Target ROAS for New-Customer-Only Campaigns
New-customer-only campaigns justify a lower ROAS floor because downstream LTV closes the gap. Here's how to derive it, what typical floors look like by channel, and how to report it without losing the budget conversation.
Quick answer
A new-customer-only campaign can run at roughly 40-70% of your blended ROAS target, because you're paid back on cohort LTV rather than the first order. Derive the floor as: (CM% × LTV over your payback window) ÷ desired CAC multiple. If your blended target is 3.0 and repeat customers double first-order revenue within 12 months, a prospecting floor of 1.5-1.8 is defensible.
Target ROAS for New-Customer-Only Campaigns
The lower ROAS floor justified when a campaign acquires only net-new customers whose downstream LTV closes the gap on first-purchase profitability.
Target ROAS for new-customer-only campaigns is the acceptance threshold you set for prospecting spend that is guaranteed — via platform audience filters or first-party suppression — to reach non-buyers only. Because the same customer will (on average) reorder within your payback window, the campaign doesn't need to be profitable on the first order. It needs to be profitable on the cohort.
The floor is derived from contribution-margin-adjusted LTV, a chosen payback window, and a CAC-recovery multiple. It sits below your blended ROAS target and above your break-even-on-first-order line. Reporting it honestly requires isolating new-customer revenue from repeat revenue that the platform may be over-attributing.
Most brands set a single blended ROAS target across all campaigns. That works until prospecting starts looking unprofitable next to retargeting, and the reflex is to cut it. What actually died was your new-customer supply — and six months later your repeat revenue drops with it.
A separate ROAS floor for new-customer-only campaigns fixes that. It lets prospecting run at a first-order return that would look bad in isolation, on the receipt that cohort revenue will make up the gap.
Why the floor is lower than your blended target
Your blended ROAS target already averages new-customer and repeat-customer revenue. Repeat orders carry no acquisition cost, so they inflate the blended number. A campaign that only touches new customers cannot mathematically hit the blended target on day-one revenue — the repeat lift is baked in downstream.
The correct comparison is first-order ROAS on the prospecting campaign versus first-order-plus-expected-repeat-revenue over your chosen payback window. If your repeat rate is 35% and average repeat order value matches AOV, a 12-month LTV of roughly 1.5× first-order revenue is realistic — which means a prospecting ROAS of 1.5 delivers the same total return as a 2.25 first-order ROAS on a mixed campaign.
The trap: first-order discount depth
A 20% welcome code erodes contribution margin on exactly the order you're using to justify the lower floor. If your CM% drops from 65% to 45% on first orders, your defensible floor rises — not falls. Model the discount before you set the target, or the math silently breaks.
Deriving your floor from LTV and payback window
Start with contribution-margin-adjusted LTV over the payback window you're willing to wait. Six months is aggressive; twelve is standard for apparel and beauty; eighteen only works if finance genuinely accepts the cash-flow drag.
The formula is: floor ROAS = (first-order revenue + expected repeat revenue in window) ÷ (first-order revenue × CAC multiple). For a Shopify beauty brand with €60 AOV, 40% 12-month repeat rate, €35 repeat AOV, 60% CM, and a 3× CAC-recovery target, the floor lands near 1.6.
The payback window choice matters more than most teams realise — a 6-month window sets a materially higher floor than 12 because you're capturing fewer repeat orders. LTV by acquisition channel matters too: TikTok-acquired customers historically show weaker 12-month LTV than Google-search-acquired ones, so channel-specific floors are more honest than one blended prospecting number.
Typical new-customer ROAS floors by channel and vertical
Indicative new-customer-only ROAS floors, assuming 12-month payback and 55-65% CM
| Vertical | Meta prospecting | Google prospecting | TikTok prospecting | Blended target |
|---|---|---|---|---|
| Apparel (€40-80 AOV) | 1.4-1.8 | 1.8-2.4 | 1.1-1.5 | 2.8-3.2 |
| Beauty & skincare | 1.5-2.0 | 2.0-2.6 | 1.2-1.6 | 3.0-3.5 |
| Home & lifestyle | 1.6-2.1 | 2.1-2.8 | 1.3-1.7 | 2.6-3.0 |
| Consumer electronics | 2.0-2.6 | 2.4-3.0 | 1.6-2.0 | 3.2-3.8 |
| Supplements / CPG repeat | 1.2-1.6 | 1.6-2.1 | 1.0-1.4 | 2.4-2.8 |
Verticals with strong repeat behaviour (supplements, skincare) can defend the lowest first-order floors because cohort LTV catches up fastest. High-consideration, low-repeat categories (consumer electronics) need floors much closer to the blended target — the LTV gap simply isn't there.
Reading the platform's reported new-customer ROAS honestly
Meta's new-customer-acquisition objective and Google Ads' new-customer value bidding both let you optimise toward net-new buyers. Both also report ROAS in their own attribution window, which typically overstates new-customer revenue by 15-30% versus first-party order tagging on Shopify or WooCommerce.
Cross-check the platform number against your store's customer tags — new vs returning — over a rolling 28-day window. If Meta claims a 1.8 new-customer ROAS and your Shopify data says 1.4, the honest number is 1.4. Set your floor against the honest number, not the reported one.
Reporting the lower floor without losing the budget conversation
A CFO looking at a 1.5 ROAS on prospecting sees a losing campaign. Reframe with a cohort receipt: show the same customers' 90-day and 180-day revenue against acquisition cost, not the day-one line. Two quarters of that report and the conversation shifts from 'why is this so low' to 'can we spend more here'.
Pair the floor with a repeat-rate monitoring rule: if cohort retention slips more than 10% versus the assumption you used to derive the floor, raise the floor. That's the guardrail that keeps this from becoming an excuse to accept bad prospecting performance forever.
Frequently asked questions
The three practical definitions are 90-day non-buyer, 12-month non-buyer, and never-purchased. Never-purchased is cleanest but the smallest audience; 12-month is the mainstream choice for most stores. Match the window your LTV model uses — if LTV is a 12-month figure, define new customer against a 12-month lookback.
Use 60% of your blended ROAS target as a placeholder. If blended is 3.0, run prospecting at 1.8 and instrument cohort revenue for 90-180 days. Once you have two full cohorts, replace the placeholder with a floor derived from CM-adjusted LTV over your chosen payback window.
Yes — channel-acquired customers show materially different 12-month LTV. Google-search-acquired customers typically repeat more than TikTok-acquired ones because intent is higher on entry. A single prospecting floor across all three channels over-funds the weakest channel and starves the strongest.
It can. A 20% code cuts CM% on exactly the order you're modelling, which raises the required first-order revenue to hit the same downstream return. Model the discount into the floor derivation, or restrict the code to a specific SKU that preserves margin.
It restricts delivery to audiences the platform classifies as non-buyers, which is what makes the lower floor honest. But Meta's reported new-customer ROAS is optimistic by 15-30% versus first-party Shopify tags — set your floor against your store data, not the platform number.
Twelve months is standard and matches most LTV models. Six months sets a materially higher floor because fewer repeat orders land in the window; use it only if cash flow constraints require faster payback. Eighteen months lowers the floor further but ties up working capital most brands can't afford.
Show a cohort receipt: the same customers' 90- and 180-day revenue versus their acquisition cost, not the day-one line. Present it alongside the blended target so the split logic is visible. Two quarters of the report and the conversation moves from defense to expansion.
Your floor becomes wrong immediately. If cohort retention slips 10% versus the assumption behind your derivation, the floor needs to rise proportionally — otherwise you're funding customers who won't pay back the LTV gap. Instrument a monthly retention check against the floor's underlying assumption.
Similar intent, different mechanic. Google lets you set a monetary delta — extra value assigned to a new-customer conversion — which effectively lowers the ROAS the algorithm needs to hit on those clicks. Set the delta to match your LTV gap: if new customers are worth an extra €40 in downstream revenue, that's the delta.
No — you'll get contradictory bidding signals. Structure prospecting and retargeting as separate campaigns with separate targets. Use audience suppression (customer-list exclusions from Shopify or Klaviyo) to guarantee the prospecting campaign truly only touches non-buyers.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.