Google Ads audit / Six final findings

Audit Findings.

Six findings on where performance slipped, where campaign routing is unclear, and what to test, hold, or verify next.

Executive account context

Park's account is functional and relatively close to its stated target.

Search recorded $20,684.26, 141 credits, and $146.70 per credit. The account's main opportunity is tighter control and clearer testing across a small number of high-value areas.

A Google Ads conversion credit is not automatically a unique person, qualified opportunity, booked job, completed job, or revenue event.

$21,749Total spend

145Conversion credits

$149.99Account cost per credit

$146.70Search cost per credit

Method and evidence definitions

Observation. Supporting Data. Action.

Realized Park delivery, current configuration, estimated market demand, and test hypotheses are kept separate. Current settings are September 3 snapshots unless a dated history proves otherwise.

01Observed: directly present in Park's exports or current configuration.
02Calculated: derived directly from observed values.
03Reported: stated by Park's public site or another named source.
04Estimated: modeled by Keyword Planner, not realized Park traffic.
05Hypothesis: a proposed change that still requires a test.

Performance recovery

Restore Hot Water Heater performance through one controlled recovery test.

A previously productive campaign lost performance while several variables changed together. The first intervention belongs in the dominant Hot Water Heater traffic.

Cost per conversion creditJune to August 2026

$120.63

June

$107.41

July

$469.70

August

July 1990 phrase keyword pauses

July 24Two budget increases

July 2920 destination changes

August 3 and September 1Budget moved to $150, then $50

Source: Park Plumbing Google Ads, Campaign report and Change history, June 1 to September 1, 2026.

Observation

The Hot Water Heaters campaign performed within target in June and July, then deteriorated sharply in August. Several keyword, destination, and budget changes overlapped the decline, preventing a reliable claim that any one edit caused it.

Supporting Data

Full-period campaign performance was $2,066.38 spent, 12 conversion credits, and $172.20 per credit. At least 114 material configuration changes were recorded across the deterioration window and immediate response: 90 phrase keyword pauses, 20 destination changes, and four budget changes. The September 1 budget reduction followed the August decline and cannot have caused it.

The Hot Water Heater ad group generated 1,302 impressions, 30 clicks, $1,184.47 spend, six credits, and $197.41 per credit. Tankless recorded $478.92, four credits, and $119.73 per credit.

Hot Water Heaters lost 52.02% of impression share to rank and 0.44% to budget. More budget was not the obvious remedy.

Change history and performance timeline

Action

Classify visible demand first. If replacement-and-installation traffic has sufficient separable exposure, add one replacement-and-installation responsive Search ad challenger in the dominant Hot Water Heater ad group.

Full test controls and decision rule

Classify visible demand into repair, replacement, installation, Tankless, manufacturer service, pricing, retail, and unclear intent.
Keep the existing advertisement eligible and use the same landing page.
Hold budget, bidding, keywords, negatives, schedule, conversion goals, geography, and Tankless fixed.
Treat the comparison as directional because volume is sparse.
If exposure cannot support a useful comparison, document a staged rollout or defer the test.
Retain only when mature cost and available lead-quality evidence improve. Otherwise revert, extend unchanged, or defer.

The July 29 destination change, the page, and the small retail-search issue are not presented as the cause of the August decline.

PRIORITY

Priority 1 of 6. Recover the previously productive campaign with one controlled, directional test.

Query routing

Give the two highest-value Near Me searches one controlled route.

The two Search campaigns control almost all observed Near Me delivery, while Performance Max adds a small third route that must remain visible in monitoring.

Observed Search campaign routesJune 1 to September 1, 2026

Regular Plumber

Spend$1,688.13

Credits9.67

Cost per credit$174.57

Plumber (BM)

Spend$372.83

Credits5.33

Cost per credit$69.95

Source: Park Plumbing Google Ads, Search terms with triggering keyword, June 1 to September 1, 2026.

Observation

The literal searches plumber near me and plumbers near me were divided mainly between regular Plumber and Plumber (BM), even though the campaigns use different bidding targets and reported materially different historical costs.

The campaigns do not bid against each other in the same auction. Google selects one eligible route.

Supporting Data

Regular Plumber recorded 645 impressions, 32 clicks, $1,688.13 spend, 9.67 credits, and $174.57 per credit. Plumber (BM) recorded 733 impressions, 13 clicks, $372.83 spend, 5.33 credits, and $69.95 per credit.

The combined family represented approximately 9.5% of account spend and 10.3% of account credits. The advertisements and destinations were identical.

Performance Max added 16 impressions, one click, $0.95 spend, and no credits across the same two literals. The exact routing correction controls Search delivery, not Performance Max.

Near Me routes across Search and Performance Max

Action

Stage one controlled exact route for the two literal searches in Plumber (BM) / Near You.

Routing controls and decision rule

First, create and enable [plumber near me] and [plumbers near me] in Plumber (BM) / Near You.
Keep regular Plumber eligible as fallback for three to seven days while confirming the intended route can serve consistently.
Track visible route integrity plus total literal-family impressions, clicks, spend, and mature conversion credits. Require at least 95 percent visible route integrity, without allowing that percentage to hide a material reduction in total Near Me exposure.
Add the same two exact negatives to regular Plumber only if the intended route has proved it can serve and meaningful leakage continues.
Preserve surrounding broad and phrase discovery. Hold advertisements, pages, goals, geography, bidding targets, and budgets fixed.
Retain only when route integrity, total exposure, mature economics, and available quality pass. Revert if traffic becomes stranded, exposure falls materially, tracking fails, or quality deteriorates. If routing works but evidence remains sparse, extend unchanged for four to six weeks.

The historical difference was observational. The $69.95 result is not a guaranteed future rate.

PRIORITY

Priority 2 of 6. Prove the exact route can serve before closing the fallback route.

Emergency coverage

Build deliberate emergency-plumbing coverage after confirming operating economics.

Emergency intent currently enters generic campaigns and generic advertisements rather than a deliberate emergency search-to-ad-to-page path.

Current operating questionConfiguration captured September 3, 2026

24/7

Park's public website reports emergency availability.

6 a.m. to 8 p.m.

Enabled nonbrand Search campaign schedule.

Source: Park Plumbing public site evidence and current campaign settings captured September 3, 2026.

Observation

Park publicly advertises 24/7 emergency availability, while enabled nonbrand Search campaigns run from 6:00 a.m. to 8:00 p.m. The schedule difference is a business and media-control question, not automatic proof of a defect.

Supporting Data

Reported Search Terms containing emergency produced 118 impressions, three clicks, $243.50 spend, two conversion credits, and $121.75 per credit.

The three clicked searches entered Plumber (BM) / Near You, Plumber / Annapolis, and Plumber / Near You. None of the exported responsive Search ads explicitly used emergency, 24 hour, 24/7, or same day language.

Three clicks are too few to treat the observed two-credit result as a forecast.

Emergency search delivery and routes

Current schedule rows; metrics use the audit window

Action

Launch one capped emergency-plumbing routing test, with structure determined by verified operating hours.

Operating prerequisites and test structure

Confirm service radius, accepted emergency job types, answering coverage, dispatch capacity, response standards, opportunity economics, approved claims, and approved maximum loss.
Use an Emergency Plumbing ad group inside Plumber (BM) when the test runs only during the current schedule.
Use a separate capped campaign only when Park approves an after-hours test requiring a different schedule, budget, bidding target, and control structure.
Do not use 24/7, response-time, pricing, licensing, or guarantee claims until each is operationally verified.
Review search terms, call handling, accepted opportunities, and mature cost weekly for 4 to 6 weeks.

Park can offer 24/7 service while intentionally declining overnight paid traffic. The selected Keyword Planner estimates are overlapping and are not added together.

PRIORITY

Priority 3 of 6. Confirm operating economics before launching deliberate emergency coverage.

Eligibility policy

Reconcile negative keywords with Park's current service policy.

The risk is not the size of the negative system by itself. The risk is that broad one-word exclusions can suppress a service Park currently accepts.

Universal negative listCurrent configuration captured September 3, 2026

3,451

Universal rows

1,571 exact1,214 broad666 phrase

Conditional conflict

Park publicly advertises Garbage Disposal Installation and Repair, while shared and direct negatives suppress garbage-disposal searches.

Source: Universal negative list, direct negative report, and dated Park public-site evidence, September 3, 2026.

Observation

Park has extensive negative-keyword coverage, but list size is not the defect. The potential problem is a mismatch between business policy, website promises, and Google Ads eligibility rules. The strongest reproduced example is garbage-disposal service.

Supporting Data

The current Universal negative list contains 3,451 rows, including 1,214 broad negatives. 1,213 of the 1,214 broad entries are one word. Another 108 direct campaign and ad-group negatives exist outside the shared-list attachment rows.

The Universal list contains broad garbage and disposal. Enabled Plumber (BM) contains direct exact negatives for garbage disposal repair, garbage disposal installation, and garbage disposal not working.

Universal garbage exclusions

Direct garbage-disposal exclusions

Public garbage-disposal service offer

Action

Confirm current service policy. If Park accepts garbage-disposal work, remove or narrow the relevant shared and direct exclusions. If Park does not accept the work, preserve the exclusions and correct the website.

Policy controls and monitoring

Confirm policy for garbage-disposal installation, repair, and diagnostic jobs.
Confirm policy for price-intent searches and the Rinnai, Ruud, and State brands.
Keep a one-word broad negative only when every query containing that literal term is invalid across every attached campaign.
Record the owner, business rationale, scope, approval date, and next review date.
Release one policy family at a time and monitor newly eligible search terms, spend, conversion credits, and agreed opportunities.

Exact negatives [water heater], [hot water], and [tankless] are not presented as errors. Negative inventory does not prove recovered traffic or savings.

PRIORITY

Priority 4 of 6. Resolve the reproduced service-policy conflict one policy family at a time.

Rank and relevance

Improve search-to-ad relevance before increasing or consolidating budgets.

Search visibility was constrained primarily by rank rather than budget, while the eligible responsive Search ad inventory offered little controlled creative contrast.

Search visibility diagnosticJune 1 to September 1, 2026

31Eligible RSAs

31Eligible cells

2Descriptions per ad

4Unique descriptions

Source: Park Plumbing Google Ads, Search campaign impression-share and responsive Search ad exports, June 1 to September 1, 2026.

Observation

Park's Search visibility was constrained primarily by rank rather than budget, while the eligible responsive Search ad inventory offered little controlled creative contrast.

Supporting Data

Search impression share was 31.20%. Lost impression share to rank was 65.55%, while lost impression share to budget was 3.25%.

The creative export showed 31 eligible responsive Search ads across 31 eligible ad-group cells, two descriptions in each ad, and only four unique descriptions across the export.

Ad Rank includes bids, competition, expected response, relevance, page experience, and asset impact. The evidence does not prove that ad copy alone caused the rank loss.

Rank loss versus budget loss

Responsive Search ad inventory

Action

Hold broad campaign budgets and bidding targets during the initial Near Me, Hot Water Heater, and emergency tests. Do not make an account-wide budget increase or consolidate campaigns during the first test window.

Test sequence and decision rule

Improve query routing and service-specific messaging first.
Evaluate mature conversion cost, conversion-action composition, available accepted-opportunity quality, click response, and rank loss together.
If a specific traffic cell improves and remains constrained by rank, review bid competitiveness for that cell.
Increase budget only when the recipient is genuinely constrained by budget and comparable mature opportunity economics show that incremental spend is likely to outperform the traffic displaced elsewhere.
Review Plumber and Plumber (BM) consolidation later, after routing, targets, goals, service mix, and opportunity economics are comparable.

Consolidation can still help later. Selected Keyword Planner volume does not prove the correct campaign boundary.

PRIORITY

Priority 5 of 6. Improve routing and relevance before changing aggregate budget structure.

Conversion quality

Confirm conversion actions are comparable before reallocating budget.

This account-wide validation finding tests whether the three bidding actions represent consistent, sufficiently deduplicated opportunities.

Campaign conversion mix145 total credits

Calls from ads72

First Time Phone Call40

Form Capture33

145 total campaign conversion credits across the three actions.

Source: Park Plumbing Google Ads, campaign and conversion-action exports, June 1 to September 1, 2026.

Observation

The account is close to its stated $145 target only if Calls from ads, First Time Phone Call, and Form Capture represent reasonably consistent and sufficiently deduplicated business opportunities. The supplied evidence does not prove that they are misconfigured, but it also does not establish that every credit equals one unique qualified opportunity.

Supporting Data

The account spent $21,749.14 and recorded 145 campaign conversion credits: 72 Calls from ads, 40 First Time Phone Call, and 33 Form Capture. That equals $149.99 per credit against the stated $145 target.

The evidence does not prove a tracking or lead-quality failure. It leaves source, optimization status, counting rules, repeat behavior, campaign-goal inclusion, attribution settings, conversion delay, and potential cross-action duplication to validate.

Current conversion configuration and All conversions counts. The 145-credit campaign mix is shown separately above.

Action

Confirm whether Calls from ads, First Time Phone Call, and Form Capture match the agency's accepted-opportunity definition and are sufficiently deduplicated.

Measurement validation

Confirm whether the $145 target measures a named Google Ads action, raw contact, inquiry, accepted opportunity, appointment, or another outcome.
Review their sources, optimization status, counting rules, repeat behavior, campaign-goal inclusion, attribution settings, conversion delay, and potential cross-action duplication.
If the existing actions already align with the accepted-opportunity definition and are sufficiently deduplicated, preserve the current setup.
If they do not align, introduce one consistently defined Qualified Lead or Accepted Opportunity action as secondary observation data.
Keep current bidding stable until the secondary action has enough mature volume for a separate decision.

This is the weakest finding as a direct performance opportunity. It is a validation finding, not a claim of broken tracking or poor leads. It does not require revenue attribution, and missing revenue is not a defect when revenue is not the operating objective.

PRIORITY

Priority 6 of 6. Validate comparability without disrupting normal account management or current bidding.

Evidence appendix

Every visual is tied to Park's local evidence.

Historical performance uses June 1 to September 1, 2026. Current configuration uses September 3, 2026. Keyword Planner values remain estimates and are not presented as realized Park traffic.

Each recommendation is tied to a dated source and a specific decision rule.

Campaign report.csv and Park_Live_Campaign_Goals_2026-09-03.csv, campaign performance and conversion-action totals.
Campaign settings report.csv, current budgets, targets, schedules, and status.
Park_Live_HWH_Change_History_2026-06-01_to_2026-09-01.csv, Hot Water Heater change chronology.
Park_Search_Terms_With_Triggering_Keyword_2026-06-01_to_2026-09-01.csv, Near Me routing.
Search terms report.csv, emergency delivery and query paths.
Park_Negative_List_Universal_2026-09-03.csv and Negative keyword report.csv, negative inventory and direct exclusions.
Park_Search_Campaign_Impression_Share_2026-06-01_to_2026-09-01.csv, rank and budget constraints.
Park_Ads_RSA_2026-06-01_to_2026-09-01.csv, responsive Search ad inventory.
park-public-site-service-proof-2026-09-03.md, public emergency and garbage-disposal statements.