# Read Your Advertising Results Without Confusing Revenue, Credit and Profit

Connect platform ROAS, blended MER, orders and contribution in one advertising review, with clear definitions, mature comparisons and accountable next steps.

Source: https://dtcup.com/blogs/insights/acquisition-what-to-measure-every-week/

DTCUP / FIELD NOTES / ACQUIRE

# Read Your Advertising Results Without Confusing Revenue, Credit and Profit

Connect platform ROAS, blended MER, orders and contribution in one advertising review, with clear definitions, mature comparisons and accountable next steps.

Published Sep 7, 2026Updated Sep 20, 202613 min read

 ![Jack Hu](https://dtcup.com/wp-content/plugins/dtcup-knowledge/assets/jack-hu-small.jpg)**Jack Hu**7 years of experience building and operating brand websites, specializing in website development, paid advertising and SEO. Built a brand website from launch to US$10 million in annual sales.



  On this page  ## The Short Answer

Keep commercial, channel and customer-journey reporting beside each other. Define the purchase, value basis, currency, dates, exclusions and customer status before calculating ratios. Use platform-attributed revenue for its stated purpose, without adding competing claims of credit into a false total.

Connect reported efficiency to order contribution and customer quality. Label incomplete recent periods, segment where the evidence can support a different decision, and record changes in pricing, stock, campaigns and measurement. End the review with a defined action and owner. A report should clarify whether to keep, adjust, revert or investigate; it should not hide uncertainty behind a single impressive ROAS figure. Define blended MER and its cost boundary explicitly, then use platform ROAS, whole-store revenue and contribution for their distinct decisions; none alone proves additional demand.

## Key takeaways

- Store revenue, attributed revenue and profit answer different questions.
- Write down definitions before comparing ratios or dashboards.
- Allow for delayed outcomes and avoid adding overlapping channel credit.
- Use platform ROAS, blended MER and contribution together without equating attributed credit with additional sales.

The advertising dashboard says sales are growing, the store reports fewer orders, and the business owner feels that cash is getting tighter. These statements can all be true. They may describe different revenue definitions, different dates and different parts of the customer journey.

The answer is not to choose the most flattering dashboard. Build a report that separates what the store sold, what each platform attributes to advertising, and what remains after the costs of serving those orders. Then use each view for the decision it can support.

A useful weekly review should make it easier to decide whether to keep, adjust, investigate or stop an activity. It should not require everyone to pretend that every system counts exactly the same thing.

## Begin with three views of the same business

Keep a commercial view, a channel view and a customer-journey view. The commercial view shows orders, revenue, refunds and relevant costs. The channel view shows delivery, spend and outcomes credited under the platform's rules. The journey view helps explain what visitors did between arrival and purchase.

Each view has limits. The order system knows a transaction occurred but does not automatically know what caused it. An ad platform can credit a conversion without proving that the ad created an additional purchase. Website analytics can describe observed behavior while missing activity affected by consent, browser restrictions or implementation gaps.

Put the views beside each other instead of forcing one to replace the others. If the store's net revenue falls while a platform's attributed revenue rises, the difference is a question to investigate. It is not automatically evidence that either system is broken.

This distinction also helps teams cooperate. Finance, marketing and website operators can use different evidence without arguing that one department owns the only legitimate number.

## Write down what counts before calculating ratios

Define the purchase event, revenue basis, currency, reporting timezone, order exclusions and refund treatment. State whether revenue includes tax, shipping, discounts or cancellations. Record which date the report uses and whether later adjustments can change historical figures.

For acquisition, define a new customer consistently. Is it someone without a previous completed order in the store's available history? How are guest orders matched? Are migrated records complete? An imperfect but documented definition is more useful than an unexplained “new customer” label.

Google's [conversion-window documentation](https://support.google.com/google-ads/answer/3123169?hl=en) describes the period after an advertising interaction in which a conversion can be recorded. That is one reason platform totals depend on settings. Keep those settings visible in the report and record changes rather than treating the resulting numbers as timeless facts.

Also distinguish a purchase count from a customer count. One person can order twice. Dividing spend by orders answers a different question from dividing it by newly acquired customers.

## Use a small set of metrics with explicit meanings

The following measures can form a useful starting report. They need the definitions above; their names alone do not make them comparable across systems.

| Measure | Calculation or meaning | Main use |
|---|---|---|
| Spend | Advertising cost in the chosen period and currency | Size of the commitment |
| Purchase revenue | Revenue under the named source and value definition | Sales associated with the reported activity |
| Purchase count | Orders or purchase events under the chosen definition | Volume and average value context |
| ROAS | Attributed purchase revenue divided by ad spend | Channel efficiency under its attribution rules |
| Cost per purchase | Ad spend divided by counted purchases | Acquisition activity per reported order |
| New-customer acquisition cost | Defined acquisition costs divided by identified new customers | Cost of growing the customer base |
| Click-through rate | Clicks divided by impressions under the report's definitions | Whether the message attracts a response |
| Conversion rate | Defined purchases or converting sessions divided by the matching denominator | Progress through a specific journey |

Do not mix denominators casually. Purchases divided by clicks is not the same measure as sessions with a purchase divided by all sessions. Both may be useful, but a report should say which it uses.

The [website measurement guide](https://dtcup.com/blogs/insights/shopify-ux-what-to-measure-every-week/) develops the journey side. Here the purpose is to connect that behavior to the advertising commitment and commercial outcome.

## Test the business meaning with a simple example

Imagine a fictional campaign that spends $2,000 and reports $8,000 of purchase revenue. Its reported ROAS is 4.0. That ratio does not tell us how much money the business keeps.

Suppose the same revenue, on a consistently defined net basis, requires $3,000 of product cost, $1,100 of fulfillment and shipping cost, $240 of payment fees and $660 of expected returns and service cost. The contribution before advertising is $3,000. After the $2,000 advertising expense, $1,000 remains before fixed overhead on this example's definitions.

Now suppose the product mix changes and the same $8,000 revenue leaves only $1,800 before advertising. The reported ROAS can still be 4.0 while contribution after advertising becomes negative $200. Neither example is a benchmark or an observed DTCUP client result.

This is why a revenue target should be accompanied by cost and customer-quality checks. The business needs to know which orders were earned, what it takes to fulfill them and whether the relationship remains worthwhile after the initial sale.

## Do not add competing claims of credit together

A customer may interact with more than one channel before buying. Each platform may credit the order under its own rules. Adding all platform-attributed revenue can therefore exceed the store's actual sales without creating any extra transactions.

Keep total store revenue separate from the platforms' claims. Use platform reporting for decisions inside the channel, then reconcile broad movement with orders, customer mix and contribution. An aggregate revenue-to-total-ad-spend ratio can provide business context, but it does not isolate the effect of advertising from repeat customers, organic demand or other activity.

If you use the label MER, or marketing efficiency ratio, write the formula beside it. For this report, suppose it means **defined net store revenue divided by total paid-media spend** for the same business scope and period. If your team instead includes agency fees, creative production or other marketing costs in the denominator, give that measure a distinct label and use it consistently. Two reports can disagree simply because their expense boundaries differ.

Compare that whole-business measure with platform ROAS; do not treat them as competing versions of the same number. Google's [conversion-value guidance](https://support.google.com/google-ads/answer/13064207?hl=en) explains its conversion value/cost reporting. Check which conversion actions and values are actually included before interpreting a platform column as purchase-revenue ROAS. A modeled lead value and a completed purchase value should not silently become the same sales claim.

Here is an **original fictional weekly comparison**, using the net-revenue-to-paid-media definition above:

| Item | Amount or calculation |
|---|---|
| Net store revenue | $50,000 |
| Google spend and credited purchase revenue | $6,000 spend; $24,000 credited; ROAS 4.0 |
| Another platform's spend and credited purchase revenue | $4,000 spend; $20,000 credited; ROAS 5.0 |
| Total paid-media spend | $10,000 |
| Blended MER under the stated definition | $50,000 / $10,000 = 5.0 |

The $44,000 of combined platform credit cannot establish that exactly $6,000 came from unpaid demand. Overlapping credit and orders with no platform credit can coexist. Nor does the second platform's higher ratio establish that it deserves the next dollar: it may reach different customers or receive different credit for similar journeys.

Now assume this fictional store keeps $20,000 of contribution before advertising, after the defined variable costs. Paying the $10,000 media bill leaves $10,000 before fixed overhead and any marketing costs excluded from this calculation. If pre-ad contribution were only $8,000, the same MER of 5.0 would leave negative $2,000 after media. The ratio alone has not established profitability.

For a next-budget decision, inspect what changed alongside both measures. Did the team spend more to reach new customers, did repeat demand rise, did a discount reduce margin, or did a stockout constrain a promoted range? Compare mature periods and credible customer cohorts. An improving blended ratio after cutting prospecting can coexist with a shrinking future customer base; that is a possibility to investigate, not a reason to excuse every weak campaign.

Use platform reporting to investigate delivery and credited outcomes within that platform, and the blended view to check the scale of the commitment against the whole business. Use contribution and capacity to assess affordability. When deciding whether extra spend caused extra business, keep that causal question separate; neither ratio answers it by itself.

When a discrepancy appears, first compare dates, currencies, purchase definitions, attribution windows, value fields and duplicate handling. Do not change a working event definition simply to make two totals match.

More precise causal questions require an appropriate experiment or other defensible method. Attribution describes how credit is assigned; incrementality asks what would have happened without the activity. They should not be used as interchangeable labels.

## Give recent results time to become comparable

An order may occur after the advertising interaction, and its reporting or refund adjustment may arrive later still. A report containing recent days can therefore be less complete than one covering older days.

Google's [conversion-delay guidance](https://support.google.com/google-ads/answer/6239119?hl=en) explains how this can make recent cost-per-conversion and return figures appear weaker. Use the observed delay for the account to label incomplete periods. Do not apply a fixed waiting period to every business without checking its buying cycle.

Keep both a current operating view and a mature comparison view. The current view helps detect failures such as a broken page or runaway spend. The mature view is more appropriate for judging the performance of a recent change.

After a meaningful adjustment, review consistent 7-, 14- and 30-day windows where they are useful, with equivalent definitions and notes about completeness. These windows are review points, not automatic proof of causation or statistical reliability.

## Segment only where the difference changes a decision

Separate new and returning customers, important product groups, markets and customer routes when the data supports it. A blended average can hide a campaign moving toward lower-margin products or customers who were likely to buy anyway.

Do not create dozens of tiny segments and rank them as though each estimate were stable. Show counts and spend alongside ratios. A segment with two purchases can have an impressive ROAS that is not a reliable guide to the next large budget allocation.

Use a finding to form a specific question. If mobile conversion declines, inspect whether product selection or payment is failing before reducing every mobile campaign. If refunds rise in one promoted range, review the promise, fit and delivery experience.

The [product-page audit](https://dtcup.com/blogs/insights/product-page-strategy-a-practical-audit/) is a relevant next step when the data points to a product decision problem. Keep the investigation connected to the observed customer route.

## Add the changes that a chart cannot explain

Record promotions, prices, stock, product launches, creative releases, website changes, measurement updates and material seasonal events. Note what changed and when, not merely that the period was “unusual.”

For every advertising adjustment, save the original setting, new setting, object name or identifier, reason, execution date and intended review. This makes it possible to distinguish a recommendation from an implemented change and to reverse the correct item if necessary.

Write a short interpretation beside the figures. For example, a fictional review might say that purchase volume rose during a discount, margin per order fell, and the recent return figures are incomplete. The next decision is to evaluate the promotion's contribution after more outcomes mature, not immediately increase the normal campaign budget.

The commentary should explain uncertainty without hiding behind it. Identify what the team can verify now and what additional evidence would change the decision.

For a website release, add enough detail to connect the change to the people and metrics it could affect. A chart marker that says “new site” is too broad when only one product template changed for phone visitors in one market. Use a release identifier that links to the actual change record.

| Annotation field | What to record |
|---|---|
| Release and effective time | Actual production time, timezone and a stable change identifier |
| Scope | Affected pages, audience, devices, markets and rollout proportion where relevant |
| Before and after | The customer-visible behavior or measurement definition that changed |
| Expected effect | The specific action or commercial outcome being investigated |
| Other changes | Promotions, stock, price, traffic mix or concurrent releases |
| Verification | What was checked, who checked it and the relevant evidence |
| Follow-up | Observation cutoff, decision owner and planned review or rollback condition |

Keep planned, released and reverted states separate. If a staged rollout expands later, record that event too. If a tracking repair changes what can be observed, mark the comparability break rather than presenting newly visible events as newly created customer behavior.

For an original fictional example, a team might release a clearer delivery estimator on selected product pages. The annotation would name those pages and the release time, while the review would inspect whether eligible visitors can obtain an estimate, continue through checkout and receive consistent terms. A simultaneous shipping promotion belongs beside the annotation because it offers another explanation for sales changes.

Do not adjust the review window after seeing which dates produce the best story. Keep the intended comparison, identify incomplete periods and state any justified revision. Retain the original exported view or extraction date so later refunds and delayed events can be distinguished from a changed analysis.

An annotation supplies context; it does not prove causation. A controlled rollout may support a stronger comparison than a whole-site before-and-after chart, but its setup still needs review. When evidence is limited, report the verified change and observed movement without claiming a lift the design cannot establish.

## End the review with an action and an owner

A weekly review needs a small amount of preparation outside the meeting. Have one owner assemble the defined metrics and extraction dates, another explain material commercial changes, and the person responsible for the website report relevant failures or releases. In a small team these may be the same person. Circulate the important differences and unresolved questions so the meeting is not spent opening dashboards.

Start by revisiting last week's decisions. Was the change actually implemented, is the relevant observation window mature enough, and did anything else change? Then review the commercial picture, investigate the few differences large enough to affect a decision, and agree what happens next. Keep a separate list for interesting questions that do not currently change the plan.

A practical decision record contains the evidence, interpretation, action, owner and next review date. For an executed campaign change, retain the exact before-and-after settings and object identifiers. For a proposed action, record it as proposed until someone verifies execution. At the next review, use the same record instead of reconstructing the reason from memory.

For example, a fictional team might see a mobile conversion decline while traffic cost stays stable. The immediate action could be to inspect the affected product and payment route, with a named website owner and a deadline. A separate budget decision can wait for that finding unless there is a clear reason to act sooner. This avoids changing media, creative and the website simultaneously and then being unable to explain the result.

A useful outcome can be to keep the current setup, adjust a defined part, revert a harmful change or investigate an unresolved issue. “More analysis” becomes actionable only when someone knows what to check and when the answer is needed.

If the report shows a tracking failure, restore measurement and document the affected period. If customer evidence reveals a misleading promise, correct it. If the economics are acceptable and the result is mature enough for the intended decision, consider a bounded next commitment with operating capacity in mind.

Do not automate major spending changes from a single unexplained ratio. The report should make the judgment clearer, while the business still owns the trade-off between growth, cost and uncertainty.

## Questions about advertising measurement

### Which dashboard should be the source of truth?

Choose a source for each fact. Use the order and finance records for transactions and commercial definitions, platform reports for platform delivery and attributed outcomes, and analytics for observed journeys. Document differences rather than expecting one dashboard to answer every question.

### Is a campaign with lower ROAS always worse?

No. Customer mix, product margin, attribution rules and the role of the campaign can differ. Compare consistent definitions and the business contribution, then consider the evidence about additional customers. A higher ratio can still accompany a weaker overall result.

### Why did last week's numbers change after we reported them?

Later purchases, reporting delays or adjustments can update a previous period. Save the extraction date, identify incomplete windows and refresh them consistently. Preserve the original report and explain the revision instead of silently presenting changing totals as if they were always final.

Explore the [paid acquisition guides](https://dtcup.com/guides/paid-ads/) for related planning and customer-journey decisions.



 ## Sources

1. [support.google.com — 3123169](https://support.google.com/google-ads/answer/3123169?hl=en)
2. [support.google.com — 6239119](https://support.google.com/google-ads/answer/6239119?hl=en)
3. [support.google.com — 13064207](https://support.google.com/google-ads/answer/13064207?hl=en)

CONTINUE / THE TOPIC

## Keep the next question connected.

[NEXT BEST READ**Which Website Metrics Actually Help You Make Better Decisions?**](https://dtcup.com/blogs/insights/shopify-ux-benchmarks-that-actually-matter/)[RELATED GUIDE**Audit a Product Page Before Paying for a Redesign**](https://dtcup.com/blogs/insights/product-page-strategy-a-practical-audit/)[RELATED GUIDE**Plan Paid Acquisition Around the Business You Can Actually Deliver**](https://dtcup.com/blogs/insights/acquisition-a-practical-audit-framework/)[RELATED GUIDE**Hire a Media Buyer or an Agency? Choose the Operating Model You Can Support**](https://dtcup.com/blogs/insights/in-house-media-buyer-or-agency/)

Still have a question?[Ask the DTCUP Community ↗](https://dtcup.com/community/)

 



  FAQ

## Questions related to this insight.



Clear scope, practical constraints and direct answers for the decision ahead.



Which dashboard should be the source of truth?Choose a source for each fact. Use the order and finance records for transactions and commercial definitions, platform reports for platform delivery and attributed outcomes, and analytics for observed journeys. Document differences rather than expecting one dashboard to answer every question.



Is a campaign with lower ROAS always worse?No. Customer mix, product margin, attribution rules and the role of the campaign can differ. Compare consistent definitions and the business contribution, then consider the evidence about additional customers. A higher ratio can still accompany a weaker overall result.



Why did last week's numbers change after we reported them?Later purchases, reporting delays or adjustments can update a previous period. Save the extraction date, identify incomplete windows and refresh them consistently. Preserve the original report and explain the revision instead of silently presenting changing totals as if they were always final.
