Skip to main content
    European Search Awards 2026 Winner - Best PPC Agency

    Guide / Commercial decisions

    Should I
    change PPC
    agency?

    It's one of the biggest commercial decisions an ecommerce business can make. Get it right and you unlock growth. Get it wrong and you've lost another six months, another peak trading period, and another chunk of profit you'll never get back.

    Reading time

    8 minutes

    Written for

    Founders and marketing leads at £3M–£100M ecommerce brands who are quietly wondering if the agency is the problem.

    The short answer

    Maybe. Read the six questions before you decide.

    The problem

    Most brands don't actually know whether they have an agency problem.

    They know revenue isn't where they want it to be.

    They know profit feels tighter.

    They know Google Ads isn't delivering what they expected.

    So they assume the agency has to go.

    Sometimes they're right. Sometimes the agency is doing exactly what they've been hired to do — just managing Google Ads.

    The problem is the business needs someone making commercial decisions, not platform decisions.

    Read this twice

    The most expensive PPC agency isn't the one with the highest retainer. It's the one that quietly costs you profit every single day while telling you everything is "performing well."

    Before you fire anyone

    Three things brands mistake for an agency problem.

    If any of these describe your situation, the switch you're considering probably won't fix what's actually broken.

    Misdiagnosis 1

    Revenue is soft, so the agency must be broken.

    Sometimes revenue is soft because demand is soft, margins have shifted, or the product mix has changed. The agency didn't cause it. Changing them doesn't fix it.

    Misdiagnosis 2

    ROAS dropped, so the account needs restructuring.

    ROAS drops when margin drops, when returning-customer share falls, or when you added a discount code. None of those are account problems.

    Misdiagnosis 3

    The dashboard looks green, so everything's fine.

    Every metric an agency optimises to can be green while the business loses money. Green dashboards are the most expensive lie in ecommerce marketing.

    The diagnostic

    Ask your agency six questions.

    Not "how are campaigns performing?" Those questions have safe answers. Ask the questions that reveal whether they think about your business — or only about the platform.

    For each one we've written what a good answer sounds like, and what a bad one sounds like. Score honestly.

    Q / 01

    Can you explain every pound we're spending?

    Not campaign by campaign. Commercially. Why is this money being invested and what business outcome does it create?

    A good answer sounds like

    • —Ties spend to a commercial job: acquisition, margin defence, inventory movement, category protection.
    • —Can name the three campaigns you'd cut tomorrow without hurting the business.
    • —Talks about spend as an investment thesis, not a media plan.

    A bad answer sounds like

    • —Reads out ROAS by campaign as if that's the answer.
    • —Explains cost distribution by platform recommendation, not commercial intent.
    • —Uses the phrase "the algorithm is learning" more than once.

    Q / 02

    Do you know our margins?

    Not an average. Not a guess. The actual profitability of the products you're advertising.

    A good answer sounds like

    • —Has margin data at SKU or product-group level inside the account.
    • —Bids differently on high-margin lines than on loss-leaders — and can prove it.
    • —Knows which SKUs shouldn't be advertised at any price.

    A bad answer sounds like

    • —Uses a single blended margin across the whole catalogue.
    • —Has never asked for a cost file.
    • —Optimises to revenue targets and calls it "efficiency".

    Q / 03

    Can you explain our customer economics?

    Contribution. CAC. Payback. LTV. New vs returning customers. When is acquiring a customer actually worth it?

    A good answer sounds like

    • —Knows your allowable CAC by product category and traffic type.
    • —Distinguishes new-customer performance from repeat-customer performance in reporting.
    • —Can defend spend on a low-first-order-margin SKU using payback and LTV.

    A bad answer sounds like

    • —Reports a single blended ROAS across new and returning.
    • —Cannot tell you what a new customer is worth in month one versus month twelve.
    • —Has never modelled a payback window with your finance team.

    Q / 04

    Do you know our inventory position?

    Which products need pushing? Which should be protected? Which should never receive another pound of ad spend until stock changes?

    A good answer sounds like

    • —Reviews stock cover before making bid changes.
    • —Has a rule for pulling ad spend on lines under a stock threshold.
    • —Coordinates with your merchandising or ops team, not just your marketing lead.

    A bad answer sounds like

    • —Bids up SKUs into stockouts because they're "performing well".
    • —Doesn't know your lead times.
    • —Treats the feed as a technical file, not a commercial one.

    Q / 05

    Which products lose us money every time they sell through Google Ads?

    Not low ROAS. Negative commercial value. Because revenue is easy. Profitable revenue is a completely different game.

    A good answer sounds like

    • —Names the SKUs with structurally negative contribution after ad cost.
    • —Has proposed excluding, deprioritising, or repricing those SKUs.
    • —Understands that a 4x ROAS product can still cost you money.

    A bad answer sounds like

    • —Insists everything is "performing" because ROAS is above target.
    • —Has never opened a P&L or a contribution report.
    • —Treats returns, shipping, and fulfilment as someone else's problem.

    Q / 06

    Have you ever recommended we spend less?

    This is the one that separates media buyers from commercial partners. Because sometimes the right answer isn't increasing budget. It's protecting profit.

    A good answer sounds like

    • —Yes — and can point to the meeting, the reasoning, and the outcome.
    • —Has actively de-scaled a campaign or paused an entire product group.
    • —Is willing to leave money on the table when the maths says so.

    A bad answer sounds like

    • —Cannot remember a single instance of recommending a spend reduction.
    • —Frames every quarterly review around adding budget.
    • —Confuses "managing your account" with "spending your money".

    Reading the answers

    Score honestly. Then decide.

    Mostly yes — four or more

    Stay.

    Seriously. Good agencies are incredibly difficult to find.

    Push them harder. Give them more data. Challenge their thinking. Share the cost file. Bring them into the commercial conversation.

    The uplift from a good agency you actually feed properly beats the uplift from switching — almost every time.

    Mostly no — three or fewer

    You don't have a Google Ads strategy.

    You have someone operating Google Ads.

    Those are not the same thing.

    One asks how to improve ROAS. The other asks how to make this business more money. Those questions produce completely different answers.

    Two questions, two agencies

    A media buyer asks

    "How do we improve ROAS?"

    Optimises to a platform metric. Reports on the platform. Wins the platform. Sometimes wins the business. Often doesn't.

    A commercial partner asks

    "How do we make this business more money?"

    Optimises to profit, cashflow, and payback. Reports on the P&L. Sometimes the answer is spending less. Sometimes it's spending more. Always it's the answer that grows the business.

    Before you switch

    Six things to have in place first.

    A switch that isn't set up properly is just an expensive way to reset the learning phase and blame someone new.

    1. 01You have twelve months of clean margin data by SKU or product group.
    2. 02You know your allowable CAC and payback window by category.
    3. 03You've shared your cost file with your current agency at least once.
    4. 04You've asked the six questions on this page and written down the answers.
    5. 05You've isolated whether the problem is the agency, the product, or the market.
    6. 06You have a clear brief for what a new agency would do differently.

    The real cost of a bad switch

    6 mo.

    Typical time lost to onboarding, learning phases, and rebuilding the attribution and reporting the old agency owned.

    1 peak

    Most switches straddle a peak trading period. If Q4 is fifty percent of your year, a botched handover costs more than the retainer ever saved.

    Profit

    You don't recover the profit lost during a bad transition. It's gone. This is why the diagnosis matters more than the decision.

    Changing agencies shouldn't be about finding someone who promises better CPCs, more AI, or another shiny dashboard.

    It should be about finding someone who understands your business well enough to tell you when not to spend money.

    That's the conversation worth paying for.

    If your answers were mostly no

    We'll audit your account against these six questions. Free.

    Send us your cost file before the call. We'll come back with a commercial read on your account — margins, customer economics, inventory, and the products that lose you money every time they sell. No pitch deck. Capped at four audits a month so it's actually done properly.

    Request the audit

    What you get

    A written commercial audit against the six questions, plus the three biggest profit leaks in your current setup.

    What we ask for

    A cost file. Read-only access to the ad account. Thirty minutes of your time.

    What we don't do

    Sales calls disguised as audits. If we can't help, we'll tell you and point you somewhere that can.

    We use cookies to improve your experience. Privacy Policy