Skip to content

The trap

Your best month of profit is already sitting
in your warehouse. It just isn't moving.

Not a metaphor. Wholesale distribution runs on a 2.37% net margin — so every dollar of loss you prevent is worth $42 of revenue you'd otherwise have to go out and sell. The money is already yours. It's stuck.

US$54.7bn

is trapped on the balance sheets of listed Middle East companies right now — cash that could be released. And only 9.4% of them have managed to improve their working capital three years running.

That study measured the Gulf, because that is where it was run. Nothing below is regional — it is arithmetic about holding cost and demand variability, and it holds in Birmingham exactly as it holds in Sharjah.

PwC Middle East Working Capital Study, 2025 ↗

Read that second number again. Nine in ten of your peers are not fixing this — not because they're careless, but because the problem has a shape that ordinary diligence cannot beat. Here is that shape.

Open the live demo — no signup Run your own numbers ↓

What's on this page

Six things, and you don't have to read them in order.

Written for distributors who import on long lead times and carry a deep catalogue — auto and machine spare parts, pharma and medical consumables, cosmetics, packaged FMCG, building materials, electrical, plumbing, industrial spares. Every figure below links to the study behind it, and the last section lists all of them in one place.

  1. 01The cycleFive sensible decisions that assemble into a trap, and how each turn takes more cash off the table.
  2. 02Why trying harder failsHold less and you stock out; hold more and you pay the carry. It is an equation, not a judgement call.
  3. 03Your own numbersPut your inventory and terms in and price the status quo. It costs what it costs today — no saving is forecast.
  4. 04Where it breaksThe one point in the loop that can actually be interrupted, and what has to be true to interrupt it.
  5. 05“I already bought software for this”You did, and it is doing its job. Recording is not the same as noticing.
  6. 06Every number, sourcedEach figure on this page, with its study, its year and its scope — including where a source sells something.

The cycle

Every step is the right decision. Together they're a trap.

Nobody chooses this. It assembles itself out of five sensible moves — and then it turns, and it keeps turning, and each rotation takes a little more of your cash off the table.

  1. 1

    Cash goes onto the shelf.

    You buy stock. That money is now inventory — and about a fifth of a wholesaler's entire balance sheet lives here. It isn't idle: it costs you at least 25% a year just to keep it, in capital, storage, insurance and obsolescence.

  2. 2

    Some of it stops moving.

    Not all of it sells. The demand pattern you bought against has quietly changed, and a slice of that stock is now slow, then dead. It doesn't announce itself. It just sits there, billing you rent every single day.

  3. 3

    So there's no cash for what does move.

    Your working capital is committed — to the wrong SKUs. The fast movers, the ones customers actually ask for, are competing for money that's already spent. This is the turn of the loop nobody sees happening.

  4. 4

    You stock out on your best sellers.

    The most expensive thing in your business is now a sale that never appears in any report. Worldwide, stockouts cost more than twice what overstock does — and when it happens, 31% of customers just buy it from someone else.

  5. 5

    So you over-order. Everything.

    Burned once, you buy deeper on everything — because you can't tell which SKU will betray you next. That's not weakness, it's the only rational response to uncertainty. And it's what puts more cash back on the shelf.

    ↺ and the loop closes — back to step 1, heavier than before

Meanwhile, out the sides:

It goes out of date. A superseded part, a short-dated batch. Neither announces itself — it just quietly stops selling, on a clock that started the day it arrived.
It gets marked down. Dead stock liquidated at auction recovers a median of 9¢ on the dollar. Acting late is barely better than not acting.
The invoices slip. 58% of UAE B2B invoices are paid late. 8% are never paid at all.

Why you can't win by trying harder

Two jaws. Closing one opens the other.

Hold less stock and you stock out. Hold more and you tie up cash and eat the spoilage. Every distributor knows this trade-off intimately — most experience it as a judgement call, a matter of nerve and experience. It isn't. It's an equation, and it has been written down.

Safety stock

= Z × σ(forecast error)

ASCM / APICS — the supply chain profession's own body, 2025 ↗

If you hold less

$1.2 trillion

is lost worldwide to stockouts every year — more than twice the $562bn lost to overstock. And when it happens, 31% of customers simply buy it from someone else. The sale that never happened appears in no report you own.

IHL Group, 2023 ↗ Worldwide out-of-stock study, 71,000 shoppers ↗

If you hold more

≥25% a year

is what that safety costs you — every year it sits — in capital, storage, insurance, obsolescence and shrink. Peer-reviewed and measured bottom-up, not estimated by someone selling software. And where a part supersedes or a batch is short-dated, that toll has a deadline attached to it as well.

Azzi et al., Int'l Journal of Logistics Management, 2014 ↗

Look at what the equation actually says. Your buffer is a tax on not knowing. It scales directly with how wrong your forecast is — so the size of the trap you're in is set by the quality of your forecast, and by nothing else.

Which means there is exactly one way out. Not more discipline. Not a better spreadsheet. Not a tighter reorder policy. A smaller error term.

Halve your forecast error and you halve the buffer you need to hold — at the same service level, with the same customers, selling the same things. That is the entire reason this company exists.

Your numbers

Put your own figures in. The math is public.

Two inputs. Every line below shows the benchmark it multiplied by and links to the study it came from — because a number you can't check is a number you shouldn't believe. Including ours.

Rough is fine. The point is the order of magnitude, not the decimal.

What you pay just to hold it inventory × 25% a year — capital, storage, insurance, obsolescence, shrink Azzi et al., Int'l Journal of Logistics Management, 2014 ↗
—
Likely sitting as excess or dead stock inventory × 15–38%. The high end is a software vendor's own SMB benchmark, so treat it as the ceiling — we lead with the floor Netstock SMB benchmark, 2024 (vendor) ↗
—
What that stock returns if you wait until it's dead excess × 9¢ on the dollar — the median recovery at liquidation auction. This is the cost of finding out late Amplio liquidation auction data, 2023 (vendor) ↗
—
Invoices out there, overdue ~50% of B2B sales on credit at ~47-day terms, of which 58% run late Atradius Payment Practices Barometer UAE, 2025 (credit insurer) ↗
—
…and never paid at all credit sales × 8% bad debt, the UAE average Atradius, 2025 (credit insurer) ↗
—

Now take the most conservative slice of all that

Ignore the stockouts. Ignore the bad debt — we can show you an overdue invoice, we can't make anyone pay it. Assume you fix one thing only: you stop paying rent on the stock that shouldn't be in your building. That is — of excess, at 25% a year, and it is the single least heroic assumption on this page.

Carrying cost you stop paying — a year — the profit equivalent of selling —

Because distribution runs on a 2.37% net margin, a dollar you don't lose is worth 42 dollars you'd otherwise have to sell to see the same dollar hit your bank account. That is the whole reason a small operational fix is worth more to you than a big sales push.

Margin computed from U.S. Census Bureau Quarterly Financial Report, 2023–2025 ↗

Three things we're deliberately not doing to this number. We are not multiplying the trapped cash itself by 42. Freeing a dollar from inventory is a balance-sheet release — a dollar you no longer have to borrow, worth its carrying cost. Real, but not profit. Anyone who multiplies it is selling you something. We are not counting bad debt or lost sales as "prevented" — we surface them; collecting and selling is still your job. And we are not predicting your business: these are published benchmarks applied to two numbers you typed. We haven't seen your warehouse. That's what the 30 days are for.

Where it breaks

The loop only turns because nobody is watching it turn.

Every arc of that cycle is a measurable event with a formula behind it. Narada runs 96 checks against your own data every night — and each one shows its working, so you can verify the number in your ERP instead of taking our word for it.

The arc it cutsWhat it computesWhat you get
Cash stuck on the shelfsteps 1–2
Finds dead and slow stock, prices what it bleeds: trapped × holding rate ÷ 365 per day — and puts a hard cliff on the ones still inside the supplier's return window.
A ranked list of what to clear, and whether to return it, discount it fast, or dump it — plus a drafted return-authorisation email to the supplier.
Stocking out on the winnerssteps 3–4
Prices the exposure in margin, not revenue — days exposed × daily demand × unit margin — because revenue would flatter the number by ~40×. And it counts down to the last day you can still order in time.
A drafted purchase order, MOQ-aware, with the arithmetic shown. This is the one thing Narada will actually send — and only after you press confirm.
The over-ordering reflexstep 5 — the one that matters
Buffer size is Z × σ(forecast error). So the engine attacks the error term: per-SKU forecasts, and a newsvendor order quantity at the critical fractile margin ÷ (margin + holding) — the mathematically correct trade-off between the two jaws, per SKU, every week.
Proof it worked. Forecast Value Add re-runs the week against a naive rolling-average plan and reports the difference in dollars. If our plan didn't beat guessing, it says so.
Money already earned, not yet paidthe third jaw
Reads your overdue invoices and prices the financing drag: outstanding × cost of capital ÷ 365 per day, weighted by how old each one is.
Who to chase, ranked by amount and age — and a reminder email written for you, grounded in that customer's actual invoices. You send it.
The clock nobody setsperishables
Every batch counts down to its own expiry date, valued at qty × unit cost, bleeding holding cost until the day it becomes a write-off.
It surfaces on the board before it's dead — while it can still be sold at a discount rather than recovered at 9¢ on the dollar.

And all of it adds up to one number

What today's inaction costs you. In dollars. Per day.

Every finding above carries its own clock. Narada de-duplicates them — so the same trapped dirham can never be counted twice — and sums them into a single live bleed rate on your board. Not a score. Not a health index. A per-day price on waiting.

See it running on live data →

The obvious objection

"I already bought software for this."

You did. And you were right to. The problem isn't that you're under-equipped — it's that your systems record what happened and stop there. They are exceptional filing cabinets. A filing cabinet has never once told you what to do on a Tuesday morning.

55%

of distributors have bought ERP, CRM, ecommerce and analytics — and never integrated them. The intelligence isn't missing. It's stranded in four systems that don't speak.

Distribution Strategy Group, 2026 · n=233 ↗

84%

of ERP implementations had to bolt a business-intelligence layer on top — because the ERP alone couldn't answer the questions leadership was asking of it.

Panorama Consulting — ERP consultants, 2024 ↗

$450k · 9mo

is the median cost and timeline of an ERP project, per the consultants who run them — a quarter of which land over budget. That is the price of the last time somebody promised to fix this.

Panorama Consulting, 2026 ↗

There's a detail in that ERP report worth pausing on. Panorama — ERP consultants, whose living depends on ERP projects going well — name the top cause of budget overrun as organisations discovering, mid-project, that "their chosen ERP system's native reporting won't meet executive dashboard needs at scale."

That is our entire thesis, written by the other side.

Narada doesn't replace any of it. It reads your ERP overnight, read-only, and does the one thing the ERP was never built to do: decide what matters, explain why, and draft the work. No rip-and-replace. No nine-month project. Connected in about 30 minutes.

The floor

Everything above is arithmetic. This is a promise.

We have been careful, all the way down this page, to never once tell you what Narada will find in your warehouse. We don't know. Nobody who hasn't seen your data does, and the ones who claim otherwise are guessing at you.

So we'll do the opposite of guessing. We'll put a number on it and stand behind it:

In your first 30 days, Narada uncovers at least 10× your monthly fee in recoverable value — $15,000 on Analyst, $25,000 on Planner — or that month is free.

Uncovered means shown to you on screen, with the formula, traceable to your own data. Not value we claim you'll eventually bank. Two conditions, both read off your data rather than your say-so — and they're written out in plain language, with nothing hidden.

Start with the guarantee → Read the binding terms

The calculator shows you the size of the prize. The guarantee is what happens if we can't reach it. You are not being asked to believe the arithmetic — you're being asked to let us run it on your data, at our risk.

The ledger

Every number on this page, and where it came from.

Narada's AI cannot print a figure it can't trace back to a calculation — we call it cite-or-omit, and it's enforced in code, not in a style guide. It would be strange to sell you that and then ask you to take this page on faith. So here is the working.

Things we cut

Researching this page killed four statistics we'd have liked to use. "82% of businesses fail from poor cash flow" traces to a slide, not a study — there is no source. "1% better forecast accuracy = 2.7% less inventory" is quoted everywhere and attributable to nobody. A stockout-switching stat we nearly used turned out to be a survey run by one of our own competitors. And a widely-repeated carrying-cost citation points at a 1995 trade magazine article that no longer exists anywhere.

Each would have made this page hit harder. Each is the kind of thing you'd have found in one search — and then you'd have been right to distrust everything else we told you.

  1. PwC Middle East ↗ 2025

    2025 Middle East Working Capital Study

    Covers: Publicly listed Middle East companies, FY2024

  2. PwC ↗ 2025

    World Working Capital Study 25/26

    Covers: 17,000+ listed companies worldwide

  3. Azzi, Battini, Faccio, Persona & Sgarbossa ↗ 2014

    Inventory holding costs measurement: a multi-case study — The International Journal of Logistics Management, 25(1), 109–132

    Covers: Peer-reviewed; bottom-up measurement across ten companies

    2014. Peer-reviewed and still the primary — the widely-quoted '25–30%' traces back here, not to the vendor blogs that repeat it.

  4. U.S. Census Bureau ↗ 2026

    Quarterly Financial Report (QFR), wholesale trade

    Covers: U.S. wholesale-trade corporations (QFR panel skews to firms ≥$50M assets)

  5. IHL Group ↗ 2023

    Retail Inventory Distortion Study: The Good, the Bad, the Ugly

    Covers: Global retail, 4 regions, 10 segments

    The full IHL report is paid; the figures here are as reported in trade press quoting it. We cite what we could verify, and we don't claim to have read the methodology.

  6. Gruen, Corsten & Bharadwaj (GMA / FMI / CIES) ↗ 2002

    Retail Out-of-Stocks: A Worldwide Examination of Extent, Causes and Consumer Responses

    Covers: 52 studies, 71,000+ consumers, global FMCG retail

    2002 — and still the study everyone cites, because nobody has repeated it at this scale. Summarised by the same authors in Harvard Business Review, 'Stock-Outs Cause Walkouts' (2004).

  7. ASCM / APICS ↗ 2025

    Safety Stock: A Contingency Plan to Keep Supply Chains Flying High

    Covers: The professional body for supply chain; standard formulation

  8. Atradius ↗ 2025 sells something adjacent — read accordingly

    Payment Practices Barometer — United Arab Emirates 2025

    Covers: UAE B2B, all industries, surveyed Q2 2025

    Atradius sells trade credit insurance — their business depends on this number being right, but they also profit from it being frightening. Read it with that in mind.

  9. USDA Foreign Agricultural Service ↗ 2024

    GAIN Report — United Arab Emirates (Grain & Feed Annual; Exporter Guide Annual)

    Covers: UAE food & agricultural imports

  10. Distribution Strategy Group ↗ 2026 sells something adjacent — read accordingly

    State of Distributor Technology 2026

    Covers: 233 North American wholesale distribution executives, Q1 2026

    North America, not the GCC — and DSG advises distributors for a living. Still the only large survey we found that is about distributors specifically rather than 'companies'.

  11. Panorama Consulting Group ↗ 2026 sells something adjacent — read accordingly

    The 2026 ERP Report (and the 2024 edition, for project cost)

    Covers: 170 organizations, median revenue $200.5M, global

    Panorama sells ERP consulting — every incentive to make ERP projects look good. We use their numbers anyway, and we quote the newer, kinder 9-month timeline rather than the older 15.5-month one.

  12. Netstock ↗ 2024 sells something adjacent — read accordingly

    2024 Inventory Management Benchmark Report

    Covers: 2,400+ SMB customers + 300 survey respondents, global

    Netstock sells inventory-optimisation software, so this number flatters their pitch — and ours. We use it because it is the only named, dataset-backed figure scoped to businesses your size. Treat it as the top of the range, not the middle.

  13. USDA Economic Research Service ↗ 2016

    Updated Supermarket Shrink Estimates for Fresh Foods (EIB-155)

    Covers: 2,900 U.S. supermarkets; 24 fruits, 31 vegetables

    2016 publication on 2011–12 data — but measured, not surveyed, and still the only large-sample shrink dataset of its kind.

  14. JPMorgan Chase Institute ↗ 2016

    Cash is King: Flows, Balances, and Buffer Days

    Covers: 597,000 U.S. small businesses, 470M transactions

    2016, and U.S. — but it is a bank's own transaction data rather than a survey, which is why we use it instead of the '82% of businesses fail from cash flow' line you'll see everywhere. That one has no study behind it at all.

  15. Central Bank of the UAE ↗ 2026

    Base Rate maintained at 3.65%

    Covers: UAE policy rate — the floor under any borrowing rate, not a lending rate

If any figure here is stale, mis-scoped, or you think we've read it wrong — tell us and we'll fix the page. That offer is not rhetorical. One wrong number is how a product like this loses the right to exist.