Inherited Google Ads Account Rescue: From A$1,700/month, Zero Conversions, Dead-URL Ads to a Structured Intent-Only Account in 14 Days

A Melbourne emergency restoration business - established operator, IICRC-certified, 24/7 dispatch from a south-east suburbs HQ - came to us after spending more than a year with a prior digital agency running their Google Ads. The trailing 30 days at intake showed A$1,702.71 spent and zero primary conversions. The bidding strategy was Maximise Clicks with a 1-cent placeholder ceiling. Multiple ads were disapproved and pointing at URLs that had never existed on the client’s website. The conversion tracking layer was a graveyard of orphaned, removed actions. Approximately 14% of spend was going to queries that could not convert by design (mould remediation, ceiling repair, DIY research, competitor brand searches, out-of-state geographies including Florida, USA).

In 14 days we restructured the account around a single rule - Google Ads captures emergency intent only, within dispatch radius - and rebuilt the keyword, negative, bid, and tracking layers to enforce it.

The client’s identifying details, and the prior agency’s identity, have been anonymised. All metrics are verified through the Google Ads UI and an append-only change log.

This case study walks through what we inherited, the intent-split decision, the 14-day cleanup, the daily review cadence that compounds from week three onward, and what real Google Ads management looks like when run to a written rule rather than ad-hoc.

What We Inherited From the Prior Agency

This was not a new account. The campaign had been running under prior-agency management for over a year on the same single campaign structure (Search channel, single ad group, single Responsive Search Ad) before we took over.

The trailing 30 days at intake:

Metric	Inherited from prior agency (30 days)
Total impressions	2,913
Total clicks	61
Total cost	A$1,702.71
Average CPC	A$27.91 (above market clearing rate)
Click-through rate	2.09%
Primary conversions	0
Search Impression Share	40.1%
Lost IS - rank	43.6% (Quality Score driven)
Lost IS - budget	16.3%
Search-term waste	~A$230 / month (~14% of spend)

That is the explicit numerical baseline. The structural debt underneath it was worse:

1. Disapproved ads pointing at a URL that did not exist. Three of the four ads still attached to the account were DISAPPROVED with destination URL /flood-cleaning/ - a path the prior agency had used in ad copy but never created on the client’s website. The disapprovals had not been actioned. The ads sat dormant, dragging account-level Quality Score history.

2. One live ad pointing at the wrong domain entirely. A fourth inherited ad had its Final URL set to totalcleaningmelbourne.com.au - the client’s sister business, not the flood damage service the campaign was meant to drive traffic to. Any clicks would have routed to the wrong landing page, the wrong offer, and a conversion tag that did not match the campaign’s tracking setup.

3. Conversion tracking was a graveyard. Multiple inherited conversion actions sat in REMOVED status - “Click to call”, “Call (0448 888 165)”, and a named action labelled in the prior agency’s internal naming convention - alongside two HIDDEN GA4-imported actions that had been disabled. Of the conversion actions still ENABLED, none had fired in the trailing 30 days. The conversion-tracking pipeline existed on paper but produced no signal in the dashboard.

4. Bidding was effectively without a ceiling. The bidding strategy was Maximise Clicks with cpc_bid_micros = 10000 - a 1-cent placeholder that the prior agency had never replaced with a real cap. With no conversion data feeding Maximise Clicks, the strategy had nothing to optimise for and was inflating CPC toward A$28 to A$39 average - well above the historical market clearing price for emergency flood and water damage terms in Melbourne.

5. The keyword set was loose phrase-only with research contamination. 45 ENABLED phrase-match keywords - several pulling research, “near me” outside the service radius, and competitor-brand queries. Zero EXACT-match keywords. The negative keyword list had not been updated in months: 40 ad-group-level EXACTs covering some specific competitor brands but missing the entire category of research and shopping signals (how to, cost, price, cheapest, best, top, vs, review, company, companies, business, wikipedia, guide, tips, definition, what is, yourself, quote, estimate, diy).

6. Legacy ad groups never deleted. Two legacy ad groups - “Flood Restoration” and “Carpet Restoration” - persisted on the account from a prior structural reorganisation but were no longer serving traffic. They held the dead-URL disapproved ads and were dragging on the account’s structure history.

7. The single primary conversion goal was a phone-call action that had never fired. Across 30 days and 61 clicks, the account showed zero conversions on its primary goal. The conversion tag was wired correctly in the page source. The conversion action was ENABLED and primary. The CSP allowed all the conversion domains. But the verification step - testing the full pipeline end-to-end - had been missing for the entire prior engagement.

The headline number is the one most clients see: A$1,702 spent in 30 days, zero conversions, no clear path forward. The under-the-hood number is worse: more than a year of prior-agency management compounding the same loose phrase keywords, the same broken tracking, the same dead-URL ads, the same Maximise Clicks bidding without a real cap.

Before: The Five Compounding Failures Underneath The Number

The account did not have one big problem. It had five mid-sized problems that compounded.

1. No written intent rule. Without a rule for what Google Ads should capture, every keyword decision had been made on auction availability rather than commercial fit. The result was loose phrase keywords like “restoration service”, “leak restoration services”, and “service restoration” - each individually defensible, each in practice matching mould, ceiling repair, Florida (USA) geographies, and competitor brand queries.

2. Maximise Clicks bidding with no conversion data. Google’s “Maximise Clicks” strategy treats every click as equal value. With zero conversion data feeding it, the strategy had nothing to optimise for and was inflating CPC toward A$28 to A$39 average - well above the historical market clearing price for this category.

3. Negative keyword list inadequate. Forty EXACT-match negatives covered some specific competitor brands but missed the entire category of research and shopping signals (how to, cost, price, cheapest, best, top, vs, review, company, companies, business, wikipedia, guide, tips, definition, what is, yourself, quote, estimate, diy).

4. Conversion tracking installed but invisible. The conversion tag was wired correctly in the page source. The conversion action was ENABLED and primary. The CSP allowed all the conversion domains. But across 30 days and 61 clicks, the account showed zero conversions. The wiring looked fine on paper - the verification step was missing.

5. One ad group, three intents. Emergency, commercial, and suburb-targeted intents were all routing through a single ad group with a single ad copy set. The ad relevance score on each individual query was diluted as a result.

The Cleanup (Days 1 to 14)

The work split into three stages, each gated against the prior one’s verification.

Stage 1 (Days 1 to 4): Intent rule, bid strategy, foundational pause

The intent split rule was written first as a one-page document and signed off before any keyword change:

• Google Ads = emergency only - active state, “emergency”, “24/7”, “now”, “near me” within dispatch radius, suburb+service combos
• SEO = everything else - comparison shopping (“best”, “company”, “professional”, “specialists”), DIY (“how to”), definitional, out-of-radius geography, planning (“quote”, “estimate”), past-tense rebuild work, mould (not time-critical), burst pipe (plumber territory before us), storm damage (seasonal + news contamination)

Bid strategy switch: Maximise Clicks → Manual CPC at A$15 cap. The cap is tied to the historical average CPC of high-intent retained keywords, not the inflated average of all keywords. Enhanced CPC disabled to prevent the algorithm from over-bidding above the cap.

Pre-cleanup pauses (8 keywords): the eight highest-spending phrase keywords with the worst search-term contamination - “restoration service” (A$456.89 / 30d, matching amtec / restorx / mould / attic / roof), “damage restoration companies” (A$200.41), “leak restoration services” (A$199.46, “leak” pulls plumber and ceiling), “service restoration” (A$134.60), “water damage home restoration” (A$55.40, matching ceiling repair and Florida queries), and three others.

Combined Stage 1 effect: ~A$1,030 per month of waste removed at the keyword layer. The campaign held at the same daily budget but stopped paying for queries that could not convert by design.

Stage 2 (Days 4 to 7): Negative keyword catalogue + Stage-1 expansion

48 BROAD-match negatives across four buckets:

• Research and shopping (23): quote, estimate, how to, how much, diy, cost, price, cheapest, best, top, vs, review, reviews, company, companies, business, wikipedia, guide, tips, definition, meaning, what is, yourself
• Wrong service (16): burst pipe, plumber, plumbing, sewer, sewage, storm, mould, mold, ceiling leak, ceiling repair, roof, fire damage, smoke damage, timber restoration, attic, balcony
• Wrong geo (9): sydney, brisbane, canberra, perth, adelaide, gold coast, jupiter, florida, usa
• Plus 18 EXACT-match competitor brand negatives at ad-group level (Paul Davis, Mould Busters, Perco Cleaning, Restorx, Steamatic, Victoria Hygiene, Elite Roof, Storm Restoration Services, FRM, Cowan Restoration)

Stage-1 expansion: 23 additional research and shopping phrase keywords paused (best water damage restoration company, water damage cleanup company, flood damage companies, professional water damage restoration, water damage restoration specialists, IICRC water damage restoration, etc).

Negative consolidation: EXACT-match competitor blocks were migrated to BROAD-match where the brand had multiple variants (steamatic BROAD blocks “steamatic”, “steamatic australia”, “steamatic near me”, “steamatic melbourne” simultaneously). Geographic negatives expanded from capital cities only to include state names, abbreviations, secondary cities, and Victorian regional towns outside the 2-hour dispatch radius.

Final negative state: 122 negatives total (86 BROAD, 36 EXACT) across campaign and ad-group level - up from 41 (1 campaign + 40 ad-group EXACT).

Stage 3 (Days 7 to 14): EXACT-match expansion + tracking verification

30 EXACT-match emergency-intent keywords added in a single batch:

• 10 generic emergency: [emergency flood restoration melbourne], [24 hour flood restoration melbourne], [24/7 water damage melbourne], etc
• 6 active service / “near me”: [water damage restoration near me], [flood restoration near me], [wet carpet drying melbourne], [flood restoration melbourne], etc
• 11 suburb + service combos within 2-hour dispatch radius from Hallam HQ + CBD coverage: [flooded carpet drying camberwell], [flooded carpet drying footscray], [flooded carpet drying st kilda], [flooded carpet drying brighton], [flooded carpet drying toorak], etc
• 3 commercial: [commercial flood restoration melbourne], [emergency commercial flood melbourne], [office flood emergency melbourne]

EXACT match was chosen over PHRASE for the literal high-intent queries because:

1. EXACT typically wins higher Quality Score than PHRASE on literal matches (less ad-keyword-LP relevance dilution)
2. EXACT typically wins auctions at lower CPC than PHRASE because of tighter relevance
3. For the suburb+service patterns, EXACT captures the specific suburb-search query directly without relying on broader phrase matching

Conversion tracking verified end-to-end. Programmatic verification confirmed the tag wiring, the conversion action enabled and primary, the CSP permitting conversion domains, and the GA4-side click event firing site-wide. Where browser-side tracking was at risk from Consent Mode v2 or tracking-prevention, an imported-conversion fallback (GA4 → Google Ads server-side import) was documented as a redundant secondary action.

The Daily Cadence: Where the Compounding Comes From

The 14-day cleanup is the foundation. The compounding comes from what runs after it.

Every weekday morning, a daily search-term sweep runs against the previous day’s data:

• Wrong-intent terms flagged for negation (with thresholds: spend over A$30 with 0 conversions, or 70%+ wrong-intent ratio across recent matches)
• Right-intent terms above an impression threshold flagged for EXACT-match expansion (locks in Quality Score for the literal query)
• Existing keywords above a spend ceiling with 0 conversions flagged for pause review
• Geographic anomalies (out-of-radius suburbs generating unexpected impression share) flagged for geo-negative addition
• Bid and budget review - rank-lost IS and budget-lost IS reviewed against the current bid ceiling and daily cap

Every mutation executed is recorded in an append-only change log with: date, category, before-state, change made, reason, expected impact, verification, and (filled later) actual impact. The log is auditable - any keyword pause, negative addition, bid change, or budget change can be traced back to the data signal that triggered it weeks or months later.

A representative two-week sample from the change log during the steady-state phase:

Date	Mutation	Reason
Day 17	NEGATE “fen australia” BROAD	Unknown entity, not emergency flood intent
Day 17	NEGATE “black water remediation” BROAD	Wrong service (sewage/blackwater is SEO-only per intent rule)
Day 17	NEGATE “flood restoration australia” EXACT	National search intent, wrong geo
Day 18	EXPAND “carpet water extraction” EXACT	2 clicks A$29.56 spend via PHRASE, clear emergency intent, lock QS
Day 18	NEGATE “mould remediation” BROAD	A$48.39 in 14 days via “service restoration” PHRASE, wrong-service intent per intent rule
Day 21	PAUSE “carpet flooded with water” PHRASE	A$104.61 / 14 days / 0 conversions; right-intent matches already covered by EXACT [carpet water extraction]
Day 21	NEGATE “steamatic” / “paul davis” / “xtreme clean” BROAD	Competitor brand leaks identified in search-term sweep
Day 23	NEGATE “footscray” BROAD	Western Melbourne, ~45 min from HQ, outside SE Melbourne service area
Day 23	PAUSE “home water damage cleanup” PHRASE	35 impressions yesterday, 0 clicks, search intent too broad

Each line is a real decision against real signal. Each line has a paragraph of reasoning behind it in the underlying log entry.

Bid Ceiling Discovery: Where The Money Is Won

The Manual CPC bid cap was deliberately conservative on day one (A$15) because the conversion data did not yet exist to justify higher. As clicks-without-conversion accumulated, the cap was raised stepwise on a documented rationale:

Day	Bid cap	Rank-lost IS	Reason
1	A$15.00	(baseline)	Conservative starting cap, below historical average CPC
8	A$25.00	24.3% → falling	Historical avg CPC was A$25.38; A$15 was losing competitive auctions
9	A$30.00	28.9%	Incremental raise to test rank dimension
10	A$40.00	~30%	Most clicks clustering at A$25-30 ceiling, raise to win more weekend auctions
11	A$50.00	(settling)	Weekend push to win first paying conversion; budget raised to A$250 in parallel

Each raise was gated against the prior raise’s settling window (two to three days minimum). The bid raised when rank-lost IS justified it; the budget raised when spend was hitting the cap.

This is a real, dispatchable bid-discovery process - not a fixed-fee “managed” account where the bid never moves.

What Wasn’t Done (And Why)

• No new ad copy. The existing Responsive Search Ad had GOOD ad strength and APPROVED status. Ad copy is a separate optimisation lever; we addressed keyword and intent first because the auction inputs determine 80% of performance. Ad copy refresh is on the backlog for week 4 onward.
• No campaign restructure. The existing campaign had one ad group; splitting into emergency / commercial / suburb-targeted is on the backlog for week 4 onward once the EXACT-match keyword adds have settled. Premature restructure would have cut Quality Score history.
• No image / lead form / promotion extensions. These are CTR-lift extensions on the backlog for the next sprint; they do not affect intent quality, which was the binding constraint.
• No new geo set. The existing 2x PROXIMITY 20km radius geo targeting was kept; switching to a “Greater Melbourne” constant is on the backlog (more robust but more brittle in the migration step).

What’s Compounding From Here

Three mechanisms continue to compound from the 14-day cleanup baseline:

1. Quality Score recovery. With the auction no longer weighting wrong-intent matches, ad-keyword-landing-page relevance tightens. Effective CPC drops below the bid cap on the same retained keywords - typical recovery window two to four weeks. This compounds without any further account changes.

2. EXACT-match coverage compounds with each search-term sweep. New right-intent terms identified in the daily sweep get added as EXACT match keywords - locking in Quality Score for the literal query and reducing the auction’s reliance on PHRASE matching. The catalogue grows weekly.

3. The append-only change log is an asset. Six months from now, when a metric moves unexpectedly, the log allows an exact diff: what was the keyword set, the negative set, the bid cap, the daily budget, the conversion tracking state on that date? Most accounts have to reconstruct that from broken memory; this account has it on file.

The headline transformation in one line: an account spending A$1,700 a month with no written intent rule, Maximise Clicks bidding, 41 negatives, zero EXACT-match keywords, and zero conversions became an account running to a one-page intent rule, Manual CPC tied to economic ceiling, 122 negatives across both levels, 30 high-intent EXACT-match keywords, conversion tracking verified end-to-end, and a daily-review cadence with append-only change log - in 14 days.

Could Your Google Ads Account Do This?

This case study is a working example of what a Google Ads account looks like when it is run to a written intent rule, daily review cadence, and append-only change log.

If your account is on Maximise Clicks with no conversion data feeding it, has fewer than fifty negative keywords, has phrase-match keywords pulling research and comparison traffic, has conversion tracking installed but never firing, or has not been touched in months - the same methodology applies.

We work intent-first (every decision traces to a written rule), tracking-verified (we test the conversion pipeline end-to-end before signing off), and report-transparent (one-page weekly reports with mutations and reasons; no vanity reporting). We integrate Google Ads with SEO so research queries paused in Ads become blog briefs in SEO, and SEO ranking data feeds Ads negative decisions. We’ll beat your current ad management spend - send us your current bill. Email-first communication: written response within one business day, no phone calls.

See our Google Ads Management service or browse all services - or read our other case study on the SEO side of the same intent-split methodology.

Frequently Asked Questions

Q: Why is the client’s brand and exact industry segment anonymised?

The client requested anonymisation while we build out a portfolio of named case studies. Once five or more clients give written permission, we’ll publish those alongside verifiable brand and industry detail. The metrics here are auditable through the underlying Google Ads account and change log.

Q: 14% waste reduction sounds small - why is it material?

A$230 per month of waste compounds to A$2,760 per year on this account, and that is just the direct ad-spend recovery. The larger lift comes from Quality Score recovery on retained keywords, which drops effective CPC across the entire account by 15-40% over two to four weeks. On an A$1,700 monthly account, a 25% effective CPC drop is A$425 per month - the equivalent of doubling the budget without doubling the spend.

Q: How does this case study square with the price-beat guarantee?

The price-beat guarantee covers your current monthly website + hosting + SEO spend, and we’ll bundle Google Ads management at the same logic - send your current ad management bill and we’ll show you the line-by-line. Ad management fee is separate from ad spend; you pay Google directly for the spend, and us separately for the management. No managed-services markup on media.

Q: How much of this is replicable on a small account spending A$500-1000 per month?

Almost all of it. The intent rule, the negative-keyword catalogue, the bid strategy switch from Maximise Clicks to Manual CPC, the conversion tracking verification, and the daily search-term sweep all apply at any spend level. The difference at lower spend is that statistical signal arrives more slowly - bid changes need longer settling windows, and search-term volume is thinner. The methodology is the same; the cadence stretches.

Q: What if my account is brand new with no spend history?

Cleaner starting point. We write the intent rule first, build the keyword set EXACT-match-first, seed the negative-keyword catalogue from the start, and install Pixel + conversion tracking with end-to-end verification before the first ad serves. New accounts that ship with this foundation reach steady-state economics two to three weeks faster than accounts that retrofit it.

Q: What’s the realistic cost-per-lead target after this kind of cleanup?

Industry- and offer-dependent. For a Melbourne emergency restoration account with a A$25-40 average CPC and an 8-12% landing page conversion rate (typical for emergency intent on an optimised LP), realistic cost-per-lead lands in the A$200-400 band. For a non-emergency Australian small business in a less competitive niche, cost-per-lead can drop to A$50-150. The cleanup methodology does not change either floor; it gets you there faster and with the maths visible.

---

Want a free Google Ads account audit? Get in touch - we’ll review your account against the intent-split methodology and show you exactly where the waste is. No obligation, no sales calls.

All metrics independently verifiable through the Google Ads UI and the underlying append-only change log.