The Channel Ledger

1. 01

   Channel revenue is not channel profit. Every door is judged by what it leaves after its own costs, not by what it sells.

   See §3 · Fully loaded channel contribution

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
2. 02

   Every door has its own P&L. Adding the doors' revenue into one number hides which door pays and which one takes.

   See §2 · Five doors, five P&Ls

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
3. 03

   Commission is a clause in your contract, not a law of the market. Know your commission headroom before you sign, not after you read the settlement statement.

   See §4 · Commission Headroom and the floor test

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
4. 04

   A promotion is paid on every discounted order, not only on the new ones. What counts is the difference against a holdout, not the difference against last month.

   See §5 · Promotions: the increment test

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
5. 05

   Customers measure the app price against the menu board in the branch. Price each door to its cost, and protect the ruler items.

   See §6 · Channel price architecture

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
6. 06

   Aggregator demand is rented, direct demand is owned, and ownership has a minimum scale. An owned channel is not free; its fixed costs need orders to carry them.

   See §7 · From rented to owned demand

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
7. 07

   Delivery is cooked in the same kitchen at the same peak. An order that passes the contract and fails the kitchen is a losing order.

   See §8 · Delivery inside capacity

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)
8. 08

   A delivery radius is a trade area. Where radii overlap, orders move between sister branches instead of growing.

   See §9 · Delivery radius and the Satiety Curve

   Share on LinkedIn (opens in a new tab)Share on X (opens in a new tab)

That is where the ledger starts, and it is the fact on which the Doctrine built its channel rules: each door has its own cost structure, so revenue across doors cannot simply be added and called performance (Doctrine §7). Every door has four questions. What does it cost? Who sets the price? Who owns the customer data? And what is its first risk?

The oldest door, and the one that carries most of the fixed cost: rent, floor space and front-of-house staff. Its own variable costs are small and its fixed costs are high, so its contribution swings with how full the room is. The group sets the price outright, and the menu board above the till is the ruler customers use to measure every other price you charge (§6). The customer data is owned in theory and lost in practice: unless the customer is identified through a loyalty programme, an app or a linked payment method, they pass through and leave no trace. Its first risk is being loaded with all the fixed costs, so that it looks like the weakest door when it is the one carrying the others.

Close to dine-in on price and far from it on cost: it uses the kitchen but not the floor, and it adds packaging. It is the fastest door to grow when its flow is well designed, and the fastest to fall when the peak queue gets long, because a drive-thru customer sees the queue before joining it. Its own costs are packaging, counter labour and payment fees. Its first risk is that its orders migrate to the platforms, orders it would have received anyway.

A powerful acquisition bridge: it reaches customers the group could not reach alone and spares it from building a fleet. In return, the customer relationship, part of the data and part of the margin stay with the platform. Its own costs are set by contract: commission, marketing and visibility fees, brand-funded discounts, refunds and adjustments, and then packaging and payment fees where the group bears them. The group sets the price within each platform's terms.

The market is concentrated. A few large platforms share the market; according to Redseer, Keeta reached about 10% of order volume within four to five months of entering in late 2024.²³ No reliable, comparable published market shares exist, so this volume does not rely on any share split. The right reading of these figures is not a judgement on any platform. It is two questions for the group: how much of your contribution passes through a single door, and what is your plan when terms change, or when the market reorders itself as fast as it did when a new entrant reached a tenth of orders in months? This door's first risk is dependence: the largest platform quietly becoming the growth plan.

The group's own app, website, and phone or WhatsApp ordering. There is no commission, which is why it looks like the cheapest door. It is not free. The group pays a delivery cost on every order, through its own riders or a logistics partner; fixed costs for the app, its maintenance and customer relationship management; and the cost of acquiring the customer the first time. Its decisive advantage is that it is the only door where the group owns the price, the customer and the data together. Its first risk is being built before there are enough orders to carry its fixed costs (§3).

Across all doors, electronic payment is now the rule rather than the exception. A secondary source reports that electronic payments were about 85% of retail transactions in 2025; treat the figure as a signal.⁴ Payment fees are a line in every door's P&L, dine-in included.

Corporate accounts, event catering and the group's products on modern-trade shelves. Orders are large and scheduled, so they can be steered away from the peak, a capacity advantage no other door has. Their own costs are different: credit terms, the cost of a sales team, listing fees and trade discounts. CAC payback here is judged against the Doctrine's limit for trade and corporate accounts, 12 months or less, not the consumer limit (Doctrine §6). The first risk is measuring this door on revenue alone and forgetting that working capital is part of its cost.

The Doctrine's channel contribution formula is the starting point (Doctrine §7), and it is enough to reject a losing order. It is not enough to compare doors, because three things are missing from it: fees that are not called commission, the fixed costs a door needs in order to operate, and the kitchen at peak hour.

Fully loaded contribution per order = Order value × GM% − Order value × (Commission% + Marketing fees% + Brand-funded discount% + Payment & packaging% + Refunds & adjustments%) − Channel labour − Delivery cost per order − Peak capacity charge

Monthly door contribution = Orders × Fully loaded contribution per order − The door's own fixed costs

Three notes keep the formula honest:

• Order value is before VAT and after discount. It is what the group keeps from the selling price before VAT. Each percentage is applied to the base the contract specifies; if commission is calculated on the pre-discount price, the model says so.
• Marketing fees are commission under another name. Visibility fees, in-app advertising and the brand's share of co-funded offers come out of the same order, so they sit on the same line.
• A door's own fixed costs are charged to that door alone: the app to direct delivery, order tablets to the aggregators, the floor to dine-in. Group overhead stays outside channel contribution, as in the Doctrine.

Proposed ruleThe peak capacity charge and its linear displacement convention (0 at 80%, 1 at 100%) are new to this ledger.

An order that arrives at peak in a full kitchen does not arrive alone. It delays or displaces another order, or pushes the branch into pausing orders on the app. That is a real cost that appears on no settlement statement, so the ledger adds a peak capacity charge to the formula:

Capacity charge per order = Share of the door's orders that fall at peak × Displacement probability × Contribution of the displaced order

• Displacement probability follows the Doctrine's Principle 3 rule (§4): zero when peak utilization is below 80%, one when it is above 100%. In between, the group enters its own estimate or uses the ledger's convention: a straight line from zero at 80% to one at 100%.
• The contribution of the displaced order is the contribution of the order that would have been served without this one, usually a dine-in or takeaway order at the same branch.
• The charge falls on the door whose growth is being decided. When you add a platform promotion, or a virtual brand in a full kitchen, the new orders are the ones doing the displacing.

The idea is not mine. Little's law in queueing theory links what is in a system to the arrival rate and the waiting time, and Goldratt's theory of constraints says that an hour lost at the bottleneck is an hour lost to the whole system.⁵⁶ The ledger's contribution is to turn it into a line on the channel P&L.

Most channel decisions are made in a single moment: when a platform offers a new contract, a fee change, a promotion programme or exclusivity. At that moment the CFO needs one number that says how much the order can bear before it loses money. That number is Commission Headroom.

Proposed ruleCommission Headroom is new to this ledger: it rearranges the Doctrine’s floor test and adds marketing fees and refunds.

Commission Headroom = GM% − Group contribution floor% − Payment & packaging%

It is the space, in percentage points of order value, that all platform-linked deductions can occupy before the order falls below the floor. With it, the Doctrine's floor test becomes a single line:

The order passes if: Commission% + Marketing fees% + Brand-funded discount% + Refunds & adjustments% ≤ Commission Headroom

The gap between the two sides is the remaining headroom. This is the Doctrine's own formula (§7), rearranged so that the question that matters at the negotiating table is visible: how many points do I have left? It is calculated at two levels: for the typical basket on each platform, and for each major menu item (§6). When the kitchen runs above 80% at peak, channel labour and the capacity charge are also subtracted as percentages of order value, giving loaded headroom.

Before any number, read the contract and the settlement statement with eight questions, and enter the answers in the model as written, not as remembered:

A contract whose answers you don't know is a contract whose earnings you don't know.

When an order, item or promotion fails the test, the question is not "do we continue?" but "which variable caused the failure?" Each cause has its response:

The rule comes from the Playbook (Gate 7): an order type that fails the floor test is repriced, redesigned or stopped, never grown. And a platform's proposal for exclusivity, a fee change or a promotion is accepted only after the floor test.

Every promotion report compares sales before and after. The question that changes the decision is the difference between what happened and what would have happened without it. That is the increment, and the second episode of Beneath Marketing explained it with the Friday queue (Episode 2). This is its accounting version.

The Doctrine's rule (§7): fund a promotion only when incremental orders × contribution per order exceeds its cost across all promoted orders, including those that would have happened anyway. As a calculation:

Net promotion effect = Incremental orders × Fully loaded contribution per undiscounted order − Discounted orders × Order value × Brand-funded discount% − Fixed promotion costs

Two numbers from it are enough for the board:

• Increment ratio = Incremental orders ÷ Discounted orders. How many new orders you bought for each order whose discount you paid.
• Break-even ratio = (Order value × Discount% + Fixed costs ÷ Discounted orders) ÷ Fully loaded contribution per order. If the increment ratio is below it, the promotion loses money, whatever the chart looks like.

A branch group has something internet companies do not have easily: similar branches. A simple design is enough:

1. Matched pairs. Choose pairs of branches that are close in sales, neighbourhood type, city and door mix: one branch runs the promotion, the other does not.
2. The same weeks. Run the promotion and the comparison over the same period, so that Ramadan, payday, holidays and weather fall on both groups.
3. A matching pre-period. Measure the natural gap between the groups for a period as long as the promotion itself, then calculate the expected result without the promotion: test-branch orders before × (control-branch orders during ÷ control-branch orders before).
4. A post-period. Watch both groups for the same length of time after the promotion ends. If the test branches fall below expected, part of what looked like increment was demand pulled forward, and it is subtracted.
5. Non-overlapping radii. Never put a branch whose delivery radius overlaps a test branch into the control group, or orders will move between the two groups (§9).

The calendar is not a detail here. According to Arab News, citing Saudi Central Bank data, food and beverage spending rose 74.9% in the week before Ramadan 2025 compared with the week before, and restaurant and café spending fell 28.3% in one week of Ramadan. Both figures are treated as signals.⁷⁸ Any promotion read on a before-and-after basis in those weeks is read wrongly.

From foreign research we take the lesson, not the numbers. Lewis and Rao found, across 25 large advertising experiments, that the natural noise in sales is far larger than the effect of advertising,⁹ and Blake, Nosko and Tadelis found that switching off eBay's branded search ads barely changed traffic.¹⁰ I could find no published Saudi study that measures the increment of restaurant promotions against a holdout, so the ledger offers no Saudi figure for promotion effect. It offers the method to measure your own.

The bigger risk in promotions is not one losing promotion. It is a "successful" promotion repeated every season until the discounted price becomes the price the customer remembers. At that point the group is no longer buying new orders; it is paying to stand still. This is subsidised habit: a customer who repeats, but only with a discount.

It shows up in two indicators, read together every month for each platform:

• The discounted share of the door's orders, and its trend.
• The increment ratio in the latest holdout test.

If the first rises while the second falls, the group is subsidising a habit, not building demand. The response is not to stop every promotion at once but to run an exit test: switch the promotion off in half of the pairs and measure what remains.

The Doctrine says: price by channel where contracts allow (§7). This chapter answers the question that comes next: which items, by how much, and without the customer feeling that something was taken from them behind their back.

Customers do not judge the app price on its own; they set it beside the menu board they remember from the branch (Episode 1). The idea belongs to others. Kahneman and Tversky found that people judge numbers against a reference point, and that a loss from it hurts more than an equal gain pleases.¹¹ Thaler found that customers pay for the item and for the feeling that the deal is fair in that place.¹² Simon and Nagle start pricing from value to the customer at that moment, not from cost alone.¹³

The gap between doors is real, visible to customers and reported in the press. In a 2023 Al-Eqtisadiah survey of 60 meals from 10 restaurants, app prices were about 22–24% higher than branch prices on most of the platforms surveyed, and about 1% higher on one. Those are that survey's per-platform figures, not a market average.¹⁴ In January 2026, Argaam reported gaps ranging from about 20% to more than 80% by meal and platform, treated here as a signal.¹⁵ Argaam has also reported that the General Authority for Competition is monitoring delivery apps.¹⁶ The conclusion: there is no single number to measure against, the gap is visible, and so it has to be designed rather than left to a spreadsheet.

Every item has a floor price in every door, below which it falls under the group's floor:

Item floor price in a door = (Item cost + Item labour in that door) ÷ (1 − Deduction rates in that door − Group floor%)

The deduction rates are commission, marketing fees, payment and packaging, and refunds, as percentages of the door's price. An item survives at the branch price if its floor price is below the branch price. The result splits the menu into four groups:

1. All door pricing starts with reading the contract. Question 5 in §4 comes before any change.
2. The basket has to pass, not every item. But the typical basket on each platform is tested monthly, because the item mix changes.
3. Ruler items are touched last. If you cannot explain the gap on one of them to a customer standing in front of you, in one sentence, the gap is too big.
4. Name the gap where terms allow. "Delivery prices" is clearer than a silent gap the customer discovers alone.
5. The direct-delivery gap is never higher than the platform gap. Otherwise you punish the customer who came to you directly.

Aggregator demand is rented; direct demand is owned (Doctrine §7). Rented demand is not bad; it is the fastest way to win a new customer. But rent is paid on every order, including the tenth order from the same customer. That is why the Doctrine says: aggregators to acquire, owned channels to retain.

The Channel bridge is one of the Bridges Strategy™ bridges: it moves aggregator customers to owned channels, within each platform's rules (Doctrine §6). It is judged on the conditions for any bridge: CAC through it below 60% of existing CAC, conversion at 80% or more of baseline, payback within the channel threshold, margin after the bridge above the group floor, and the ability to track it and switch it off.

The bridge is not built on every customer, only on those who chose you twice: repeat customers. The tools are familiar and are all used within each platform's terms: a loyalty programme in the branch that identifies dine-in and takeaway customers, a benefit in the app that exists nowhere else (an item, a reservation, faster pickup), and the brand's presence in its own channels. The aim is not to pull customers away from the platform. It is to give the customer who already knows you a direct route to you.

Proposed ruleThe Demand Ownership Ratio and its 90-day definition are new to this ledger.

Demand Ownership Ratio = Repeat orders through owned channels ÷ All repeat orders, over the last 90 days

To be comparable month after month, it is defined precisely:

• Owned channels: the doors where the group holds the customer record, the means to contact the customer, and the price: app, website, phone and WhatsApp, dine-in and takeaway when the customer is identified through loyalty or an account, and B2B accounts.
• Repeat order: an order from an identified customer who placed another order with the group, through any door, in the 90 days before it.
• Denominator: repeat orders in owned channels, plus repeat orders on platforms as shown in the platform's merchant reports where available. Where they are not available, they are estimated by a written method and the ratio is marked "estimated".
• Exclusions: orders from unidentified customers are excluded from both numerator and denominator. Report the identification rate alongside the ratio: identified orders ÷ all orders. A high ownership ratio on a low identification rate means you are measuring a small part of your customers.

The ledger sets no target for the Demand Ownership Ratio, because it varies by format and city. It sets a direction rule: if the ratio falls for two consecutive quarters while the platforms' share of orders rises, the group is growing on rent.

The Doctrine's retention metric is the 90-day repeat rate for each cohort of new customers (§8). The ledger splits it by door: of this month's new customers in each door, how many ordered again within 90 days, and through which door? The resulting table answers two questions. Which door brings customers who stay? And which door hands its customers to another? A door whose customers come back only with a discount is buying orders, not customers.

Before funding a programme to move customers to direct delivery, calculate its payback as you would for any acquisition:

Monthly gain = Migrated customers × 90-day retention × Monthly orders per customer × (Fully loaded contribution in direct delivery − Fully loaded contribution on the platform) − Additional fixed costs

Payback = (Migrated customers × Migration cost per customer + Set-up cost) ÷ Monthly gain

It is read against the Doctrine's limit for CAC payback in consumer channels: 6 months or less (§6). If the contribution difference is zero or negative, because direct delivery costs more than the platform's deductions at your current volume, nobody is moved yet.

The Doctrine's fifth channel rule: delivery orders compete with dine-in for the same kitchen at the same peak, and they belong in U (§7). The third episode of Beneath Marketing tells what happens when that rule is forgotten: the campaign succeeds and the branch collapses (Episode 3).

U = (Dine-in + takeaway + delivery orders in the peak window) ÷ (Kitchen velocity in orders per hour at standard × Hours in the peak window)

The decision comes from Principle 3 (Doctrine §4). Below 80% there is room to grow. At 80%, capacity is added or demand is redirected before any new campaign, promotion or platform. Above 100%, demand generation into that branch stops, including platform promotions and paid visibility. One sign of crossing the line is the minutes spent in "busy" mode on the apps, a form of revenue leakage in the Doctrine.

In Ramadan the peak map itself changes: two peaks instead of one, before iftar and at suhoor. U is calculated for each window separately during the season, and the ordinary months' calculation is not carried over.

A virtual brand, a menu sold on the platforms from an existing kitchen, is not a marketing decision. It is a decision about how to use kitchen capacity. It is accepted on three conditions: most of its orders fall in hours when U is below 80%; it does not push peak U above 80%; and its contribution is measured after deducting the orders it takes from the parent brand on the same platforms. A virtual brand that sells to the same customer at the same hour adds a name, not an order.

A dark kitchen, a delivery-only kitchen, is a new unit with capex, so it goes through the full opening gate (Doctrine §5): satiety, payback, capacity signal and replication. The capacity signal is clear here: nearby branches full at peak, or delivery orders from the target district going unserved. Its cannibalization is estimated from the overlap of its delivery radius with existing branches (§9). A dark kitchen can itself be a Location bridge: a test of demand in a district before a full branch lease is signed (Playbook, Gate 2).

For dine-in, a trade area is a walking or driving distance. For delivery, it is the circle drawn by the delivery radius. The Branch Satiety Curve applies to the circle just as it applies to the neighbourhood (Doctrine §5).

When two branches' delivery radii overlap, both can serve the same addresses. Orders then move between sister branches instead of growing, and the move shows up in reports as growth at the new branch and an "unexplained" decline at the old one. That is why cannibalization is estimated from the group's own data before any opening or radius extension (Playbook, Gate 2):

Overlap share = Existing branches' orders from addresses inside the new radius over the last 90 days ÷ All of those branches' delivery orders over the same period

The overlap share feeds the estimate of C_cannibal in the S_b formula, and cannibalization is re-measured at months 3 and 6 after opening (Playbook, Gate 4).

1. Every delivery radius is mapped alongside its sister branches' radii before any opening or extension.
2. Extending a radius is a decision like an opening: its effect on sister branches and on peak U is calculated, and it is never taken to lift revenue alone.
3. Narrowing the radius at peak is a legitimate capacity tool when the branch is above 80% and a nearby sister branch has capacity.
4. A new branch is never judged on its own sales, but on the sales of the branches it overlaps, before and after opening.

What the board does not see per door, it cannot govern. The fifth metric on the Doctrine's dashboard is contribution by channel (§8). This chapter turns that metric into a monthly P&L and a single board page.

One column per door, a separate column per platform, and the same lines for all:

1. Orders, and average order value before VAT.
2. Net sales.
3. Gross margin.
4. Commission, marketing fees, brand-funded discounts, refunds and adjustments.
5. Payment and packaging, channel labour, delivery cost.
6. Peak capacity charge, for branches above 80%.
7. The door's own fixed costs.
8. Fully loaded contribution, in riyals and as a percentage, against the floor.
9. Remaining Commission Headroom, per platform.
10. Increment ratio of the latest promotion, and the discounted share of orders.
11. 90-day repeat by door, Demand Ownership Ratio and identification rate.
12. The largest platform's share of orders and of contribution.

The source for these numbers is the actual settlement statements and the books, not platform dashboards alone and not estimates.

• CFO: owns the definitions (order value, fully loaded contribution, the floor) and the reconciliation of settlement statements to the books.
• CRO: owns the doors, pricing, promotions and the Demand Ownership Ratio, with one relationship owner per platform reporting to them.
• COO: owns peak U, delivery radii and the decision to go "busy".
• CEO: owns the ceiling on dependence on any one platform, and the exclusivity decision.
• Board: approves the contribution floor and the dependence ceiling, and reviews the channel page quarterly.

A metric with two owners has none, and a platform with two owners is negotiated with in two voices.

One page, six lines, read in five minutes:

1. 01Commission creep

   Symptom

   Revenue flat or rising; contribution slipping without a clear decision

   Metric that detects it

   Remaining headroom per platform, monthly, from settlement statements

   First move

   Rebuild the actual deductions with the eight questions (§4)

2. 02Promotion dependency

   Symptom

   Orders drop whenever a promotion stops

   Metric that detects it

   Discounted share of orders and increment ratio

   First move

   An exit test in half of the branch pairs (§5)

3. 03Ghost growth

   Symptom

   Platform orders rise; the group's total orders in the same trade area stay flat

   Metric that detects it

   Orders and contribution across all doors per trade area, before and after the growth

   First move

   A P&L by door for each trade area, not for each platform alone

4. 04Radius overlap

   Symptom

   A new branch grows in delivery; a nearby branch declines for no visible reason

   Metric that detects it

   Overlap share, and measured cannibalization at months 3 and 6

   First move

   Map all radii on one map and recalculate S_b (§9)

5. 05Data blindness

   Symptom

   Nobody knows who the repeat customers are or where they order from

   Metric that detects it

   Identification rate and Demand Ownership Ratio

   First move

   Identify customers at dine-in and takeaway before any direct-delivery campaign (§7)

6. 06Price-parity trap

   Symptom

   App price equals branch price; platform contribution is below the floor

   Metric that detects it

   Number of items and baskets that fail at the branch price

   First move

   The survival test for every item, then channel items, within each contract's terms (§6)

The sixth failure mode has a mirror image that is just as dangerous: one large gap across the whole menu. Parity eats margin; a blind gap eats trust. The fix in both cases is the same: price architecture, item by item.

Figures as reported by the client groups, for the periods stated where stated. None of them is presented as having used this ledger's tools in their current form. None was measured against a control group.

Five tools accompany the ledger. Every input is your own number; none contains a ready-made commission rate or market benchmark:

1. Channel P&L builder: fully loaded contribution for every door and platform, side by side (§3 and §10).
2. Commission Headroom and floor test: remaining headroom per platform and per item, and the effect of any proposed change in terms (§4 and §6).
3. Promotion increment calculator: the holdout result, increment ratio, break-even ratio and net effect (§5).
4. Direct-migration payback: monthly gain and payback, with the owned channel's minimum volume (§3 and §7).
5. Delivery-in-capacity check: peak U across all doors, the effect of a proposed promotion or virtual brand, and the capacity charge (§8).

1. 1
   Week 1 — Truth. Collect every platform contract and the last three months of settlement statements. Agree one definition of order value and of the floor. Output: a table of actual deductions per platform.
2. 2
   Week 2 — Doors. A P&L for every door and platform, then for every branch; Commission Headroom for the typical basket. Output: the list of doors and platforms below the floor.
3. 3
   Week 3 — Prices and promotions. The survival test for the top 30 items by sales; the ruler items identified; the last three promotions recalculated across all discounted orders; the first holdout designed. Output: a door-by-door price plan and a promotions calendar conditional on testing.
4. 4
   Week 4 — Capacity, ownership and decisions. Peak U across all doors for the highest-volume branches; the delivery-radius map; Demand Ownership Ratio and identification rate; the first board page. Output: a board memo with the renegotiate, reprice and reduce-exposure decisions.

Start with the Branch Gate on your own numbers: five numbers about the site and a three-word verdict, sign, sign with conditions, or not yet. Then every engagement starts with a 30-minute Scale Review. If there is a fit, the 30-day channel audit is delivered within the Branch Expansion Diagnostic. Groups that want continuing support move to a retained Revenue Engineering Advisory.

Try the Branch Gate on your numbers

Book a 30-min Scale Review

1. Zakat, Tax and Customs Authority: VAT at 15% since 1 July 2020. https://zatca.gov.sa/en/RulesRegulations/VAT/Pages/default.aspx
2. Rest of World, 2025, on Keeta's entry and delivery-platform shares in Saudi Arabia. https://restofworld.org/2025/delivery-app-keeta-hungerstation-saudi-arabia/
3. Redseer, 24 February 2025, on Keeta reaching about 10% of Saudi order volume within months of entry. https://redseer.com/articles/keeta-has-gained-10-market-share-in-ksa-whats-next/
4. Signal (secondary source): Voxire, on the electronic share of retail transactions in 2025. https://voxire.com/blog/mada-payment-conversion-saudi-2026/
5. John D. C. Little, "A Proof for the Queuing Formula: L = λW", Operations Research 9(3), 1961. https://doi.org/10.1287/opre.9.3.383
6. Eliyahu M. Goldratt & Jeff Cox, The Goal, North River Press, 1984.
7. Signal (press report of official data): Arab News, citing Saudi Central Bank (SAMA) data, week before Ramadan 2025. https://www.arabnews.com/node/2592495/business-economy
8. Signal (press report of official data): Arab News, citing SAMA data, a week during Ramadan. https://www.arabnews.com/node/2634644/business-economy
9. Randall A. Lewis & Justin M. Rao, "The Unfavorable Economics of Measuring the Returns to Advertising", Quarterly Journal of Economics 130(4), 2015, pp. 1941–1973. https://doi.org/10.1093/qje/qjv023 . US digital-advertising figures, cited as evidence for the idea, not as a Saudi number.
10. Thomas Blake, Chris Nosko & Steven Tadelis, "Consumer Heterogeneity and Paid Search Effectiveness: A Large-Scale Field Experiment", Econometrica 83(1), 2015, pp. 155–174. https://doi.org/10.3982/ECTA12423 . A US eBay figure; not transferable to the Saudi market.
11. Daniel Kahneman & Amos Tversky, "Prospect Theory: An Analysis of Decision under Risk", Econometrica 47(2), 1979, pp. 263–291. https://doi.org/10.2307/1914185
12. Richard Thaler, "Mental Accounting and Consumer Choice", Marketing Science 4(3), 1985, pp. 199–214. https://doi.org/10.1287/mksc.4.3.199
13. Hermann Simon, Confessions of the Pricing Man, Springer, 2015; and Thomas Nagle, Georg Müller & Evert Gruyaert, The Strategy and Tactics of Pricing, 7th ed., Routledge, 2023.
14. Al-Eqtisadiah, 31 March 2023: a survey of 60 meals from 10 restaurants comparing app prices with branch prices. The percentages are per platform in that survey, not a market average. https://www.aleqt.com/2023/03/31/article_2522281.html
15. Signal (press report): Argaam, January 2026. https://www.argaam.com/ar/article/articledetail/id/1875844
16. Signal (press report): Argaam, on the General Authority for Competition monitoring delivery apps. https://www.argaam.com/en/article/articledetail/id/1875921