Store Advertising in Türkiye: Budget, CAC, ROAS and Campaign Profit

سفير تركيا · Monday at 12:03 AM

Scale advertising when order contribution after costs supports it
A limited trial plan with calculations reconcilable to store orders

Illustrative online-shopping image, not attributed to a store or location in Türkiye.

An advertising dashboard may show strong sales while store cash falls because of shipping, discounts and returns. Sound decisions start by defining measured revenue and comparing campaign spend with the contribution its orders retain. This guide builds a practical trial for a store serving the Turkish market, from product choice to stopping or scaling. Every amount, percentage and count in the tables is hypothetical and expressed in Turkish lira, rather than advertising prices or Turkish performance averages. Results represent contribution before fixed expenses, financing and taxes, rather than promised net profit.

Start with the question to test

Write one question before creating the campaign: can advertising a particular product generate retained orders at a cost allowing positive contribution? Define customer type, service area, product and landing page. Advertising the whole catalog immediately makes failure harder to explain: people may respond to the offer but stop at unavailable sizes or late-revealed delivery charges.

Choose a product with shippable stock, clear photographs and a workable return policy. Review contribution before advertising, beyond the difference between selling and supplier prices. Include payment, preparation, store-funded shipping and normal return allowances. Weak contribution means advertising may increase both orders and financial strain.

Separate new-customer acquisition trials from existing-customer retargeting where data allows. Acquiring a first-time customer and reminding a previous buyer are different objectives. State the decision expected from the trial and who makes it, rather than changing the question after an uncomfortable result.

Define CAC, CPA and ROAS before reading the dashboard

Cost per order, CPA, divides expenditure by orders under your chosen definition. CAC concerns new-customer acquisition, so all orders are not a suitable denominator: one customer may buy twice. Specify whether the numerator is media spend alone or a fully loaded acquisition cost including design, tools and agency fees.

ROAS is attributed revenue divided by advertising spend. Values including KDV and shipping differ from tax-exclusive product revenue. Check your actual account configuration; advertising tools accept conversion values, and the displayed figure cannot automatically be assumed to mean profit or net revenue.

In management records, define retained revenue as product sales after discounts, cancellations and refunds, excluding KDV in this example. Do not count the full order amount and a partial payment again. Keep the definition fixed throughout comparisons and record the data extraction date beside it.

Metric	Calculation	What it does not establish
CPA	Advertising spend ÷ orders under the specified definition	New-customer count or profitability
Media CAC	Defined acquisition spend ÷ its new customers	The company’s full cost
ROAS	Attributed revenue ÷ advertising spend	Net profit or causation of every sale
Campaign contribution	Order contribution before advertising − advertising − additional costs	Profit after fixed costs, financing and tax

Set a trial cap your cash flow can support

The trial budget is an amount you can spend while meeting supplier, shipping and refund deadlines. Do not fund a new campaign with payments from orders that may return in two weeks. Reserve for returns and agree a total spending cap with alerts as it approaches. A daily budget does not necessarily mean each final bill follows a rigid daily amount; review your platform’s budgeting and billing mechanism.

Divide the trial into setup, measurement and review. Hypothetically, allocate ten thousand lira to media during a defined test period, plus a two-thousand-lira cap for expected additional costs such as return losses. Record design costs outside this example or add them explicitly. Moving costs between tables can turn an auditable result into a misleading figure.

Stop operation if measurement fails, stock runs out or customer service cannot handle orders. These are direct operational reasons that do not require waiting for the marketing trial to finish. Define the alert owner, daily review time and authority to increase the budget; approval to start is not unlimited spending authorization.

Fix measurement before buying traffic

Complete a test order from a trial advertisement through payment and confirmation, checking that a unique transaction ID, correct value and currency arrive. Refreshing the thank-you page must not create another purchase, and adding to cart must not count as a completed sale. Use the primary purchase event for sales decisions and keep supporting events distinct rather than combining them as one conversion type.

When linking Google Ads and Analytics, review the account, property, permissions and key events rather than only seeing a link in settings. Value and currency are essential purchase fields. Compare an order sample between the store and tool, using dates and values without excessive personal data.

Test successful and failed payments, duplicate orders and cancellation. Maintain a discrepancy log explaining causes such as event delays, consent limits, different devices or incorrect configuration. Do not manually inflate revenue to match a preferred figure or expect every tool to match the store perfectly; reconciliation identifies errors and defines measurement limits.

Illustrative image from the approved e-commerce photographs; advertising success starts with a clear purchasing experience.

Read the conversion funnel before blaming all advertising

Follow impressions, clicks, product views, cart additions, payment and retained orders in sequence. A high click-through rate may reflect a promise the page fails to deliver; strong cart additions with weak payment may suggest unexpected fees, unsuitable payment methods or technical problems. Try the page on a real phone with an ordinary connection.

Clearly state price, size, material, preparation time and delivery coverage. Translate checkout elements and policies into the customer’s language while keeping merchant details and rights understandable. Test discount codes and conditions before promotion; traffic sent to a broken discount wastes budget and increases support requests.

Make one change with a defined reason: an image, sizing description or placement of shipping information. Record the date so a comparison between old and new interfaces does not falsely imply that only the audience changed. Improving page speed or an ad headline alone promises no numerical result; the improvement is a hypothesis needing data.

Convert reported sales into retained sales

In the example, the campaign spends ten thousand lira and records one hundred orders at one thousand lira each excluding KDV, giving initial ROAS of ten. After follow-up, ten orders are canceled and ten returned in full, leaving eighty orders and retained revenue of eighty thousand. Cost per retained order becomes 125 lira and retained ROAS becomes eight.

These figures alone do not judge performance. Assume contribution before advertising of 300 lira per retained order, with two thousand lira of additional campaign cancellation and return costs not already included in that margin. The calculation is 80 × 300 − 10000 − 2000 = 12000 lira of campaign contribution. Avoid including return cost in order margin and deducting it again.

If the campaign has sixty retained new customers, dividing all spend by them gives 166.67 lira as a simplified media measure per new customer. Label it accordingly; do not claim a purely causal acquisition cost when the spend also served previous customers.

Example item	Value	Calculation or explanation
Initial dashboard orders	100	Before cancellations and returns
Retained orders	80	100 − 10 − 10
Retained CPA	125 lira	10000 ÷ 80
Retained ROAS	8	80000 ÷ 10000
Campaign contribution	12000 lira	24000 − 10000 − 2000

Calculate break-even ROAS from contribution

If contribution before advertising is 30% of retained revenue and there are no additional campaign costs outside it, break-even ROAS is 1 ÷ 0.30, approximately 3.33. Under that assumption it covers advertising only, without automatically covering rent, management salaries, finance or tax. Choose a higher target where those items need coverage and calculate them separately.

With a ten-thousand budget and two thousand in additional costs, required contribution before advertising is twelve thousand. Required revenue = 12000 ÷ 0.30 = forty thousand, or ROAS of four. The difference between 3.33 and four comes from a specific cost, rather than a general advertising rule.

Break-even changes with product mix. Selling a high-price, low-contribution product can raise ROAS while reducing retained money. Sum actual product contributions where possible instead of applying one product’s margin to the whole catalog. Otherwise use a conservative assumption and document uncertainty.

Hypothetical retained revenue	ROAS with spend of 10000	Contribution at 30%	After advertising and 2000 additional costs
30000 lira	3	9000	−3000 lira
40000 lira	4	12000	0 lira
50000 lira	5	15000	3000 lira
80000 lira	8	24000	12000 lira

Results are contribution before fixed expenses, financing and taxes; all figures are hypothetical.

Choose a follow-up window that lets orders reach an outcome

Distinguish click, purchase, delivery and last-refund dates. Last week’s campaign may still have undelivered orders, so do not compare it with an older period whose returns are complete. Produce an initial spend-and-orders report and a later retained-orders report, using equally mature cohorts for comparison.

Do not add revenue attributed by two platforms to the same order and count two sales. Attribution allocates marketing credit under settings and windows, which tools may divide differently. Use order records as the reference for actual sales counts, retaining each tool’s perspective for optimization without claiming every sale was caused by one advertisement.

Google Ads supports conversion value adjustments or retractions subject to method and configuration requirements; verify eligibility and deadlines before execution. Use order identifiers and status records and test a small file first. Some retractions cannot be reversed in the same way, so do not submit a cancellation file without financial and operational reconciliation.

Illustrative e-commerce working image; the example figures are hypothetical.

Organize the test so you know what worked

Give each advertisement an identifier, promise, page and target product, and name campaigns so records are retrievable. Do not launch dozens of versions on a small budget and scatter data across unevaluable cells. Start with a manageable number, then change a specific element and record the result and concurrent factors.

Separate unavailable from available products, served from unserved areas and new from previous customers where a lawful, accurate method exists. Review search terms or placements the platform permits you to see and look for unsuitable intent, such as a different product or service you do not provide. Do not exclude terms based on clicks alone when the sample is small.

Record seasons, holidays, price changes, payment problems and unavailable sizes, which may explain results. Offer tests provide weak conclusions if the page, audience and shipping change simultaneously. Describe results modestly as preliminary and requiring follow-up rather than declaring definitive success from a handful of orders.

⁦⁩ Link scaling decisions to money and operations

Scaling increases expenditure under conditions; it is not an automatic reward for higher ROAS. Review retained contribution, measurement quality, cash, stock and preparation time. Sales beyond warehouse capacity create cancellations and complaints that may erase campaign gains.

For a negative result with sound measurement and a complete sample, stop or reduce the losing part and define a realistic corrective hypothesis. If measurement is broken, fix it before judging the audience. If contribution is positive but most orders remain undelivered, keep the result preliminary until it matures.

Raise budgets gradually under a written plan and compare each increase with the additional contribution generated. Historical averages do not guarantee the next order at the same cost. Allocate a limited new-channel test rather than withdrawing the full budget from a proven useful channel. Monthly decisions bring marketing, accounting and customer service into one table.

Review status	Practical action	What to track afterward
Incorrect values or duplicate purchases	Fix measurement and suspend financial judgment	Reconciliation of test orders
Negative contribution after cases complete	Reduce or stop and test one cause	Next cohort’s contribution
Preliminary profit with immature orders	Follow up without aggressive scaling	Delivery and refunds
Positive contribution and available shipping capacity	Limited increase under a new cap	Incremental order cost and contribution

⁦⁩ Balance tracking with customer permission and data protection

Collect only data needed for measurement and the stated purpose. Review legal basis, privacy notice, cookie settings and cross-border transfers with the appropriate specialist. A general accept button does not replace identifying what tools do and who receives data. Keep customer phone numbers, emails and names out of URLs and campaign test files.

Use internal order numbers for reconciliation and restrict report-download access. Separate order updates from retargeting and offers. Do not make marketing consent a condition of payment or purchase completion, or use a verification code concealing several consents in one step; the 2025 KVKK decision emphasized separation, clarity and freedom of choice.

Consent-setting changes can alter visible data volume. Record this rather than interpreting every decrease as weak advertising. Limited measurement requires disclosed limitations, rather than invented purchases or bypassed user choices to fill gaps.

⁦⁩ Frequently asked questions and checklist before further spending

Does ROAS of five mean I earned five times the budget in profit?
No. It means attributed revenue is five times spend under the value definition. Deduct products, fulfillment, returns and advertising, then fixed costs, financing and tax when calculating full profit.

Should I stop after one day without an order?
One day supplies no universal rule. Review the defined loss cap, measurement reliability, visits and purchase-completion timing. Stop immediately for a clear operational problem, and assess the hypothesis with sufficient data without exceeding the cap.

Can future customer value justify high CAC?
Use demonstrated repeat purchases from mature cohorts and contribution after costs, with a conservative scenario. Unproven future purchases do not pay present shipping and refund obligations.

Written definitions of new customer, revenue and retained order.

A test order with correct value, currency and unique identifier.

Spending cap, returns reserve and designated decision owner.

Break-even calculation matching product margins and costs without duplication.

Equally mature cohort comparisons and consistent cancellation records.

Inventory, fulfillment and support ready before increasing the budget.