Boom Media

Ad Operations · Agency Playbook

Meta & Google Ads — How to Run & Optimize Client Campaigns

A working playbook for managing paid ads for local-service clients: how each platform is structured, how to launch, how to optimize week to week, and the KPIs to hold yourself to. Built to review and keep updating as the platforms change.

Living doc · last updated 2026-07-22 · pairs with the Lead-Gen Engine build plan

Pick the platform by intent

The single most important call: Google captures existing demand; Meta creates it. Match the platform to where the customer is.

Use…	When the client…	Because
Google Search / LSA	sells something people actively search for (HVAC repair, roofer, dentist, "med spa near me")	You meet high-intent buyers at the moment of need. Expensive per click, high conversion.
Meta (FB/IG)	has an offer/promo, a visual result, or needs to build demand (body-comp scans, aesthetics, new location)	Cheap reach + interruption. You create want with creative before intent exists.
Both	has budget > ~$2k/mo and wants volume + efficiency	Meta fills the top of funnel, Google captures the branded/high-intent searches Meta generates.

Rule of thumb for local service: if there's search volume for the service, start on Google (or Local Services Ads); if the value is shown visually or driven by an offer, start on Meta. Most clients end up on both, with a landing page + tracking shared between them.

The funnel & the money math

Everything routes to a conversion — a lead form, call, or booking. No conversion event = no optimization = wasted spend.
Know the client's numbers before launch: average customer value, close rate, and target cost-per-lead (CPL). Target CPA should be well under (customer value × close rate).
Budget floor: a campaign needs enough spend to exit the learning phase (~50 conversions/week on Meta). Under ~$1–1.5k/mo per platform, results are noisy.

What you control vs. what the algorithm controls

Modern ad platforms automate targeting and bidding. Your leverage has shifted to three things the machine can't do for you:

Offer — the single biggest lever. A strong, specific offer beats any targeting tweak.
Creative — on Meta especially, creative is the targeting. Feed it many angles.
Data quality — clean conversion tracking (pixel/CAPI, GA4, enhanced conversions) so the algorithm optimizes toward real money, not clicks.

Account foundation

Set up once, correctly

Business Manager (business.facebook.com) — owns everything. For clients, use Business Asset sharing (partner access) rather than logins/passwords.
Ad Account + client's Page + Instagram + Pixel. One pixel per website/domain.
Conversions API (CAPI) — server-side tracking to survive iOS/ATT signal loss. Set it up alongside the pixel (via GTM server, Zapier, or a partner integration). This is not optional in 2026.
Domain verification + Aggregated Event Measurement — rank your 8 priority events (Lead / Purchase highest).
Special Ad Categories — if the client is Housing, Employment, Credit, or Social/political, you must flag it; targeting gets restricted (no age/gender/zip narrowing).

Campaign structure

Campaign → chooses the OBJECTIVE + (Advantage) budget └─ Ad Set → audience, placements, optimization event, schedule └─ Ad → the creative (video / image / carousel) + copy + CTA

Keep it simple. Fewer, better-funded ad sets beat many tiny ones. A common local-service structure: 1 campaign, 1–2 ad sets (broad + one tighter), 3–5 creatives each.

Objectives — pick the outcome, not the vanity metric

Objective	Use for
Leads	Instant Forms (Lead Ads) or website leads — the workhorse for local service.
Sales	E-commerce / purchases with the pixel firing Purchase.
Traffic / Engagement	Rarely for lead-gen — only for warming audiences or content tests. Don't optimize a lead client to Traffic.
Awareness	New location launches, brand lift. Cheap reach, weak on direct response.

Lead Ads vs. landing page: Instant Forms convert cheaply but leads are lower-intent (and need instant follow-up). Website leads to a good landing page are higher-intent but cost more. Test both; add qualifying questions to Instant Forms to raise lead quality.

Audiences

Advantage+ Audience / broad — in 2026 the algorithm targets best when you go broad and let creative + the conversion signal do the work. Start broad (location + age range only) for most local service.
Custom Audiences — retargeting: website visitors (pixel), customer lists, video viewers, page/IG engagers, lead-form openers.
Lookalikes — 1–3% off a strong seed (purchasers or high-value leads). Still useful, but broad + good signal often wins now.
Always run retargeting — a small always-on ad set to site visitors + engagers is the cheapest conversions you'll get.

Budget & bidding

Advantage Campaign Budget (CBO) — set budget at the campaign; let Meta distribute across ad sets. Default choice.
Bid strategy: start with Highest Volume (lowest cost) until you have data, then move to Cost per Result Goal to control CPL. Use Bid Cap only when you know your numbers.
Learning phase: ~50 optimization events in 7 days per ad set to stabilize. Don't edit significant settings during it — every meaningful edit resets learning.

Creative — the real lever

Hook in the first 1–3 seconds (thumbstop). Most spend is lost on weak openers.
UGC & native-feeling video outperform polished ads for local service. Faces, real results, phone-shot authenticity.
Angles, not just variations — test different problems/offers/audiences, not five colors of the same ad.
Specs: 9:16 (Reels/Stories) and 1:1 / 4:5 (Feed). Design for sound-off (captions), safe zones for Reels UI.
Copy: lead with the offer/benefit, one clear CTA. Advantage+ Creative can auto-generate variations — use as a base, not a crutch.
Refresh cadence: watch frequency; when it climbs (>2–3 in a tight local radius) and CTR drops, creative is fatigued — refresh.

Optimization workflow

Let it run 3–4 days / through learning before judging. Resist daily knee-jerk edits.
Kill ads with high spend + no conversions (e.g., spent ~2× your target CPA, zero leads).
Scale winners: vertical (raise budget ~20% every few days) or horizontal (duplicate the winner into a new audience). Big budget jumps reset learning.
Keep a small retargeting ad set always on.
Check the funnel: high CTR + low conversions = landing page/offer problem, not an ad problem.

KPIs to watch

CPM CTR (link) CPC Cost/Lead CPA ROAS Frequency Hook/thumbstop rate — but the only ones that pay the client are CPL / CPA and lead-to-sale close rate.

Account foundation

Conversion tracking first — Google Ads conversion actions + GA4 + Enhanced Conversions. For calls, track call-from-ad and calls to a number on the site. No tracking = Smart Bidding is blind.
Google Tag / GTM installed sitewide. Import GA4 key events as conversions.
Google Business Profile linked (for LSA, Local, and location assets).
Access via Manager Account (MCC) so you manage clients without owning their logins.

Campaign structure

Account → Campaign (type + budget + bid strategy) └─ Ad Group (a tight theme of keywords) ├─ Keywords (+ match types) & Negative keywords └─ Responsive Search Ads (headlines/descriptions) + Assets

Campaign types — pick the right one

Type	Use for	Notes
Local Services Ads (LSA)	Home services, legal, health, beauty — pay-per-lead, "Google Guaranteed" badge	Often the best first move for local service. Top of page, pay per lead not click, badge builds trust. Separate from regular Google Ads.
Search	Capturing high-intent queries	The core of intent marketing. Full control of keywords + negatives.
Performance Max	Scaling across all Google inventory once conversion data exists	Powerful but a black box — can cannibalize brand/Search and burn budget on junk placements. Use with strong asset groups + audience signals + brand exclusions.
Demand Gen	Visual, social-style demand creation (YouTube, Discover, Gmail)	Google's answer to Meta. Good for the awareness layer.
Display / Video	Retargeting, awareness	Cheap reach; weak for direct lead-gen alone. Great for remarketing.

Local-service default stack: LSA (pay-per-lead) + a tight branded/high-intent Search campaign + Display/YouTube retargeting. Add PMax only after Search is proven and you can exclude brand.

Search deep-dive

Match types

Exact [emergency plumber] — tightest, highest intent.
Phrase "emergency plumber" — meaning preserved, some flexibility. Good default.
Broad emergency plumber — widest; only safe with Smart Bidding + strong conversion data + aggressive negatives.

Negative keywords — the discipline that saves budgets

Mine the Search Terms report weekly and add negatives (jobs, "free", DIY, cheap, competitor names you don't want, irrelevant nearby services). This is the #1 way to stop wasted spend on Google.

Quality Score & Ad Rank

Quality Score = expected CTR + ad relevance + landing page experience. Higher QS → lower CPC and better position. Improve it by tightly theming ad groups (keyword → ad → landing page all match).

Responsive Search Ads + Assets

Give 10–15 headlines + 4 descriptions; pin only when legally/brand necessary. Aim for "Good/Excellent" ad strength.
Assets (extensions): Sitelinks, Callouts, Structured snippets, Call, Location, and Lead Form — add all relevant ones; they lift CTR and real estate for free.

Bidding

Start: Maximize Conversions (no target) while you gather ~15–30 conversions.
Then: Maximize Conversions with a Target CPA, or Maximize Conversion Value with Target ROAS for revenue-tracked clients.
Manual CPC only for tiny/new accounts with no conversion data.
Smart Bidding needs clean conversion data — bad tracking makes it optimize toward the wrong thing.

Optimization workflow

Search Terms report → add negatives + promote good terms to their own keyword.
Pause/limit keywords with spend but no conversions; shift budget to converters.
Improve Quality Score on expensive keywords (tighten ad group + landing page).
Watch Impression Share (lost to budget / rank) to know whether to raise budget or bids.
Guard PMax with brand exclusions so it doesn't take credit for branded searches.

KPIs to watch

Impressions CTR CPC Quality Score Conv. rate CPA ROAS Impr. share — pay on CPA + close rate, same as Meta.

New-client launch checklist

Discovery: customer value, close rate, target CPL, service area, best offer.
Tracking: pixel + CAPI (Meta), conversion actions + GA4 + Enhanced Conversions (Google), call tracking. Verify a test lead fires before spending a dollar.
Assets: access via Business Manager / MCC (never shared passwords), Page, GBP, ad accounts funded.
Landing page or Instant Form built, with consent language (TCPA) on lead forms.
Creative: 3–5 angles (Meta) / RSAs + assets (Google).
Structure: 1 campaign, 1–2 ad sets/ad groups, correct objective + bid strategy.
Set a clear target CPA and a kill rule (e.g., pause an ad at 2× target CPA with 0 conversions).

Weekly optimization ritual (per client, ~30 min)

Pull spend, leads, CPL, CPA vs. target.
Meta: check frequency + creative fatigue; kill losers; scale winners ~20%.
Google: Search Terms → negatives; pause dead keywords; check impression share.
Confirm conversions are still tracking (broken tags are silent killers).
Note lead quality feedback from the client — optimize toward booked jobs, not raw leads.

Monthly client reporting

Spend → leads → CPL → booked → revenue → ROAS. Tie it to jobs closed, not clicks.
What changed + what's next month's plan. One clear recommendation.
Deliver via a Dashee dashboard (live) rather than a static PDF — that's the productized edge.

Common mistakes to avoid

Launching without verified conversion tracking.
Editing during the learning phase / judging on day 1.
Optimizing to Traffic/clicks instead of Leads/Sales.
No negative keywords on Google (bleeds budget on junk terms).
Letting PMax cannibalize branded search.
Reporting vanity metrics (impressions, CTR) instead of CPL/close rate.
Weak offer + hoping targeting fixes it. It won't.

Benchmarks & glossary

Benchmarks vary wildly by industry, geography, offer, and season. Use these as rough sanity checks, not targets — the only benchmark that matters is a CPA that's profitable for that client.

Rough local-service reference ranges

Metric	Meta (FB/IG)	Google Search
CTR	~1–2%+ (link)	~4–8%+ (high-intent)
CPC	~$0.50–$2	~$2–$15+ (service-dependent)
Conversion rate (LP)	~5–12%	~5–15%
Cost per lead	~$10–$60	~$20–$150+
Frequency (healthy)	< ~2–3 in tight radius	n/a

Glossary

CPM	Cost per 1,000 impressions
CTR	Click-through rate = clicks ÷ impressions
CPC	Cost per click
CPL / CPA	Cost per lead / cost per acquisition (conversion)
ROAS	Return on ad spend = revenue ÷ spend
Frequency	Avg. times a person saw the ad (fatigue signal)
CAPI	Conversions API — server-side tracking (Meta)
LSA	Local Services Ads — pay-per-lead, Google Guaranteed
PMax	Performance Max — cross-inventory automated campaign (Google)
Quality Score	Google's 1–10 keyword quality (CTR + relevance + LP)
Learning phase	Period while the algorithm gathers data (~50 conv/wk on Meta)