ROAS Calculator: Inputs, Scenarios and Decision Rules

Take delivered media cost from the report for the selected campaign and period. A deposit funds the account; it is not automatically the amount spent during the test. A bid is a buying setting; it is not the final campaign bill. Keep credits and later charge adjustments visible so the denominator can be reproduced. If you also want to include agency or creative expense, calculate a separately labelled total-cost view. Mixing that broader cost into one source report while leaving it out of another makes their ROAS values difficult to compare.

Choose the sales or conversion-value record that answers your buying question. For a shop, that may be attributed order revenue after the agreed reversals. An affiliate buyer should use the commission they are entitled to receive, rather than the merchant's entire basket value. A lead campaign can use an assigned estimate, but the output must say value-based ROAS because an estimate is not collected revenue. Keep pending approvals separate from accepted amounts. We recommend a named value source and an adjustment policy before the first comparison, so a larger dashboard number does not quietly replace a more relevant business measure.

Recalculate the numerator when an eligible order is refunded or an affiliate commission is reversed. In a hypothetical campaign, USD 1,500 of initially attributed revenue on USD 500 of media gives 3.0x. If USD 150 is later reversed, the adjusted revenue is USD 1,350 and adjusted ROAS is 2.7x. Retain both dated observations so the change is understandable. A refund policy also needs a cutoff: otherwise an older source can appear worse simply because it has had more time to receive reversals. Your operating report should show the same adjustment treatment for comparable groups.

Label recent results provisional when customers can still convert after the reporting cutoff. Compare campaign groups at a similar age since acquisition, rather than giving last month's visitors more time to buy than yesterday's. Record an initial observation and a later update using your actual buying cycle. Google Ads' Target ROAS documentation, reviewed on 25 October 2026, also discusses conversion delay within its own bidding context. Its account requirements do not transfer to FroggyAds. The useful principle for your worksheet is to retain the delay and observation scope instead of declaring an incomplete result final.

Convert revenue and cost to one reporting currency before dividing them. A numerator in euros and a denominator in dollars produce a meaningless ratio unless you apply and record an exchange convention. For a multi-market campaign, keep the original amounts, the rate source, the conversion date rule and the normalized totals. Rounding belongs at presentation, after the underlying totals are reconciled. If a reliable conversion basis is unavailable, report the currency groups separately. That is more useful for deciding which FroggyAds market test deserves further spend than showing a precise-looking blended number with incompatible inputs.

Choose a revenue basis that matches your business question and use it consistently. Tax collected for another party and shipping revenue that offsets a delivery expense need different interpretation from product margin. The calculator cannot determine your accounting treatment. Ask the responsible finance owner to define the numerator and any unrecoverable cost included in the denominator. Preserve a media-only ROAS view if that is the buying team's usual comparison, alongside a clearly named business-cost analysis. A documented basis makes the result understandable without presenting advertising revenue as spendable profit.

For leads, calculate assigned value from the accepted enquiries your business can justify. An illustrative 15 accepted leads at USD 20 each give USD 300 of modeled conversion value; divided by USD 100 of media cost, that is modeled value ROAS of 3.0x. Collected sales revenue remains a separate observation. Affiliate buyers need the same status discipline: pending, approved and paid commissions can differ. Choose FroggyAds to test the media purchase for your offer, while the receiving programme applies its own commission eligibility and reversal terms.

Compare their definitions before choosing a winner. One report may credit a sale to the last eligible click while another uses a different attribution model. Reports may also use different time zones, conversion dates, revenue adjustments or event coverage. Locate the first difference in a sample of matched records, then explain any unmatched portion. Do not add two platforms' credited revenue totals and assume the sum is unique sales. A shared customer order can receive credit in more than one reporting view. Keep the business's order or accepted-value record available for reconciliation.

Put the accepted order count beside revenue and spend. A single large purchase can dominate a small test, and a high ratio based on one order may change sharply after a reversal. Compare similarly mature groups and inspect the distribution of order values before assuming an average will persist. There is no universal order count that makes every campaign dependable. Consider the range of values, repeat behavior and consequence of a mistaken increase. A measured next purchase can gather further evidence while keeping the potential learning loss within your planned budget.

Change one assumption to see how much room the campaign has. Suppose the accepted revenue estimate is USD 1,800 and spend is USD 600: base ROAS is 3.0x. A hypothetical 20% reduction in eligible revenue gives USD 1,440 and 2.4x. The difference matters if your business requires 2.8x on the same margin basis. This is a sensitivity calculation, not a forecast that revenue will fall by 20%. Use the scenario to choose a tolerable test size and decide which uncertain input needs better evidence.

Divide the revenue cell by the spend cell for a ratio. If revenue is in B2 and cost in C2, the ratio expression is =B2/C2 when C2 is positive. Display that result as a number with an x label, or format the same value as a percentage. Multiplying by 100 and also applying percentage formatting would inflate the display again. Keep the raw amounts nearby so a colleague can reconstruct the answer. Missing cost should remain a missing-input state, rather than being filled with a convenient zero that suggests exceptional performance.

Estimate the share of net revenue available for advertising after the variable costs included in your chosen scope. If that contribution share is hypothetically 40%, the media break-even ratio is 1 divided by 0.40, or 2.5x. Fixed overhead and desired profit may require more. This simplified relationship assumes the contribution share stays stable at the proposed volume; discounting, shipping and product mix can change it. Replace the example with your own economics before setting a target. A high revenue multiple alone cannot establish that another traffic purchase will be profitable.

Choose an amount you can afford to spend while the campaign produces uncertain information. Base it on the offer's economics, the expected delay and the action you need to observe, rather than a universal test-budget claim. Enter the campaign budget and confirm the available bid controls in FroggyAds. A funded account, a maximum bid and a campaign budget answer different questions; check each in the actual setup. Record what would stop further purchasing, including a broken destination or unusable measurement route, before the first report encourages a larger commitment.

New and returning customers can have different acquisition economics. A blended ratio may look strong because existing customers spend more, even when the new-customer purchase is weak. Where your own records support the distinction, show each group's revenue and matching cost basis with any allocation limitations. Do not invent a person-level connection that your setup cannot establish. The same care applies to products with different margins. A source that favors lower-margin items can show more revenue while leaving less money for acquisition than another source with a smaller basket total.

Average ROAS covers the whole measured purchase. Marginal ROAS asks what the additional spend produced on the same attribution and value basis. In a hypothetical staged test, an initial USD 400 earns USD 1,600 and a later additional USD 200 earns USD 500. The combined result is USD 2,100 divided by USD 600, or 3.5x; the additional slice is 2.5x. Review that later slice before increasing again. A healthy historical average can conceal a weaker incremental buying decision, even though the arithmetic of both displayed ratios is correct.

Check inputs and operating changes in a useful order. Confirm that cost imports and value records are complete, then inspect refunds, unusually large orders and observation age. Compare source mix, offer price, availability and destination behavior before blaming the bid. If the numerator stopped updating, lowering bids will not repair the missing value feed. If accepted revenue falls while data coverage remains sound, review the affected media segment and offer route. FroggyAds source reporting helps locate the buying context; the receiving records explain whether the visitor action became usable business value.

A promotion can raise order count while lowering revenue per order and contribution. Build a separate scenario for the actual proposed price, expected discount treatment and variable costs. Avoid carrying the full-price ROAS target into the sale unchanged. You can compare the media purchase under each offer version, but the records should identify which offer was active. We recommend changing one commercial variable deliberately and retaining the prior result, so the next FroggyAds test answers a clear question about the new offer rather than mixing incompatible pricing periods.

Attributed revenue is revenue credited under a reporting rule. It does not prove that every credited sale would disappear without advertising. A calculator can describe that credited value efficiently; it cannot create a comparison group or establish incrementality. If your decision requires a causal claim, use an appropriate experiment design and its own analysis. Keep ordinary ROAS useful for day-to-day buying without overstating what it proves. You can still inspect source economics and control the next campaign purchase while a separate study evaluates the wider effect.

Retain the campaign and source scope, report dates, observation cutoff, value definition, currency basis and raw cost and value totals. Add the formula, any excluded records and the reason for the next action. When a late approval or refund changes the numerator, save a dated successor with the original available for comparison. This small record makes account handover easier: a colleague can distinguish a genuine performance change from a new counting rule. It also lets your team repeat a useful FroggyAds buying decision with the same commercial standard.