ROAS (Return on Ad Spend) is useful, but it does not tell the whole story. A campaign can look profitable while generating orders that are later returned or canceled, or orders from non-genuine traffic. NetROAS helps measure the value that remains after those outcomes are accounted for.
The limit of ROAS
Traditional ROAS divides the revenue attributed to a campaign by its advertising spend. It shows the revenue associated with the campaign, but has important limitations:
It does not consider different margins between products.
Ignore returns and cancellations.
It does not take chargeback or unpaid orders.
It does not value traffic quality, so it may include fraudulent clicks or poor leads.
In other words, ROAS measures the initial event, but not necessarily the final result for the business.
Why value changes after conversion
An order or lead is never a static accounting element. Their value evolves over time and can degrade for many common reasons:
Order returned or cancelled.
Failed payment or unretired mark.
Chargeback or fraud.
Lead duplicates, fakes or off targets.
invalid calls or no-show bookings.
Google Ads, for example, already supports conversion adjustments, i.e. the ability to update or remove previously recorded conversions to better reflect the real final value.
What is NetROAS
NetROAS is a measure that considers the net value held after removing all the costs and losses that occur after the initial conversion. The formula is simple:
NetROAS = Net value held ÷ Advertising expenditure
The net value held is calculated by subtracting from revenues attributed:
Discounts
Cost of sale (COGS)
Shipping costs
Payment commissions
Packaging and handling
Returns and cancellations
Invalid or suspicious orders
Traffic or customer quality adjustments
This metric is certainly not a net accounting profit, but rather an estimate of the real economic value that the business retains and can attribute to the advertising campaign.
NetROAS: a practical example
Let's imagine a campaign with:
Advertising fee: 1.000 €
Revenue attributed: 5.000 € → ROAS = 5,0
After costs and adjustments:
COGS: 2,400 €
Discounts: 300 €
Shipping: 250 €
Payment fee: 100 €
Returns and cancellations: €800
Chargeback: 150 €
Sustained/non-responsible orders: 300 €
Net value held: 700 € → NetROAS = 0.7
In the advertising report it seems that for every euro spent there were 5, but in reality the business holds only 70 cents.
Why NetROAS improves Smart Bidding
Advertising platforms like Google and Meta optimize campaigns based on the signals they receive. If the algorithm learns only from gross turnover, it will optimize to maximize turnover, although many conversions then turn out to be useless or harmful.
NetROAS, on the other hand, provides a cleaner and more realistic signal, helping the algorithm to learn from real economic results, not from partial or distorted data. This is crucial to avoid wasting budgets on low-quality traffic or fictitious conversions.
This metric is particularly useful in contexts such as:
Ecommerce with variable margins and high return rates (fashion, beauty, electronics, supplements).
Business with mark or postcard payments.
Performance Max and Shopping Campaigns.
Lead generation with integrated CRM.
Lead generation with Value-based bidding.
Sectors exposed to invalid traffic or advertising fraud.
How Adefence builds NetROAS
Adefence sets up a pipeline that:
It acquires the initial conversion and connects it to the advertising click.
It collects data from ecommerce or CRM, including costs and order status.
Monitor order evolution over time, considering returns, chargeback and cancellations.
Detects suspicious or low-quality traffic.
Calculate the net value held by updating the signal to advertising platforms.
In addition, Adefence automatically manages the rollout to this new bidding strategy, without changing everything at a glance. The NetROAS must first be observed, compared to traditional data, used as a secondary signal and finally, set as the primary objective of optimization. The gradual process allows algorithms to recalibrate without destabilizing campaigns.
Move to NetROAS: try Adefence for free
A registered sale is not always a sold sale. And a paid click is not always genuine interest. Traditional ROAS measures what happens in the advertising platform, not always what remains in the business. In an increasingly competitive and complex market, driving advertising strategies towards real value is the key to sustainable and efficient growth.
Adefence is a real advertising signal firewall, which helps campaigns optimize on what business really holds, avoiding misleading waste and optimizations.
Questions and answers
What is the main limit of traditional ROAS?
ROAS calculates the relationship between advertising spending and turnover at the time of conversion, but does not take into account what happens after: returns, chargeback or unexpected costs can drastically reduce the real value of that sale. Two campaigns with the same ROAS can therefore generate very different margins.
Why can the value of a conversion change after it has been recorded?
An order can be made, disputed with a chargeback or generated by an abuse of coupons or fake accounts, events that occur only after checkout. The gross ROAS never updates itself to reflect these changes, remaining anchored to the initial value of the transaction.
What is NetROAS?
NetROAS is the ROAS calculated net of returns, chargeback and other costs related to non-genuine conversions: it is obtained by subtracting from the gross turnover generated by the campaigns the value lost after conversion. It is therefore an indicator closer to the real margin that the company retains.
Why does NetROAS improve the Smart Bidding?
By providing the algorithm a correct value signal for each conversion, the Smart Bidding can concentrate the budget on traffic sources that generate solid and profitable sales, instead of those that produce apparent turnover but intended to be rendered or contested.
How does Adefence build NetROAS for my campaigns?
Adefence integrates data on the actual outcome of orders, such as returns and chargeback, with traffic analysis after clicking, to calculate the net value of each conversion and send it to advertising platforms. In this way campaigns can optimize directly on NetROAS instead of only gross turnover.
Can you give me a numerical example of the difference between ROAS and NetROAS?
In the example listed on this page, a campaign with 1.000€ of expenditure and 5.000€ of turnover shows a ROAS of 5,0, a number that looks great. After subtracting costs, returns and suspicious orders, however, the net value held falls to 700€, for a NetROAS of just 0.7. The advertising report therefore tells a very different story than the real one: it seems that every euro spent generates five, but in practice the business holds only 70 cents for each euro invested.
How often should NetROAS be recalculated?
Typically every month or every quarter, depending on how quickly returns and chargeback come in your industry. Too frequent update is likely to introduce noise in data for events that are still being processed, while one too rare leaves the algorithm to work long on signals that have already been overcome. The right cadence also depends on the volume of orders: the more sales you have, the more often you can afford to update the calculation without losing stability in data.
Is the difference between ROAS and NetROAS equal in all sectors?
Yes, and it is normal that it is so: sectors with high physiological returns, such as fashion and clothing, will of course have a wider gap between ROAS and NetROAS than sectors with rare returns, such as consumer electronics or non-perishable goods. There is therefore no equal value of NetROAS for all sectors: it must be compared over time with the historian of your own business, not with generic market benchmarks that could mislead more than help.
Google Analytics alone is enough to calculate NetROAS?
No, not entirely, even if it remains a useful tool for a first general control. Google Analytics 4 can record repayment events at product level, but does not automatically link this data to the specific advertising campaign that generated the order, nor send them back to the offer algorithm to correct future decisions. It therefore needs an additional level that closes this circle, connecting the actual outcome of the order to both the source of traffic and the bidding system itself.