Skip to main content
    European Search Awards 2026 · Best Small PPC Agency

    Problem-Solution Hub

    Your ROAS is hitting targets. Your profit isn't.

    This is the most common problem we see in accounts spending £10k+ a month. The good news: it's fixable. The bad news: most agencies don't even know it's happening.

    The ROAS Illusion

    Why 4.5x ROAS can mean negative profit

    Revenue
    Profit
    JanFebMarAprMayJun£0£40k£80k£120k£160k£-5k£0£5k£9k£14k

    ROAS

    4.8x

    POAS

    -0.1x

    Trend

    ↓

    Symptoms

    Recognise any of these?

    If you're nodding to more than one, your POAS is probably broken. These aren't edge cases - they're the default state of most ecommerce Google Ads accounts spending over £10k a month.

    ROAS looks healthy but bank balance is flat

    The platform reports revenue against spend and calls that ROAS. It does not know your cost of goods, your returns rate or your payment processing fees, so a ratio that looks comfortably above target can still sit on top of a loss once those costs come off. The gap between the two numbers is exactly the cost data Google Ads never sees.

    Scaling spend doesn't scale profit

    Increasing budget does not increase profit at the same rate, because the algorithm chases whatever converts most easily as it spends more, and that is often your lowest-margin, highest-volume product. Margin compression accelerates as the account scales unless the bidding is told which products actually carry the profit.

    High-revenue SKUs are low-profit SKUs

    Your best sellers in Google Ads reporting are not necessarily your best performers in contribution margin. A hero product can generate the majority of attributed revenue while carrying thin margin after returns, while a lower-revenue product with much stronger margin gets almost no spend, simply because Google has no way to see the difference unless you tell it.

    Returns and BNPL are eating into cash

    The revenue Google Ads reports is captured at checkout. It does not subtract the return processed weeks later, the buy-now-pay-later settlement that arrives on a delay, or the restocking cost on returned goods, so reported ROAS can overstate what actually reaches the bank.

    Worked Example

    How 4.5x ROAS can still mean a loss

    Simplified teaching maths showing the mechanism, not a client result. It is representative of the kind of gap we see between dashboard ROAS and P&L reality for a mid-market fashion ecommerce brand.

    Illustrative exampleWhat the dashboard shows vs what the P&L shows

    What the dashboard shows

    Ad Spend£25,000
    Reported Revenue£112,500
    ROAS4.5x
    Agency verdict"Performing well"

    What the P&L shows

    Gross Revenue£112,500
    COGS (42%)−£47,250
    Returns (28%)−£31,500
    Shipping & Handling−£9,000
    Payment Processing (3%)−£3,375
    Ad Spend−£25,000
    Actual Profit−£3,625
    True POAS−0.15x

    The gap: in this illustration the brand reports 4.5x ROAS while the account is actually losing money once COGS, returns, shipping and payment processing are counted. The fix isn't to reduce spend - it's to redirect spend towards the SKUs that generate positive contribution margin.

    Simplified teaching maths. Not client performance. Figures are hypothetical.

    Methodology

    The four-stage POAS correction process

    We follow a structured diagnostic and implementation process from audit to full POAS optimisation. Exact timing depends on account size, catalogue complexity and how complete your cost data already is.

    01

    Commercial Audit

    We extract COGS data from your ERP or ecommerce platform and match it against Google Ads performance at SKU level. This reveals which products are genuinely profitable and which are subsidised by the algorithm's revenue bias.

    • Map landed COGS per SKU including duties, freight, and packaging
    • Calculate true contribution margin after returns and payment processing
    • Identify the break-even ROAS for each product category
    • Flag 'hero SKUs' that drive revenue but destroy margin
    02

    Feed Restructure

    We inject margin data into your product feed as custom labels, creating the foundation for profit-aware bidding. Products are segmented into four bands using our SKU Job Framework.

    • Scale: >30% margin, high volume - maximise spend
    • Protect: 15-30% margin - bid to break-even, let volume carry profit
    • Recover: <15% margin but strategically important - cap CPC aggressively
    • Pause: Negative contribution margin - exclude from paid activity
    03

    Campaign Architecture

    We rebuild campaign structure so each margin band operates with its own tROAS target. This prevents the algorithm from cross-subsidising loss-making SKUs with profitable ones.

    • Separate Shopping campaigns per margin band
    • Set tROAS targets calibrated to each band's break-even point
    • Implement negative keyword isolation between bands
    • Configure conversion value rules to reflect true margin
    04

    Post-Sale Calibration

    We feed actual post-sale data (returns, cancellations, BNPL adjustments) back into Google Ads conversion values. This creates a feedback loop where the algorithm learns from real profit, not projected revenue.

    • Implement conversion value adjustments for returns
    • Factor BNPL settlement delays into cash-adjusted ROAS
    • Build rolling margin correction models
    • Establish monthly P&L reconciliation against Google Ads reporting

    Solutions

    How do you fix low POAS?

    Four interconnected fixes that transform your Google Ads from a revenue engine into a profit engine.

    1. Audit SKU-level margins - Inject COGS data into your feed so every bid knows true profit potential.
    2. Implement POAS bidding - Restructure campaigns to bid on contribution margin, not revenue.
    3. Account for post-sale erosion - Factor returns, shipping, and payment delays into calculations.
    4. Align spend with cash flow - Sync ad spend timing with actual cash receipt.

    Verified case study

    Thermos

    +94% contribution margin

    Shifted from ROAS to POAS targeting after correcting double-counted conversions, restructuring campaigns around COGS tiers.

    Read the full case study

    +94%

    Contribution margin

    +86%

    Revenue

    5 months

    Engagement period

    Period, comparison basis and source are stated on the case study and evidence register.

    Verified case study

    Husk & Seed

    +45% average order value

    Moved from loss-making samples to profitable bundles, with contribution margin doubling and acquisition cost held flat.

    Read the full case study

    +45%

    Average order value

    2x

    Contribution margin

    Period, comparison basis and source are stated on the case study and evidence register.

    What is POAS and why does it matter more than ROAS?

    TLDR: POAS measures profit per £1 of ad spend. ROAS measures revenue. The difference determines whether your ads make or lose money.

    POAS (Profit on Ad Spend) measures the actual profit generated per pound of ad spend, after deducting COGS, shipping, returns, and payment processing. Unlike ROAS which only tracks gross revenue, POAS reveals whether your advertising is genuinely profitable. A campaign can show 5x ROAS while generating negative profit if the products sold carry thin margins and high return rates.

    Break-even ROAS at 25% margin:
    4.0x(JudeLuxe)

    How do you fix low POAS in ecommerce Google Ads?

    TLDR: Inject COGS into feeds, segment by margin bands, adjust for returns, and sync spend with cash flow.

    Fixing low POAS requires four steps: inject COGS data into your product feed so bids reflect true margins, restructure campaigns into margin-banded segments with separate tROAS targets, account for post-sale margin erosion (returns, shipping, payment delays), and align ad spend timing with actual cash receipt. Timing to see profit change depends on account economics, product mix, data quality and starting point.

    What data do I need to implement POAS bidding?

    TLDR: COGS per SKU, shipping costs, return rates, and payment fees - fed into Google Ads via custom labels.

    At minimum you need landed COGS per SKU, average shipping cost per order, return rates by product category, and payment processing fees. Ideally, add SKU-level return rates, BNPL settlement timelines, and seasonal margin variations. Most ecommerce platforms can export this data, and it gets structured for feed injection via custom labels.

    Questions

    Common questions about fixing low POAS

    POAS (Profit on Ad Spend) measures the actual profit generated per pound of ad spend, after deducting COGS, shipping, returns, and payment processing. ROAS only measures gross revenue against spend. A campaign can show 5x ROAS while generating negative profit if the products sold carry thin margins and high return rates. POAS reveals the truth that ROAS obscures.

    Divide 1 by your average contribution margin percentage. If your average margin after COGS, shipping, and returns is 25%, your break-even ROAS is 1 ÷ 0.25 = 4.0x. Any ROAS below this means you're losing money on every sale, regardless of what the platform dashboard says. For SKU-level accuracy, calculate this for each margin band separately.

    Partially. You can start by adding COGS data as custom labels in your product feed, which gives you reporting visibility. But to truly bid on profit, you need to restructure campaigns into margin-banded segments with separate tROAS targets. The feed changes are quick; the restructure and the algorithm's subsequent learning period take longer, and the length varies with account maturity and conversion volume.

    There is a transition period while campaigns are restructured and the algorithm relearns against contribution margin instead of revenue, during which reported volume can move in either direction as spend is redirected towards higher-margin SKUs. What we monitor throughout is contribution margin, break-even ROAS by band, and cash-adjusted profit, not just the headline ROAS. Scale and timing depend on your account's starting economics, product mix, data quality (how complete your COGS, returns and shipping data is) and how much of the catalogue needs restructuring, so we do not quote a single timeline or improvement range up front.

    At minimum: landed COGS per SKU, average shipping cost per order, return rates by product category, and payment processing fees. Ideally, you'd also have SKU-level return rates, BNPL settlement timelines, and seasonal margin variations. Most ecommerce platforms can export this data, and we help clients structure it for feed injection.

    POAS is arguably most important for low-margin businesses, because the gap between revenue and profit is largest. A fashion brand at 20% margin needs 5x ROAS just to break even - and that's before returns. POAS bidding prevents the algorithm from chasing high-revenue, low-margin sales that look good in reports but bleed cash.

    Next step

    Ready to fix your POAS?

    Book a commercial review. We'll diagnose your profit leaks in 30 minutes and tell you if we can help.