A campaign that returns $4 in revenue for every advertising dollar can leave less money in the business than one that returns $3.50. The difference lies in the cost of fulfilling those orders.
That changes the budget conversation. The campaign with the higher ROAS may be selling discounted products at high delivery costs. Its less impressive neighbor may bring in fewer sales while leaving more contribution after advertising. Looking only at revenue makes those outcomes difficult to distinguish.
POAS, or profit on ad spend, puts the cost of the sale into the comparison. During Q4, when offers change what an order is worth, that can help marketers make better spending decisions. But it works only when everyone agrees on what “profit” includes, and the bidding system receives a value it can use.
Key takeaways
- ROAS measures the revenue per advertising dollar. POAS adds a defined measure of profit or contribution to that calculation.
- This guide uses contributions before advertising as the POAS numerator. A POAS of 1.0 covers ad spend, but does not necessarily cover overhead.
- Discounts must be reflected once, and the refund treatment must be consistent. Double-counting either distorts the result.
- Google value-based bidding can work with business values beyond revenue. Meta and TikTok implementations should be verified separately rather than assumed to work identically.
- Use recent POAS as a provisional signal while conversions and returns mature. Scale against expected additional contribution, not simply the highest historical ratio.
What ROAS leaves out
ROAS is attributed revenue divided by advertising spend. It answers a useful question: how much revenue was credited to each dollar spent?
It does not explain how much of that revenue remains after the order is fulfilled. Two products sold for the same price can have different production costs. A free-shipping offer can change the economics again.
ROAS can still reflect discounts if the revenue sent to the platform is the actual discounted transaction value. The limitation is that revenue alone does not tell the bidder what it cost to deliver the sale.
An account-wide revenue target therefore deserves scrutiny when product margins vary substantially. A target that works for one part of the catalog may be insufficient for another. Q4 promotions make that difference worth calculating before budgets increase.
The POAS formula, with a clear definition of profit
POAS is often described as gross profit divided by ad spend. The difficulty is that different implementations include different costs. A ratio based only on revenue less COGS tells a different story from one that also accounts for fulfillment and payment processing.
For this guide, we use contribution before advertising:
Contribution before advertising = net revenue − COGS − other variable order costs
POAS = contribution before advertising ÷ advertising spend
Contribution after advertising = contribution before advertising − advertising spend
Other variable order costs may include outbound shipping, packaging, payment fees, and marketplace commissions. Document the costs included in your calculation so comparisons remain consistent.
A POAS of 1.5, or 150%, means $1.50 of contribution before advertising for every $1 of ad spend. After advertising, $0.50 remains to help cover overhead and profit. It does not mean the business earned $1.50 in net profit.
Start with the revenue you can reconcile
Use net revenue after discounts and refunds, with a consistent treatment of taxes and customer-paid shipping. Then subtract costs not already reflected in that figure.
If your order feed already records a $100 purchase as $75 after a discount, do not subtract the $25 again. If refunds have already reduced net revenue, a second refund deduction understates the contribution.
Returns also affect more than the refund amount. A returned item may be resalable, while shipping or processing fees may not be recoverable. Agree on that treatment with finance rather than applying a blanket deduction to every return.
A lower ROAS can leave more contribution after advertising
Consider two hypothetical campaigns. In this example, revenue is after checkout discounts but before the refunds listed below. No cost is deducted twice.
Campaign A generates $10,000 in revenue from $2,500 in ad spend, producing a 4.0 ROAS. Its orders incur $4,200 in COGS, $900 in shipping, $300 in fees, and $500 in refunds. That leaves $4,100 before advertising. Its POAS is 1.64, and its contribution after advertising is $1,600.
Campaign B generates $7,000 in revenue from $2,000 in ad spend, producing a 3.5 ROAS. Its costs are lower: $2,100 in COGS, $500 in shipping, $200 in fees, and $100 in refunds. It also leaves $4,100 before advertising. Its POAS is 2.05, and its contribution after advertising is $2,100.
Campaign B leaves $500 more after advertising despite producing $3,000 less revenue. That is the distinction POAS makes visible.
It is a reason to investigate further investment in B, not proof that B can absorb any amount of additional budget. Its next customers may cost more to acquire. The current ratio describes the spend already observed.
Calculate the advertising break-even before setting a target
Using the contribution definition above:
Advertising break-even ROAS = 1 ÷ pre-ad contribution margin
Express the margin as a decimal. At a 40% contribution margin, advertising break-even ROAS is 2.5. At 25%, it is 4.0. At 15%, it is approximately 6.67.
This is the point at which contribution covers advertising spend. A business that must also fund overhead from those orders needs more room above that threshold.
Why does a discount change the target
Suppose an order sells for $100 and has $60 in variable costs. Its $40 contribution gives it a 40% margin and a 2.5 advertising break-even ROAS.
Now discount it by 25%. Revenue falls to $75. If the same $60 in costs remains, the contribution falls to $15, or 20% of revenue. Advertising break-even ROAS rises to 5.0.
The assumption matters: this example holds variable dollar costs constant. Some costs, such as percentage-based transaction fees, may fall with the selling price. Recalculate from the actual offer rather than applying the example mechanically.
For bundles, calculate the basket's economics. Include every item supplied, even when the promotion describes one as free, and account for the cost of shipping the whole order.
Set a contribution target that the business can afford
A universal “good POAS” is not especially useful. The target depends on what the remaining contribution must support and whether the campaign is intended to acquire customers whose value arrives later.
One planning calculation makes the requirement explicit:
Required POAS = 1 + required contribution after advertising ÷ planned advertising spend
If planned spend is $10,000 and the business needs $5,000 left after advertising, the required POAS is 1.5. That is a planning threshold, not a forecast that the campaign will achieve it.
For acquisition, agree on a payback period. A subscription business may accept a negative first-order contribution if repeat purchases reliably recover the cost. Ground that decision in observed cohort performance and available cash, rather than an optimistic lifetime-revenue estimate.
Prepare the value signal before changing bids
The reporting model and the bidding signal serve different time horizons. Reporting should eventually reconcile to actual order economics. Bidding needs timely information, often before the return window closes.
Calculate an estimated pre-ad contribution when the order arrives. Where returns are material, use a documented allowance based on relevant historical data. Later, reconcile that estimate against actual outcomes. Review whether the estimates consistently overstate or understate particular products.
Keep revenue separately available. Someone comparing platform reports with the finance system should be able to tell whether a value represents sales or estimated contribution.
Before testing, check a sample of ordinary purchases alongside discounted orders and refunds. Confirm order IDs, currency, and cost coverage. A missing-cost order should be flagged under an agreed fallback rule, not silently treated as having a 100% margin.
Calculate sensitive cost data in a controlled backend where possible. A browser tag can transmit a value; it should not become an exposed product-cost database.
How to approach profit-led bidding by the platform
Google Ads: Use a clearly defined conversion value
Google's conversion value guidance recognizes business goals, including profit. A contribution-based value can inform value-based bidding when the conversion setup supports it.
If the selected conversion value is pre-ad contribution, a 150% target means aiming for $1.50 in that value per advertising dollar. The interface may still label the strategy Target ROAS. Document the meaning so the team does not compare it with a revenue-based target.
Avoid counting both revenue and contribution for the same purchase in the bidding objective. Check which conversion actions are selected and preserve separate revenue reporting.
Standard conversion value rules are not a general SKU-margin mapping tool. Their documented conditions include audience, geography, and device. Calculate order-specific contribution in the measurement workflow rather than assuming a margin-tier multiplier is available in those rules.
Google's cart-data reporting documentation also distinguishes gross-profit reporting from a full contribution model. Cart data and matching Merchant Center COGS data are required; additional order costs still need treatment in your own calculation.
Give the bidder time to learn the new values
Google recommends supplying values for at least three weeks or one to two conversion cycles, whichever is longer, before activating value-based bidding in the transition described in its setup guidance.
That creates a readiness requirement, not a promise that every account will be prepared in three weeks. Validate the values and account for conversion delay. Use historical performance on the new value basis to inform the initial target.
If you already bid on revenue values, changing their meaning also changes what the target represents. Plan the transition deliberately rather than replacing the feed while leaving the old target untouched.
Meta: Verify the supported event and reporting setup
Treat profit-based optimization as a measurement project before treating it as a campaign setting. Confirm with your implementation team or measurement provider which value-based optimization options are available in the account and whether the proposed contribution signal is supported.
Do not overwrite purchase revenue simply because another platform accepts custom business values. Establish how the implementation preserves sales reporting and avoids duplicate purchases across browser and server events.
If that setup is not validated, use contribution reporting to guide controlled budget decisions while retaining a reliable bidding signal. A financially useful report can improve decisions even before automated bidding uses the same value.
TikTok: Check VBO eligibility and value semantics separately
TikTok's website value-based optimization documentation describes purchase events with value and currency parameters. Its VBO overview focuses on finding purchasers likely to generate higher purchase value.
That does not make every custom profit-value implementation equivalent to Google's. Confirm the supported setup for your campaign type and integration before replacing purchase values. Check eligibility in the account and test reporting consistency before scaling.
Allocate BFCM budget against the next increment of spend
Margin tiers can help organize a catalog, but they are not a complete allocation strategy. A high-margin item may have expensive acquisition costs or limited demand. A lower-margin product may contribute more overall at a sustainable acquisition cost.
Keep both POAS and total contribution after advertising in the review. Optimizing the ratio alone can favor a small campaign that leaves less money in the business.
For example, a campaign spending $100 at a POAS of 3 leaves $200 after advertising. A campaign spending $1,000 at a POAS of 1.5 leaves $500. The higher ratio and the larger contribution belong to different campaigns.
The next question is what additional spend is likely to return. If an extra $500 produces only $400 in additional pre-ad contribution, that increment loses $100 even if the campaign's overall average remains attractive. Historical results can help estimate this, but do not establish the causal return on the next dollar.
Build a few decision scenarios before the promotion starts. Model a deeper discount or a higher return allowance, and identify when the planned spend no longer meets the business's contribution requirement. Our budget simulation guide provides a framework for working through spending scenarios.
Monitor provisional results without mistaking them for final profit
BFCM decisions happen before every conversion, and the return is known. Label current estimates accordingly.
A low same-day POAS can reflect delayed conversion reporting. An unusually high one can fall as refunds arrive. Use reporting maturity and sample size when evaluating a floor; do not automatically stop a campaign because an incomplete window looks weak.
Separate an operational failure from an economic concern. Broken tracking warrants immediate investigation. A contribution decline on a handful of recent orders may warrant observation before a budget change.
Cross-channel comparisons also need consistent attribution. The same order may receive credit in more than one platform. Do not add those claims together and treat them as separate business revenue. Our guide to comparing ROAS across Meta, Google, and TikTok covers the reporting differences to examine before reallocating spend.
POAS does not resolve incrementality. Even an accurately costed attributed order may have happened without the ad. Where practical, use holdouts or other incrementality tests to assess that separate question.
Where Pixis Prism fits
Pixis Prism supports conversational campaign analysis, performance monitoring, and optimization across connected advertising platforms. It also offers scenario simulation to help teams examine proposed changes.
Those capabilities can support the operating rhythm around a Q4 plan: investigate performance changes, review spend, and assess the available options. Profit-led decisions still depend on the contribution data and definitions established upstream.
Before using any campaign manager for POAS analysis, confirm that the implementation can access the required cost information and distinguish the estimated contribution from revenue. Advertising data alone cannot supply missing fulfillment costs or future refunds.
Bring that requirement into the product evaluation. Ask the team to demonstrate which data is available, how campaign values are interpreted, and what decisions the connected setup can support.
A practical rollout before peak spending
Begin by agreeing on the contribution calculation with finance. Select an owner for cost updates and document refund treatment. Reconcile sample orders before changing the bidder's objective.
Next, observe the proposed values over an adequate reporting period. Compare them with revenue and investigate missing costs. Readiness depends on the platform's requirements and your conversion cycle, rather than a universal conversion-count threshold.
Pilot on a limited, representative scope. Record the change date and, where possible, avoid making unrelated major changes. Judge the pilot on contribution after advertising as well as delivery volume and signal reliability.
Expand only when the measurement holds up. If the setup is not ready before BFCM, use validated contribution reporting to inform budgets while preserving reliable campaign measurement. Peak week is a poor time to discover that the new value feed is inconsistent.