Friday 9:40 a.m. in a Richmond warehouse office: the Google Ads dashboard shows $18,400 spent this week in AUD, Shopping is “winning,” and last-click ROAS looks like 3.7x. Then finance forwards the Shopify payout file. After GST, returns from the last Afterpay wave, and the Metro Melbourne shipping surcharge you never modelled into the bid strategy, contribution on the hero SKU is closer to break-even. That gap is the real job of PPC for e-commerce companies in Australia, not a prettier ROAS tile.
Australian catalogues live in a nasty mix of high CPCs, GST-inclusive pricing, and buyers who compare you against Amazon AU, Kogan, and a dozen Shopify stores before they add to cart. Australian retail CPCs in Google Shopping frequently run $1.50-$4.50 depending on category, and they do not get kinder when you bid on “buy running shoes melbourne.” Performance Max often carries the bulk of spend. Meta sits closer to 2.8x average ROAS while Google’s blended 4.2x still hides SKUs that never should have been in the feed. If you run paid search for an Australian store, you are not fighting a UK playbook with the word “Sydney” swapped in. You are fighting GST, returns, freight, and a conversion signal that quietly drops 20%+ to ad blockers and privacy gaps.
Where Australian Shopping and Search spend dies before checkout
You will usually find the leak in three places at once. The feed still ships “Navy / M” as a title Google cannot match. Search still harvests informational queries like “how to wash merino” that never buy. The PDP takes 4.1 seconds on 4G in Parramatta, and mobile already accounts for 63%+ of paid search clicks while converting 30-40% worse than desktop. Only 55.9% of origins pass all three Core Web Vitals in the May 2026 CrUX set, so a slow PDP is not a design nit. It is a bid tax.
Twenty to thirty percent of SEM budgets still vanish on irrelevant queries, neglected negatives, or a homepage that does not match the ad. If someone on your team says they just send Shopping traffic to the collection page, you already know the outcome: the auction paid for “women’s trail runners size 8” and the page asked the shopper to start again. Dedicated landing paths and feed hygiene beat adding another channel.
GST-inclusive RPAs also poison smart bidding. If your conversion value in Google Ads is ex-GST and Shopify is inc-GST, tROAS chases a number finance will never recognise. Afterpay and Zip add another delay: the “sale” fires, then a 14-day return lands, and last-click still congratulates Search. Last-click also double-counts the same order when PMax, branded Search, and a Meta retargeting ad all claim it. Attribution is not “dead,” but last-click as a CEO answer is. Use it tactically. Do not staff a board pack on it.
High CPC pressure makes new accounts want to flee back to manual CPC. That feels like control. It usually just freezes learning while you underbid the SKUs that could actually cover freight. AI Max style Search setups showing 12-18% lower CPCs versus standard Search only help if the conversion you feed them is a real, refund-adjusted order value, not a thank-you page ping.
If paid clicks already hit a generic homepage, stop scaling until the path matches the query. A short look at Google Ads and paid search management is useful when the account, the feed, and the PDP are out of sync, not when you need another dashboard.
How a GST-aware Australian catalogue actually bids
Treat Search, Shopping, and Performance Max as one inventory system with different jobs, not three ways to show the same hero SKU. Search owns high-intent modifiers and competitor interception (“wool dryer balls vs dryer sheets”). Shopping owns visual discovery and long-tail colour/size queries. PMax should harvest incrementality on the rest, not clone the same product group at a higher blended CPC. “Performance Max drives the bulk now” is fine if the asset groups are split by margin band, not by whoever built the campaign in 2024.
Signals beat keyword volume. First-party purchase events, Enhanced Conversions (unified toggle), and a clean server-side path protect you when 20%+ of conversion data never reaches the browser. Google still needs keywords for the queries you refuse to buy. Negatives are not optional theatre. They are how you stop paying for “jobs,” “wholesale,” and “DIY pattern” when you sell finished goods.
Meta is a different machine. Creative is your targeting. Broad plus a systematic test engine beats stacked lookalikes from 2020. UGC that peaked in 5-6 days will not carry a EOFY sale. Watch CPM before you watch volume: fatigue shows in auction cost first. Do not let Meta’s in-platform CPA become the source of truth when it diverges from Shopify contribution. Meta’s current delivery system (Andromeda, Meta’s current delivery algorithm) rewards hooks and formats that fit the scroll naturally, not disruptive creative that tanks watch time. Weak creative plus broad targeting is how you buy cheap, useless traffic.
Australian specifics you cannot skip: bid modifiers and value rules for metro versus regional shipping, separate product groups for SKUs that cannot profitably ship to WA, and promo assets that state GST-inclusive prices so you do not win the click and lose the cart on GST shock. Afterpay messaging on the PDP has to match the RSA. If the ad implies four payments and the page hides it under a widget, you paid for a bounce.
PMax in cleaned-up accounts often lands in a 4x-8x ROAS band. That is not a promise. It is what happens when the feed, the values, and the landing experience are not fighting the algorithm. Nobody can guarantee a ROAS or CPA before they have seen your account, unit economics, and PDPs. Realistic lift takes 3-6 months of signal work, not a week of campaign reshuffles that reset learning.
Get a free marketing audit - we review your search, ads, and landing pages and send back what to fix first.
Get a free audit
Feed, bid, and PDP scorecard Australian store marketers can run
Work this as an audit, not a vibe check. Score the account against contribution, not last-click vanity.
Audit scorecard
-
Conversion value matches Shopify contributionMap purchase value to GST-i
Conversion value matches Shopify contributionMap purchase value to GST-inclusive revenue minus a rolling return rate for the last 90 days. If Google’s conversion value and the payout file disagree by more than a few points, pause tROAS increases until they match. Include Afterpay/Zip as the same order, not a second conversion.
-
Feed titles match how Australians searchRewrite titles as brand + produc
Feed titles match how Australians searchRewrite titles as brand + product type + material/size/colour in language people type (“Merino crew socks 3-pack black”), not internal SKU poetry. Exclude out-of-stock and unprofitable freight SKUs from the Shopping primary feed.
-
No SKU triple-served across Search, Shopping, PMaxBrand Search should no
No SKU triple-served across Search, Shopping, PMaxBrand Search should not also sit inside a PMax asset group bidding the same exact query. Use campaign priorities, brand exclusions, and inventory filters so one SKU is not competing with itself at three CPCs.
-
Negatives for the queries that never buyPull 30 days of search terms
Negatives for the queries that never buyPull 30 days of search terms. Negate jobs, wholesale, free, PDF, meaning, DIY, and competitor support queries. This is how you claw back part of the 20-30% waste band without touching bids.
-
PDP is the conversion eventAd headline, price (AUD, GST-inclusive), size
PDP is the conversion eventAd headline, price (AUD, GST-inclusive), size selector, shipping promise, and primary CTA must be visible without a hunt. If Interaction to Next Paint on the add-to-cart control is ugly, smart bidding is bidding on a page that cannot close.
-
Enhanced Conversions and first-party events are onTurn on the unified En
Enhanced Conversions and first-party events are onTurn on the unified Enhanced Conversions path. Deduplicate browser and server events. If you lose a fifth of purchases to blockers, tROAS will overbid cheap sessions and underbid the buyers who use Safari.
-
Freight and metro rules in the bid layerValue rules or separate product
Freight and metro rules in the bid layerValue rules or separate product groups for WA/NT and remote postcodes. If a $42 hoodie cannot survive $18 freight, it should not share a tROAS target with a $189 jacket that ships from a Sydney 3PL.
-
Mobile path measured separatelyDo not let desktop ROAS hide a mobile PDP
Mobile path measured separatelyDo not let desktop ROAS hide a mobile PDP that drops 30-40% relative. Split reporting. Fix sticky ATC, size charts, and Afterpay clarity on phone before you raise Shopping budgets.
Action checklist
- Export 90 days of SKU-level spend, revenue, return rate, and shipping cost in AUD. Rank SKUs by contribution, not revenue.
- Kill or cap the bottom contribution decile in the Merchant feed before you touch bids.
- Rebuild one money PDP: price, stock, freight promise, reviews, and a single primary CTA above the fold on mobile.
- Align RSA and Shopping titles to that PDP. No “sale” language the page does not honour.
- Turn on Enhanced Conversions, verify purchase value, then give smart bidding two full learning windows without daily structural edits.
- Only then raise budget on the product groups that already cover GST, returns, and freight.
Free tools - try these yourself
Session recordings and heatmaps in HeyLead Insights show where size charts, sticky ATC, and freight copy leak intent after an expensive Shopping click. Guessing at form friction is how catalogues scale spend into a wall.
Two Australian catalogue failures that looked like “the algorithm”
A performance lead at a Melbourne athleisure brand kept PMax and a standard Shopping campaign on the same 40 SKUs. CPC on the hero tights sat at $1.91 in Shopping and $2.64 in PMax for overlapping queries. They excluded the brand terms from PMax, split asset groups by margin above and below 42%, and rewrote feed titles to include fabric and inseam. Contribution on that line moved from roughly break-even to a 4.2x last-click that finance could actually reconcile, because they stopped paying twice for the same click path. The mechanism was inventory overlap, not “PMax is broken.”
A Brisbane homewares operator blamed automated bidding for inflating spend. Manual CPC on a new account felt safer. Conversion tracking only fired in the browser, Safari users vanished, and tROAS had been chasing a 22% hole in the data. They turned on Enhanced Conversions, passed order value inc-GST, and left the bid strategy alone for 11 days. CPC did not magically drop. Qualified add-to-carts did, because the model finally saw the buyers it had been under-counting. Automated bidding was not the villain. The missing 20% of the signal was.
A third pattern shows up every EOFY: Meta CPMs creep for four days, then “lead volume” (really add-to-cart) falls. Teams refresh the offer instead of the creative. UGC that worked on day one is dead by day six. Swap the hook and format first. Keep the same SKU and landing URL so you do not reset learning on the funnel and the ad at the same time.
When the PDP is the bottleneck, more budget is a more expensive way to watch the same bounce. If you need a second pair of eyes on the post-click path, keep it on the page, not another channel.
FAQs
Should Australian e-commerce teams still use keywords if PMax carries most spend?
Yes. PMax needs clean inventory and values. Search still needs exact and phrase coverage for the queries you must win, plus negatives for the ones you must never buy. “Signals over keywords” does not mean an empty Search campaign.
Is a 4.2x Google ROAS a sensible target for an AU store?
It is a blended 2026 average across industries, not a merchandising plan. Your number has to clear GST, returns, payment fees, and freight in AUD. A 4x last-click with a 28% return rate on fashion can still lose money.
Why does Meta look cheaper than Google and then fail at contribution?
Average Meta ROAS near 2.8x often includes view-through credit and a softer conversion definition. If creative dies in days and you retarget the same purchasers, in-platform CPA diverges from Shopify. Judge Meta on incrementality and contribution, not the ads manager tile.
Will AI Overviews kill Shopping clicks?
AI Overviews now reach huge query volume and can cut CTR sharply on informational searches. Product queries with a feed, price, and Merchant Centre still buy. Do not staff the whole plan on informational Search CTR you no longer own. Prioritise structured data (price, availability, review schema) on PDPs so Shopping units appear inside or alongside AI Overview panels. Monitor Search Console for impression-to-click ratio decay on top-of-funnel product queries month on month.
How often should we rebuild campaign structure?
Constant rebuilds reset learning. Fix feed, values, negatives, and PDPs first. Change structure when inventory roles conflict, not because a weekly ritual demands a new campaign type.
What clients say
“We were winning Shopping auctions and still breaking even on hero SKUs after GST and returns. Cleaning the feed, splitting PMax by margin, and passing inc-GST order value lifted contribution on our best homewares lines in one EOFY cycle - finance could finally reconcile ads to Shopify.” - Ecommerce manager, Australian homewares brand
Putting it to work
Execution sprint
This week
- Pull 30-90 days of performance for e-commerce (Search Console, ads, CRM, or call logs - whatever you have).
- Flag the top leak: wrong intent, weak page, slow response, or dirty conversion tracking.
- Ship one fix on the highest-traffic money path (page, campaign split, or response rule).
- Run the free tools below on that same URL or account and log the findings.
Free tools for this sprint
Next 30 days
- Expand the fix to the next one or two money paths only after the first one shows cleaner bookings.
- Align creative, keywords, or content with the same offer the page now states.
- Review booked outcomes weekly; cut anything that still only produces unqualified volume.
This week
-
Pull 90 days of SKU spend versus contribution in AUD, including returns and shipping.
-
Negate the obvious non-buyer search terms and pause SKUs that cannot cover freight to WA.
-
Run the Core Web Vitals checker on your top three paid PDPs and fix INP on add to cart.
-
Confirm purchase value and Enhanced Conversions match Shopify inc-GST.
-
Tag one Shopping and one Search URL with the UTM builder so GA4 and ads stop arguing about source.
Next 30 days
-
Split PMax asset groups by margin band and stop overlapping brand Search.
-
Rebuild one hero PDP for mobile freight, size, and Afterpay message match.
-
Stand up a weekly creative swap on Meta so ads do not die on the same UGC for two weeks.
-
Reconcile ads-reported ROAS to contribution, not last-click alone.
Start by lining Google’s conversion value up with GST-inclusive contribution on your top 20 SKUs, then stop bidding the rest like they are equally profitable. HeyLead can own that GST-aware feed, bid, and PDP loop so you are not reconciling Shopping overlap and missing purchase signals every EOFY. Chat with us on WhatsApp
Free marketing audit, or reach Martin directly:
Get a free audit Connect on WhatsApp · martin@heylead.com