On every lever, the difference is whether the decision is operated:

The void unowned
The decision operated
The call is made monthly, in a meeting, by the loudest function.
Continuous — the decision re-fires whenever the signal moves.
Demand, supply, finance and commercial each optimize locally.
One model fuses all four into a single committed call.
Exceptions are verbal and untracked; "just this once" repeats weekly.
Every call carries an owner, a corridor and a logged reason.
No one can prove the decision moved the P&L — so no one is accountable.
Each decision is attested against a counterfactual baseline.
The scale

How big is the void? Trillions — and it drains harder closer to home.

A directional sizing of the prize across all four levers, cascaded from the world economy down to Brazil. The void costs more per dollar of revenue in markets with volatile currencies, heavier tax complexity and more manual decisions — which is exactly where LATAM and Brazil sit, and exactly PRAGMA's beachhead.

Globalworld commerce
~$110T GDP
~$2T/yr
directional range ≈ $1–3T per year

Triangulated bottom-up from the sector leak rates in this brief — retail returns & shrink, CPG trade spend, freight-audit gaps, healthcare net-revenue leakage, banking & SaaS price erosion — and cross-checked top-down at ~1.5–2% of global business revenue.

leak intensity · ~1.8% of revenue
  • Every sector leaks the same way — the void is structural, not regional.
  • 1% on any lever ≈ 8–12% of operating profit, sitting unowned.
  • Most firms have never reconstructed the decision to even see it.
Latin America~6% of global GDP
~$6.5T
~$200B/yr
directional range ≈ $150–260B per year

LATAM is ~6% of world commerce but leaks at a higher rate. Chronic FX & inflation pass-through lag, some of the world's highest logistics costs, and largely manual pricing & order-to-cash governance widen every band.

leak intensity · ~3.0% of revenue
  • FX & inflation volatility → cost-pass-through lag is permanent, not occasional.
  • Fragmented, partly informal distribution → off-invoice sprawl no one reconciles.
  • High logistics cost & infra gaps → cost-to-serve leaks by lane and customer.
Brazil~⅓ of LATAM
~$2.2T GDP
~$90B/yr
≈ R$480B+ · range $70–110B

The region's largest economy, amplified by tax-driven pricing complexity, "Custo Brasil" logistics, a promo-heavy CPG & varejo culture, pulverized retail and fast-growing e-commerce returns — every leak in this brief, magnified.

leak intensity · ~3.6% of revenue
  • Tax complexity (ICMS-ST, PIS/COFINS, interestadual) → realized price is genuinely opaque.
  • Custo Brasil: logistics + bureaucracy → structural cost-to-serve.
  • Trade-spend-heavy varejo → gross-to-net drift, CPG-style, magnified.

PRAGMA directional model — not published totals. Absolute figures are order-of-magnitude estimates triangulated from the cited sector sources, regional GDP shares (World Bank / IMF, 2025), and observed leak-rate uplifts in emerging markets. BRL converted at ≈ R$5.4/US$ (2026). Read these as the size of the prize, not an audited number — the point is the order of magnitude and the regional gradient: the harder a market is to govern, the more it leaks.

How PRAGMA solves it

Don't advise the decision. Operate it.

PRAGMA doesn't sell a deck or another dashboard. It runs the decision as a continuous loop — binding the commodity agents you already own, adding the few that compound, and proving every call against a counterfactual. You keep the authority; the system carries the outcome.

Bind · connected
Sense
Signals in — your data, BI and market feeds. PRAGMA plugs into what exists; it doesn't rebuild it.
Build · native
Frame
Builds the decision model, generates the options, and sets the counterfactual baseline — what would've happened anyway.
Build · native
Decide
Domain agents propose the call — demand, operations, pricing or capital — with rationale and alternatives.
The gate · your call
Commit
Passes your authority rules; high-stakes routes to a human, then executes through the planning & RPA systems you already run.
Build · PAV
Attest
Proves it: counterfactual + Shapley + cryptographic attestation that this decision — not the market — moved the P&L.
Build · native
Learn
Codifies the call into a Playbook and sharpens the model. Then back to Sense — continuous, not a monthly meeting.
Bind the commodity agents — Sense & Act. Swappable plumbing; never rebuilt. Build the ones that compound — Frame · Decide · Attest · Learn. The moat.

Own what compounds. Bind what's commodity.

Sensing and acting are commoditizing — copilots, planning systems, RPA. PRAGMA never rebuilds them; it operates the vendor-neutral layer above, swapping o9 for Anaplan with nothing else changing. What it builds is the layer no tool owns: framing the decision, deciding the call, attesting the value, learning the playbook. Built from your context, they sharpen every cycle — the consultant's labor curve runs the other way.

No proof, no fee.

Proof is a stage of the loop, not a quarterly report. Every committed decision is measured against the counterfactual baseline framed before it ran; Shapley attribution isolates PRAGMA's contribution; a cryptographic record makes it auditable. That's what makes an outcome fee honest — and what lets a CFO put the decision layer near the P&L at all.

The PRAGMA read

This is Enterprise Decision Operations: PRAGMA operates the decision — the rung between strategy and execution that consultants only advise and software only feeds — while you keep the call. Each operated call is codified into a playbook, so the next one starts sharper. The system is worth more every cycle.

The prize is the ~1% on each lever — demand, operations, pricing, capital — that drops 8–12% to operating profit, already inside your P&L, waiting on a decision no one currently operates.

And it's paid when the P&L moves, proven against the counterfactual — not billed by the hour. Predictability is the moat — and it starts by giving the void an owner.

Give the void an owner

PRAGMA operates the decision — and proves it in your P&L.

A consultancy leaves you a deck; a platform leaves you a dashboard. PRAGMA runs the decision loop across demand, operations, pricing and capital — binds your existing stack, keeps you on the call, and is paid when the P&L moves, proven against the counterfactual.

Book a demo Back to the model

Sources

Findings paraphrased and attributed; figures are directional, drawn from the publications below (2025–2026 unless noted). The pocket-price waterfall in the pricing lever originates with Marn & Rosiello, "Managing Price, Gaining Profit," HBR 1992. The decision-ladder, four-lever and continuous-loop (Sense→Frame→Decide→Commit→Attest→Learn), bind-vs-build, and Proven Attested Value framing is PRAGMA's own Enterprise Decision Operations model.