Pride Plumbing Services, LLC · Mount Gilead NC · callprideplumbing.com · Service area: Stanly, Richmond, Montgomery, Moore, Randolph counties plus Winston-Salem. The audit reflects the read-only window below. Changes now roll out as items clear the sign-off list on the hub, and each push gets logged there.

The reported $102 is inflated by overlapping conversion actions counting the same phone call several times, so we do not trust it and neither should you. Cost per booked job (a booked appointment in ServiceTitan) is the one number that means the same thing across every P3 OpCo, and $295 is Pride's baseline, locked as of this window, the number we manage to and beat from here. For context, the portfolio blended baseline across all three priority accounts is also $295 (170 booked jobs on $50,178).

Five findings, in the order they matter. The one-line version: fix tracking first so the bidding signal is honest, then feed the non-brand campaign, then tidy the keyword and competitor negatives. In that order.

• Conversion tracking is the constraint. 16 conversion actions are enabled and 10 are set as primary, including three or four different call trackers plus booked jobs, JVP, and completed jobs. One real phone call can count 3+ times. That is why the account reports 150 conversions at a $102 CPA when ServiceTitan shows 52 booked jobs at $295 per booked job. Bidding is Maximize Conversions, so Google is optimizing to the inflated number, not to real work.
• The non-brand campaign is budget-starved while it is the growth engine. Non-brand spent $9,864 (64% of spend), but its daily budget runs out before the demand does: its ads showed for only about 2 in 5 relevant searches, and about 3 in 10 of the missed showings were missed purely because the budget ran out. Brand ads showed for nearly every search for Pride's own name on a much smaller budget.
• Search-term waste is real but smaller and more targeted than a raw scan suggests. The genuine, clean waste is competitor-brand searches (about $445 with zero booked jobs). The bigger zero-conversion bucket, roughly $866, is in-service-area city searches (Asheboro, Southern Pines, Rockingham, Albemarle). Those are core demand, not waste, and should not be negated.
• Structure is basically sound underneath the clutter. Of 32 campaigns only 3 are active (non-brand search, brand search, LSA). The other 29 are paused legacy campaigns from prior agencies. The active non-brand campaign has a clean, service-themed ad group build. The weak spots are match type (80 broad-match keywords) and Quality Score (15 keywords at QS 3).
• Governance needs one owner. Five people or logins touched the account in July, including two shared Gmail marketing accounts, plus one auto-apply setting (Optimize ad rotation) left on. Nothing is on fire, but there is no single point of control, which is how the tracking got this tangled.

The 150 reported conversions are not real leads. They are the sum of overlapping conversion actions firing on the same events. Per ServiceTitan reporting, the cost per booked job is $295, which is not bad for plumbing, and it is the number to plan against. The problem is not the $295, it is that the account cannot currently see the $295, so it cannot optimize toward it. The ServiceTitan numbers below are as recorded in Google Ads through the offline import, pending reconciliation with ServiceTitan's own reports.

Note that brand reports more conversions (77.8) than non-brand (53.2) on 40% of the spend. That is the double-counting showing up: brand callers get logged across several call actions. It is not that brand is three times more productive than non-brand.

• Non-brand is the growth lever and it is capped. $500/day budget, but the budget runs out before the demand does: ads showed for only about 2 in 5 relevant searches, and about 3 in 10 of the missed showings came down to budget alone. Average daily spend ran about $329 against the $500 cap, so it is not pinned every day, but it still runs out on high-demand days. Raising the budget only pays off once the bidding signal is fixed, otherwise we would be buying more phantom conversions.
• Brand is efficient and near-saturated. Brand ads showed for 98% of searches for Pride's own name, and its budget almost never ran out. Brand does not need more money. Its current budget reads $100/day against $4,075 of 30-day spend, which implies the budget was higher earlier in the window and was cut around July 25. Worth confirming, but brand is healthy either way.
• LSA is a small, efficient third leg at $1,103 for 15 leads. Leave it, watch it.

The sequence matters: tracking fix first, then a measured non-brand budget increase once bids are chasing booked jobs. A budget increase before the tracking fix would scale the waste.

16 conversion actions are enabled, 10 are primary. Call trackers, JVP, completed jobs, and booked jobs are all primary at once, so a single call-that-becomes-a-job counts 3+ times. Reported conversions (150) are roughly 2.9x the ServiceTitan-reported booked jobs (52). The good news: the two ServiceTitan Booked Job actions already sum to exactly 52, matching ServiceTitan. The ServiceTitan reporting is already in the account. We just need to make it the goal and mute the noise.

Primary (counted, bid toward): ServiceTitan Booked Job (Originated from Call Extension) and ServiceTitan Booked Job (Originated from Website). Together these are "a booked job," split only by source. They sum to 52 for the window, which matches the ServiceTitan reporting and gives a clean $295 cost per booked job. At 52/month this clears the roughly 30-conversions threshold Maximize Conversions needs to bid well, so we do not need a second primary to prop up volume. Everything else goes secondary: still recorded and visible for insight, not counted in the bidding goal.

A note on the "one qualified call action primary" idea: if we added a call action to primary alongside booked jobs, we would re-introduce call-to-job double counting and push reported volume above 60, which muddies the $295 booked-job number. Because booked-job volume (52/mo) already clears the bidding threshold, the cleaner answer is booked-jobs-only as primary. If booked-job volume ever dips below about 30/mo, the designated backup is to promote ServiceTitan Calls Conversion Action (the de-duplicated upload call action) to primary as a bridge, accepting the tradeoff. Until then it stays secondary.

Actions #15 and #16 are Local Services Ads actions and live in the LSA product, which optimizes on its own. They do not pollute the search-campaign bidding goal, so they stay primary within LSA. The demotions above all concern the search campaigns. Note the odd current state: the two Booked Job actions, the ones that carry the ServiceTitan reporting, are partly set to secondary today while noisier actions are primary.

• Reported conversions drop from about 150 to about 52 per 30 days.
• Reported CPA rises from $102 to $295 per ServiceTitan-reported booked job ($15,350 / 52). The CPA number goes up on purpose. It is not a performance drop, it is the ServiceTitan reporting becoming visible.
• Maximize Conversions begins optimizing to booked jobs instead of raw call counts, so spend should tilt toward the keywords and times that produce actual jobs.
• Expect a 1 to 2 week relearn period where volume and CPA wobble. This is normal after a goal change. We will not react to it, and we will not change budgets during the relearn.

The corrected story first: this account is not bleeding money on out-of-area geography. Its geo targeting is granular (86 positive locations, 126 location exclusions, plus proximity radius on the active campaigns). The biggest zero-conversion bucket, about $866 of in-service-area city searches, is core demand and stays. The clean waste is local competitor-brand searches, about $445 with zero booked jobs. Only terms the data justifies are listed below. Nothing has been pushed.

Total zero-conversion spend in the pulled terms is $2,016 ($1,919 of it non-brand). One item flagged pre-audit as a possible routing bug, "pride plumbing near me" ($97, 0 conversions), verified fine: it serves from the brand campaign, which is correct. It is simply 3 brand clicks that did not convert in the window. No action needed.

Existing shared negative sets on the account: Negative keywords for Pride (2 members, attached only to two paused campaigns, effectively unused), Brand Terms (28 members, attached to non-brand so brand queries route to the brand campaign), Competitors (119 members, attached to brand + non-brand), and Master Negative List (103 members, attached to brand + non-brand). The active non-brand campaign already has Brand Terms, Competitors, and Master Negative List attached, so anything we add to those sets takes effect immediately on the campaign that matters.

Two more competitor names appeared with a conversion attached, so we hold them for review rather than auto-adding: "brandon plumbing co clemmons" (1 conv) and "hall plumbing reidsville" (1 conv on one variant). We recommend adding the brand token but confirming with Natalie first, since they did produce a booked job. Estimated recovery for the theme: $400 to $450 per 30 days, redeployable to in-area demand.

Optional, lower confidence: "handyman" (from "handyman toilet repair near me", $12, 0 conv). Pride is licensed; handyman-seekers may be price-shopping below Pride's service level. We recommend it as a broad negative on Master Negative List, but flag it for Natalie since some handyman searchers do convert to real jobs.

No clear wrong-service waste surfaced in this window. Pride's ad groups (water heater, gas line, hydro jetting, repiping, etc.) are matching relevant intent. Nothing to add today. We re-check after 30 days on the clean tracking signal.

jobs · hiring · salary · careers · apprentice · job openings

Housekeeping: the "Negative keywords for Pride" shared set (2 members) is attached only to two paused campaigns, so it protects nothing live. When we consolidate, we recommend either retiring it or repurposing it as the account's single working negative set attached to the active campaigns. Not urgent, but worth cleaning up so there is one obvious place negatives live.

Summary of what we push on approval: 17 competitor phrase negatives into "Competitors" (about $445/mo recovery, 2 held for Natalie's sign-off), 4 DIY/research broad negatives into "Master Negative List" (plus "handyman" pending sign-off), 6 employment broad negatives into "Master Negative List" (preventive), and the geo settings check (Presence-only, trim out-of-area positive locations). No city negatives.

Structure is basically sound underneath the clutter. The weak spots are match type, Quality Score on the money keywords, and governance. None of this is a crisis; it is the normal residue of an account that has passed through several agencies. The fix is one owner and one clean process.

• 32 campaigns, 3 active. Active: non-brand search, brand search, LSA. The other 29 are paused legacy campaigns (PMax, Display, Demand Gen, DSA, call-only, seasonal septic, "April-June", etc.) left behind by prior agencies. They carry zero spend now but clutter the account and make it hard to see what is real. We recommend archiving in a later phase, not urgent.
• Non-brand ad groups are well themed. General Plumbing, Water Heater Installs + Services, Gas Leak Detection, Hydro Jetting & Sewer Line, Kitchen Plumbing, Water Line Repair, Bathroom Remodeling & Toilet Repair, Commercial Plumbing, Repiping, Water Leak Detection, plus a "Core 2H" catch-all. A solid service-line structure to build on.
• Match type is loose. Across the active campaigns: 149 phrase, 80 broad, 18 exact keywords. 80 broad-match keywords on a budget-limited Maximize Conversions campaign that is being fed a corrupted conversion signal is a spray risk. Broad match trusts the bidding signal, and the bidding signal is currently wrong.
• Quality Score is weak on generic non-brand terms. Of the scored keywords: 15 at QS 3, 4 at QS 2, 1 at QS 1, 3 at QS 4. "plumbers near me" is QS 3. Brand keyword "pride plumbing" is QS 8. Low QS on the money keywords means Google charges more per click and shows the ads less often, which a budget increase alone will not fix.
• Multiple hands on the account in July. Changes came from natalie.bracamontez@p3services.com, farrukh.saleem@p3services.com, and two shared logins (p3servicesmarketingteam@gmail.com and p3servicesmarketingteam3@gmail.com). Shared Gmail logins make it hard to know who changed what, and they are a security exposure. Our suggestion for the Aug 3 conversation: consolidate to named accounts and agree one point of control (us, once the MCC link is approved).
• Auto-apply: only one setting is on, "Optimize ad rotation." Low-risk on its own, but combined with Maximize Conversions on a corrupted signal it means Google is auto-rotating ads toward the wrong outcome. We recommend turning it off when we do the tracking fix. All other auto-apply and target-CPA recommendation subscriptions are already paused. Good.
• Recommendation-applies: five recommendation applies in July (keyword removals via the Recommendations screen). Not harmful, but ad-hoc. Worth moving to a deliberate cadence.

Sequenced deliberately: the tracking fix comes first because everything downstream depends on an honest signal. This audit follows the Google Ads Success Formula (right traffic, right tracking, right bidding) and the conversion-tracking-first principle. The constraint here is tracking, so we stop and fix that before anything downstream.