The conversion column in a Sydney or Brisbane Search account can look heroic by Thursday. CPL in AUD is down. Smart Bidding is “learning”. Sales still has an empty calendar for next week. That gap is the real cost of Google Ads lead quality mistakes in Australia: you paid GST-inclusive media for people who never intended to buy, while the dashboard treated every thank-you page as a win.
In competitive service categories, Search CPC in Australia regularly clears $8-$15 and rising, so every junk enquiry is more expensive than it used to be. In accounts we audit, irrelevant queries routinely consume 15-35% of spend before a proper negative list is built - neglected negatives, or a landing page that does not match the ad. In most AU service accounts we manage, mobile delivers 60-70% of clicks but 20-30% fewer conversions per session - the gap varies by form length and page speed. If you count form volume as the marketing number, you will scale the leak.
This piece is about the settings, signals, and post-click habits that fill CRM rows without filling jobs. Not a rant about “strategy”. The actual switches teams leave on.
Six Google Ads habits that pad Australian dashboards with junk enquiries
Most of these sit in the account, not in a slide deck. Someone chose a conversion action, a match type, a Performance Max asset group, or a homepage as the destination. Each choice has a mechanism. Each one has a fix you can run without rebuilding the whole media plan.
Common mistakes
Action checklist
- Optimising to every form submit, including the brochure download In Google Ads, primary conversions train bidding. If “Contact”, “Quote”, “Newsletter”, and “Pricing PDF” all sit as primary, Maximize Conversions chases the cheapest action. In AU B2B that is often a student, a competitor, or a strata committee member collecting brochures. Mechanism: Enhanced Conversions and tCPA still fire on the easy event, so CPA looks cheap while booked work does not move. Fix: one primary action that maps to a sales-ready enquiry or a qualified booked call. Everything else secondary. Recency of 30 days of clean data before you trust Smart Bidding again. If you lost 20%+ of conversion data to ad blockers, turn on Enhanced Conversions with a unified first-party setup so the remaining signal is the expensive one, not the noisy one.
- Broad match on commercial terms with an empty negative list Google's internal benchmarks cite 12-18% lower CPCs when broad match is paired with Smart Bidding versus exact-only - though results vary sharply by vertical and negative list hygiene. It also pulls “jobs”, “salary”, “course”, “DIY”, “template”, and “free” unless you fence it. Search terms reports in Melbourne and Perth accounts still show internships sitting next to “enterprise” queries. Mechanism: you never uploaded a shared negative list for careers, education, and competitor homepages, and you never reviewed search terms after the last structure change. Fix: a shared negative list at account level, phrase and exact negatives on the junk clusters, and a weekly (not daily) search-term pass so you do not reset learning. Keep the high-intent commercial terms. Kill the research and HR traffic.
- Sending every ad to the homepage because “the site already converts” If someone says they just send traffic to the homepage, treat that as a quality warning. The ad promised a quote in 24 hours. The homepage opens on brand story, three CTAs, and a slow hero video. In most AU service accounts we manage, mobile delivers 60-70% of clicks but 20-30% fewer conversions per session - gap varies by form length and page speed. Mechanism: message mismatch plus form friction. People bounce, then a handful of low-intent browsers still submit, so the account learns the wrong visitor. Fix: one landing page per intent cluster (demo, quote, emergency call, partnership). Same promise in headline as in RSA. Proof above the fold. If you need to see where the form actually dies, session behaviour in HeyLead Insights shows scroll depth and abandon on the Australian page, not a guess from bounce rate. Roughly half of all origins still fail at least one Core Web Vital (HTTP Archive, 2024); a laggy AU landing page quietly taxes every click you already paid for.
- Letting Performance Max eat Search without brand exclusions or asset discipline “Performance Max drives the bulk now” is true in a lot of AU accounts. It is also where brand search, YouTube junk, and Display junk get mixed into one ROAS number that looks healthy because branded queries did the work. Mechanism: no brand exclusion campaign, weak audience signals, and feed or asset groups that do not match the offer. Industry ROAS benchmarks mask enormous variance - a 4x blended figure means nothing if branded queries are doing the heavy lifting. Fix: keep high-intent Search as its own campaign. Exclude brand from PMax if brand already converts on exact. Feed PMax clean first-party lists and a tight creative set. Do not judge PMax on last-click form volume. Judge it on booked jobs after sales has sat with the leads for a fortnight.
- Counting thank-you page views while call and chat quality is invisible Plenty of Australian companies still live on the phone: tradies, clinics, field services, even B2B when the buyer wants a human. If Ads only sees a pageview conversion, click-to-call and missed after-hours calls never train bidding. Mechanism: GTM fires on /thanks/, call tracking is a separate spreadsheet, chat is untagged. Smart Bidding then prefers cheap form fillers over callers who actually book. Fix: import qualified calls (duration threshold that matches a real conversation, not a 10-second hang-up) as a conversion. Tag chat starts that reach a human. Keep UTM discipline so GA4 and Ads do not invent two sources for one job. Last-click already double-counts across channels; adding a second untagged phone number makes the lie worse.
- Resetting the account every time CPC jumps, then blaming automation High CPC pressure makes teams flip from tCPA to Manual CPC, split campaigns, pause PMax, then switch back in 11 days. Each change dumps the learning phase. Automated bidding can inflate spend if the conversion you fed it is junk; it does not magically invent bad leads if the signal is a booked job. Mechanism: you treated two expensive days in Adelaide as a strategy failure. Fix: freeze structure long enough for 30-50 primary conversions. If quality is bad, change the conversion definition and the landing offer first. Only then change bid strategy. Constant structure edits are how good accounts look “volatile” on a Monday report.
DIY tools for this section
Catch the tracking, snippet, and landing gaps before you add budget
If the account is already spending and the landing handoff is the weak joint, a tighter SEM / Google Ads program that treats conversions as booked work, not form counts, is usually cheaper than another month of “cheap CPL”.
Get a free marketing audit - we review your search, ads, and landing pages and send back what to fix first.
Get a free audit
How to fix Google Ads lead quality in Australia: the right cleanup order
Do not launch three new campaigns while the conversion action is still a newsletter. Sequence beats volume. Creative and extra geos can wait until the signal is honest.
-
Tracking first. Start with the conversion inventory in Google Ads and GA4. List every primary action. If two of them can fire on the same session, you are training the algorithm to farm the easier one. Align UTM parameters so a Gold Coast paid click does not show as “direct” after a redirect.
-
Match types and negatives second. Dump 14 days of search terms into a sheet and tag each row: bookable, research, job-seeker, competitor, or junk. That tagging exercise is more useful than another RSA test.
-
Landing speed and message match third. Dedicated pages, not homepage hope. Check INP and LCP on mobile from an Australian location, not a US lab. Snippets that over-promise (“instant approval”, “free forever”) attract the wrong click before the page even loads.
-
Structure and PMax last. Once those three layers hold, you can broaden match types or give PMax more budget without teaching it to love brochure hunters.
What can wait: new Display experiments, extra remarketing lists, and bid strategy theatre. What cannot wait: a primary conversion that sales will not accept as a real lead. If sales rejects 40% of form fills on sight, no amount of AI Max will save the ROAS story.
One more mid-course check: if in-platform CPA and CRM-accepted leads have diverged for more than two weeks, pause scale. You are not “in learning”. You are paying for a metric sales already stopped believing. That is the moment to fix events, not to add Canberra as a new geo.
The Perth B2B account that stopped celebrating demo forms
A head of growth at a Perth B2B software company had a Search + PMax mix that printed 62 demo forms in a month at a CPL the CFO liked. Sales qualified nine. The rest were students, consultants fishing for pricing, and people who wanted a free migration they were never going to buy. Pipeline from Search was basically a rounding error while tCPA sat on every form submit, including the white paper.
The one change was not a new keyword list. They demoted the “Book a demo” thank-you event from primary to secondary and imported only CRM-accepted opportunities (sales marked them within 48 hours) as the primary conversion. Same ads, same landing URL, same budget band. Smart Bidding had fewer conversions to chew on for 19 days, which felt ugly in the UI. Then demo-to-opportunity rate moved from 14.5% to 41%, and the team stopped arguing about “lead volume” in the Monday stand-up.
They also pulled brand out of PMax so branded ROAS stopped laundering the prospecting number - PMax had been most of spend, and most of the ROAS was their own brand queries. That was the second change, and it only worked because the conversion definition was already honest. Once brand Search was split, prospecting looked expensive overnight, which was the first honest week they had. If they had restructured first, they would have reset learning on a still-broken signal.
If your own CRM and Ads numbers have not been in the same room for a quarter, you already know how this story ends. Tightening the event is faster than hiring another SDR to work junk. For ongoing bid, negative, and landing alignment without living in the UI, the same SEM / Google Ads work is the operating layer, not a one-off audit PDF.
FAQs
Should we pause Smart Bidding if lead quality is poor?
Not as the first move. If the primary conversion is a cheap form, Manual CPC will still buy cheap forms, just with more of your time. Change the conversion definition and the landing offer, give the bid strategy a few dozen clean conversions, then judge it. Flip-flopping bid strategies is how AU accounts stay stuck in learning.
Is Performance Max always bad for lead quality?
No. It is bad when it is the only campaign, mixed with brand, and scored on last-click forms. Used with Search for high-intent queries, brand controls, and a sales-accepted conversion, it can carry scale. Used as a black box for homepage traffic, it fills the CRM.
How many negatives is enough for an Australian Search account?
Enough to cover careers, education, DIY, free, and competitor brand if you are not bidding that fight. Shared lists beat campaign-by-campaign folklore. Review search terms on a cadence, not after every expensive day. Over-negating commercial synonyms will starve volume. Under-negating is the more common sin - it starves sales while CPL looks fine.
Do we need a new landing page for every RSA?
No. You need a page per intent cluster, with the ad promise in the H1, proof, and a form that asks for what sales actually uses. Five near-duplicate pages with different headlines and the same 14-field form will not save quality. Speed and message match will.
What should we show the board instead of CPL?
Cost per sales-accepted lead or cost per booked job, in AUD, with GST called out so finance is not reconciling in Slack. Form volume can sit as a diagnostic. If those two numbers move in opposite directions, you are still counting the wrong conversion.
Putting it to work
Execution sprint
This week
- Pull 30-90 days of performance for google ads mistakes that fill forms but not jobs (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
-
Export the last 60 days of Google Ads conversions and mark each primary action as sales-ready or vanity.
-
Pull search terms for your top three spend campaigns and tag junk clusters (jobs, free, course, DIY).
-
Confirm the paid landing URL is not the homepage for at least your highest-intent ad group.
-
Run the Core Web Vitals checker on that URL from a mobile view and note LCP and INP.
-
Build UTMs with the UTM link builder so the next sprint of spend is reportable in GA4 and the CRM.
Next 30 days
-
Make one sales-accepted event the only primary conversion and demote the rest.
-
Split brand Search from PMax if brand is inflating ROAS.
-
Ship one intent-matched landing page and watch form quality, not just conversion rate.
-
Import qualified calls if the phone still books the work.
-
Freeze major structure changes until you have a clean month of signal.
Pull your last 60 days of Google Ads leads by conversion action and sit with sales on which ones became booked work. If the primary conversion is still a form that nobody will take, HeyLead can take over the conversion design, search-term fencing, and landing match that stop Australian accounts from buying junk enquiries. Chat with us on WhatsApp
Free marketing audit, or reach Martin directly:
Get a free audit Connect on WhatsApp · [email protected]