Rabehni is an e-commerce growth partner. We connect the ad to the store, the order, the confirmation, the delivery and the return, so we improve the money that actually lands, not the number the platform reports.
The ad platform sees the first four stops and closes the report. The next four decide whether the month earned or lost.
The report closes at orders and ROAS, and the work is optimised for those four.
A placed order is not money, it is a promise of money.
Confirmation, delivery and returns are what turn the promise into money, and the ad decision is made on that number.
Same month, same account, two completely different numbers. Not because the platform lies, because it measures the promise, and we measure what was collected.
Illustrative Illustrative numbers to explain the maths, not a client's numbers. Contribution margin is 3% of collected revenue, before any fixed costs. Any scaling decision built on the 4.0× spends well above the break-even ceiling. Bars are scaled to 1,000 orders in the first group and to attributed revenue in the second.
Read the full definition of real ROAS →We start with the diagnosis, not the prescription. Find the symptom that describes your account today: the problem is usually not in the column you are looking at.
High traffic, low conversion rate
The ad brings people; the page, the price or the checkout stops them. More spend here buys more visits and more loss.
Good CPA, poor confirmation rate
The order is cheap because it is easy, the ad brings weak intent, or confirmation replies late. An order that never confirms cost full price to acquire.
Good confirmation, poor delivery
The customer said yes and the parcel did not land or came back. That leak carries the shipping cost twice, out and back, and appears on no ads screen.
Strong ROAS, weak contribution margin
Price, cost of goods or stacked discounts eat the margin before the ad even runs. The account can keep looking successful while losing on every order.
Good acquisition, almost no repeat purchase
Every pound buys a customer once, so the spend ceiling is set by the first order alone. Lifting repeat purchase lifts the ceiling without touching the campaigns.
Falling CTR, rising CAC
The one row here whose fix is inside the ads: the message is burned and the angle needs to change, not the targeting.
Five out of six are not fixed in the campaigns. That is why we do not start with a campaign. We start by finding which stop leaks most, then decide where the work begins.
Rabehni Marketing Agency is an e-commerce growth agency working with brands in Egypt and the Gulf on profitability and scale through performance marketing, creative strategy, conversion rate optimisation (CRO), tracking and measurement, and e-commerce strategy.
We work mainly with live e-commerce stores that already sell and want to scale ad spend without losing their unit economics. The number we judge by is not the revenue reported in the ad platform, it is collected revenue minus cancellations, returns, shipping and goods.
We have worked with brands across Egypt and the Gulf: fashion, beauty, jewellery, F&B, home, automotive and restaurants.
Strategy, creative, performance and data in one team, in a written method you can review yourself, so the result repeats instead of happening by luck.
The offer, the audience, the angle and the message are decided and written down before the account runs. An account that starts without that is not spending on growth, it is paying tuition.
Creative moves the number, not targeting. We ship new ads every week rather than every month, and we try several ideas for the same product to find the line that actually stops people. What is not working gets killed early before it eats budget, and what works gets more versions.
The budget, the validation window and the pass and fail thresholds are written before the creative goes live. Winners and losers are decided on a number agreed up front, not on an opinion two weeks later.
No account is handed to a junior after the first month. The people in the first call are the same people running it.
Every marketing and growth service we run for e-commerce stores, ordered the way the money travels. Not a menu: a decision in one layer changes a number in the next.
Managing the accounts on Meta, Google and TikTok, and producing the creative that moves the number, weekly, with more than one idea per product.
The angle that wins the click sets the intent of the visit, and that intent shows up in conversion rate, then days later in the confirmation rate.
The store itself: building and rebuilding, work on the product page, cart and checkout, and the measurement every later number rests on.
Measurement here is not a report, it is what the platform learns from. Optimising on the confirmed order instead of the checkout changes who the platform brings in the first place.
From the moment the order lands to the moment the customer says yes. We run managed order confirmation, and we read order quality and cancellation reasons back into the campaign.
An order that never confirms was paid for in full. Confirmation rate is not an ops number, it is a direct multiplier on the true cost per order.
We are not a courier and we do not run your warehouse. What we do is measure delivery, returns and their true cost, and turn that into a decision on the campaign, the product and the governorate.
Delivery rate sets the acquisition-cost ceiling. An account delivering 62% and one delivering 80% cannot take the same scaling decision, even on identical ROAS.
The company's numbers, not the ad's: unit economics, pricing, the acquisition-cost ceiling and customer value, and the final call on whether scaling makes sense now.
And the decision here loops back to the first layer: the acquisition-cost ceiling dictates the creative, the audience and the product we spend on tomorrow. That is why this is a loop, not a ladder.
Five phases, each with a defined deliverable, and none starts before the one before it closes. How fast you move depends on the account readiness.
We open the account from the inside: structure, Pixel, Conversions API, attribution, creative library and the post-click funnel. Deliverable: a written report of current bleed and quick wins, ranked by impact.
Restructure with clean naming and a clear split between testing and scaling, unified Pixel and server events, and the store, Meta and Google wired into one dashboard. Every critical event is verified manually before anything scales.
We run continuous production on a weekly rhythm: shooting, writing and editing. We try several ideas for the same product and know within days which one works. Every file stays with you, photos, videos and copy.
Every creative enters a strict testing framework: a defined budget, a defined validation window and pass/fail thresholds written before launch. Winners scale, losers are killed and logged in a per-account learning log.
Once 5 to 8 stable winners are identified, structured scaling begins: 20-30% weekly steps with net-profit guardrails, cannibalisation protection, and a weekly review with you.
Not every problem is an ads problem. In three of these six, what fixed the account was measurement, pricing and unit economics, not creative.
A new brand with no history and no working pixel, running entirely on assumptions.
The bottleneck was measurement. Any number before the pixel and Conversions API work correctly is a guess.
Layer 02An account that stopped growing: spend rising, revenue flat, the account bidding against itself.
The bottleneck was account structure. Campaigns bidding on the same audience: spend rises, reach does not.
Layer 01A premium brand attracting discount hunters, with ads opening on price instead of value.
The bottleneck was the message and the offer. Leading with price filters for discount hunters.
Layer 01The account reported 5.48 ROAS while the real figure was 1.80, wrong attribution window, no UTMs, revenue counted before it landed.
The bottleneck was the attribution window. Revenue counted at checkout, not at collection.
Layer 02“Zero orders was never an ads problem. It was a 150-pound problem”, priced 13% above market with nothing to justify it.
The bottleneck was price. Zero orders caused by an EGP 150 gap, not by creative or targeting.
Layer 05A 618-product catalogue with no defined break-even, a hidden discount leak, and prices 27% above market.
The bottleneck was unit economics. No known break-even, and a hidden discount eating margin before the ad runs.
Layer 0514 snapshots from accounts we actually run. The numbers are real and unedited; brand names are withheld to honour NDAs. Tap any card to enlarge.
Most brand owners here get a screenshot at the end of the month. Our clients get a link that reads from the ad platforms and the store directly.
| Platform | Spend | Collected | ROAS |
|---|---|---|---|
| Meta | 310,000 | 760,000 | 2.45× |
| 120,000 | 300,000 | 2.50× | |
| TikTok | 70,000 | 130,000 | 1.86× |
The question the dashboard answers: how much did the business make this month? Not "what was the ROAS?", that is one of the numbers inside it, not the answer.
This dashboard is not a tool we bought, it is a layer of the Rabehni platform itself, which runs orders, shipping and inventory for other merchants. That is why our marketing team talks about delivery and return rates: it is data they work in every day, not a term they borrowed.
The difference is scope, not skill. A media buyer owns the campaigns; a growth agency owns the system the number moves from.
| Dimension | Media buyer | Rabehni, growth agency |
|---|---|---|
| Scope | Running the campaigns | Running the whole growth system |
| Governing metric | In-platform ROAS and CPA | Growth and profit on the delivered order |
| What gets tested | The ads | Offer + creative + funnel + ads |
| Data source | Mostly the ad platform | Ad data joined to the store and the commercial numbers |
| Focus | Acquisition | Acquisition + conversion + unit economics |
We put this before you apply, not after you pay.
Not sure which side you are on? Send the request, if the account is not ready we will say so plainly, and we will not start.
Not a "free marketing audit". It is a diagnostic that opens thirteen points in your store and your account to answer one question: what is limiting growth right now?
The numbers above illustrate the shape of the report, they are not a client result. The content is written on your own numbers, and if the audit finds the bottleneck is not in marketing we will say so.
Answer these so we understand how your store runs, and we will get in touch.
Only three fields are required; the rest is optional.
We will review your store details and get in touch.
The same maths we run inside the accounts: open and free, use it whenever you like.
Max cost per delivered order.
It computes the true acquisition cost after goods, shipping, collection fees and the orders that never land, on a contribution-margin model and a COD-adjusted one.
Input: one order's dataBreak-even after cancellationsSpend ceiling per creativeOpen the break-even ROAS calculator CalculatorWhat did this order actually earn?
It computes net profit on a single order after every variable cost and its share of returns, the number that decides which product can carry ad spend and which cannot.
Input: one product's numbersNet profit per orderShare of returnsOpen the order net-profit calculatorNo. Media buying is one of five layers we work on: acquisition, on-site conversion, order operations and confirmation, delivery and returns measurement, and unit economics with the scaling decision. The difference is that our campaign decisions are taken on the delivered order, not the placed one. Equally, we are not a courier and we do not run your warehouse, we measure fulfillment and turn it into an advertising decision.
In-platform ROAS divides revenue attributed at checkout by spend. Real ROAS divides the money actually collected, after cancellations, failed deliveries and returns, by the same spend. In cash on delivery the gap is wide; in one of the accounts on this page it was 5.48 against 1.80. The second number is the one that governs a scaling decision, because the first one is not money that arrived.
It examines thirteen points: traffic, creative, offer, acquisition, store, CRO, tracking, orders, confirmation rate, shipping, returns, retention and unit economics. The output is a written page naming the primary and secondary bottleneck, where revenue is leaking, the nearest growth opportunity, and priority actions ordered by impact. The example shown on the page is illustrative, your own report is written on your own numbers.
The difference is scope: a media buyer owns the campaigns and the in-platform metrics, while an e-commerce growth agency owns the system the number moves from, offer, creative, store, measurement and unit economics alongside the ads. Our method is written down: strategy is settled before the first pound, pass and fail thresholds are written before launch, and every decision is logged, so you can review the work rather than take it on trust.
No. Platform ROAS measures revenue attributed at checkout, not money collected, and in cash on delivery the gap between the two is wide. In one of the accounts on this page the platform reported 5.48 while the real figure after rebuilding measurement was 1.80. What we measure is collected revenue minus cancellations, returns, shipping and goods.
It is measured on the delivered order, not the placed order. The calculation starts from collected revenue and deducts goods, shipping, collection fees and the order's share of cancellations and returns, then compares that to ad spend over the same window. That is what sets the maximum acquisition cost a product can carry, the same maths as the break-even calculator published on this site.
When the store is already selling and ad spend reaches a size that justifies structured work on it. The minimum stated here is EGP 400,000 a month on ads, with inventory or a supplier that can absorb more demand and a confirmation team replying the same day. Below that the bottleneck is not the ads, so the work does not pay for itself.
The metrics that turn reported revenue into collected money: order confirmation rate, cancellation rate, finalized delivery rate, average order value, goods and shipping cost, and contribution margin per delivered order. Without them ROAS does not tell you whether the account earns or spends.
Yes. The service is declared for Egypt, Saudi Arabia and the UAE. Scope is agreed on the audit call, depending on the ad platform, shipping and collection model in that market.
No. We work with Shopify, WooCommerce, Salla and others. On Rabehni we start faster because the profitability data is already there without setup.
No, and anyone guaranteeing you a number is selling hope. What we do guarantee is that the work is written and reviewable, and that we will tell you plainly if the account is not ready.
Yours, always. We take working access, and if the work stops tomorrow the account, pixel and audiences stay with you.
The first phases go to the audit, the rebuild and fixing tracking; a result you can judge appears after that. The duration depends on the account readiness, and we do not promise a date.
You submit the request and we get in touch to agree scope and price. If it turns out the work will not help you right now we will say so plainly, and we will not start.
Page last updated · edited by the Rabehni team