09 / Paid

Paid ads management for local service businessesFive surfaces.
One number.
One team.

Five platform agencies will quote you five different numbers. We run Google, Meta, LinkedIn, TikTok, and YouTube as one motion, against one cost per booked job, with new creative every week per surface, because the ceiling on every paid motion past month two is creative, not budget.

/01
5+
platforms run as one motion by one team
/02
Weekly
new creative per platform, made in house
/03
One
cost per booked job, measured across paid together
/04
1 team
audit, build, buy, measure, all under one roof

Most agencies pick a platform.
We run the system.
That difference is the entire pitch.

Most paid agencies optimise for their own platform’s last-click number. You have five of them, and the totals never add up. The only paid motion that compounds is the one paced to a single cost per booked job, refreshed weekly, and measured by holdout. We just decided to run it that way.

The 24-second showreel

Paid, run as a system.

Five glass screens above a war-room desk, native creatives playing on each, and a precision cluster reading spend, cost per booked job, and what those jobs were worth.

showreel.paidSHOWREEL
The five surfaces

Five surfaces. Run natively. Run together.

Each surface needs its own creative grammar, its own bidding posture, its own audience cut. They also need to add up to one number on Friday. That balance is the whole job.

The paid war room: five platform benches running as one motion.
the.warroom
  • Surface 01

    Google

    Search, Performance Max, Demand Gen, YouTube, Display. Tight-keyword Search for the person typing your trade and their postcode at 2am; PMax with proper asset feeds and exclusions; Demand Gen and YouTube for the people who will need you next season. All built on one conversion setup, so the numbers add up.

  • Surface 02

    Meta

    Facebook and Instagram feed, Reels, Stories, Advantage+. Creative built for the platform: vertical, sound on, the owner on camera, real crews on real jobs, before and afters, with new versions every week. Conversions API deduped against the pixel, so one enquiry is counted once.

  • Surface 03

    LinkedIn

    Sponsored Content, Doc Ads, Conversation Ads, Thought Leader. The expensive surface, and where the commercial work lives: property managers, facilities leads, general contractors, insurance adjusters. Targeted to the companies and job titles that hire your trade in your area, with creative that reads like a post.

  • Surface 04

    TikTok

    In-feed, Spark Ads, Smart+. Ranks creative first, so we ship a batch of platform-native versions each round, lean on your own crews and customers, and let the platform find the winners. Spark amplifies the posts that already earned attention on their own.

  • Surface 05

    YouTube + Programmatic

    YouTube In-stream, Shorts, Bumper, Demand Gen video. Programmatic across DV360, Reddit, X, Pinterest, Criteo, StackAdapt for surfaces the others can't reach. Used to reach the people who will need you before they start searching, measured by what actually changed rather than last click.

The operating loop

Six stages. One compounding motion.

Most paid engagements stall around month three because the motion has no loop. Audit, rebuild, ship creative, measure, scale, iterate, the same six stages every week, never skipped, never out of order.

Six-stage paid operating loop visualised as a precision flight-deck.
operating.loop● 6 stages
  • /01

    Audit

    Account archaeology across every platform, where money leaks, what compounds.

  • /02

    Rebuild

    Account architecture, naming, bidding, audience, exclusions, conversion model.

  • /03

    Creative

    Platform-native creative made in house. New work every week per platform.

  • /04

    Measurement

    One cost per booked job, holdout tests, asking every caller how they found you.

  • /05

    Scale

    Once the cost per booked job holds steady, we scale with guardrails.

  • /06

    Iterate

    Every channel ships a named experiment weekly with a pre-registered metric.

The attribution bench: precision instruments measuring paid causation, not last-click theatre.attribution · 5 truths
How we measure

Most attribution is theatre.

Multi-touch models are coin flips in spreadsheets. We operate with five honest disciplines instead: holdout tests, asking callers how they found you, one cost per booked job, weekly experiments, monthly review. Slower than a dashboard. More accurate than one.

honest · per channel · per quarter
  1. /01

    Incrementality over attribution

    Geo-blackouts and percent-holdouts, at least quarterly, on every channel we spend on. The only honest measurement of paid is the version that asks 'what would have happened anyway?'

  2. /02

    Survey-on-conversion

    'How did you hear about us?' as a free-text field on every quote form and booking, and a question your office asks on the phone. Coded monthly. Ugly but honest, and the only thing that catches the job that started with a Facebook video and ended with someone typing your name into Google.

  3. /03

    One cost per booked job

    Each click does not turn into a job on its own, and we don't pretend otherwise. We pace budget to one cost per booked job across all paid sources and let the platforms compete inside it. The math gets clean very quickly.

  4. /04

    Weekly experiments

    Every channel ships one named experiment per week, with a pre-registered outcome metric and a pre-registered kill rule. Anything that isn't an experiment is just spend.

  5. /05

    Monthly review

    One 45-minute call. Three numbers: what you spent, what it cost per booked job, and what those jobs were worth. What we'd change for next month, in writing, by Friday.

How a week shapes up

Five days. One paid motion in motion.

  1. MON

    Audit + plan

    Past-week scorecard pulled, winners flagged, kill list signed off, this-week experiment briefed.

  2. TUE

    Creative ship

    New versions approved, trafficking briefed, naming conventions held, campaign names a person can still read in six months.

  3. WED

    Build + QA

    Campaigns built, exclusions audited, conversion model checked, UTMs validated end-to-end before any budget moves.

  4. THU

    Launch + watch

    Live-fire launches in batched windows. Hourly watch on day-one. Pause-rules written and agreed before anything goes live.

  5. FRI

    Score + refresh

    Cost per booked job updated, pacing reviewed, creative refresh queued for the next Tuesday's ship.

What we hold every week

What the war room actually runs on.

Six precision media-buying instruments along a horizon, platforms, refresh, cost per job, team, holdout, experiments.
  • /01
    5+
    platforms operated as one motion by one team
  • /02
    Weekly
    new creative per platform, made in house
  • /03
    One
    cost per booked job, measured across paid together
  • /04
    1 team
    audit, build, buy, measure, all under one roof
  • /05
    Quarterly
    holdout test on every channel we spend on
  • /06
    Per week
    one named experiment per channel, metric agreed up front
Deliverables

Every motion, scoped & shipped.

Fixed scope per engagement. Audit, rebuild, creative studio, measurement model, weekly experiments, daily ops, and a monthly review written in plain English that you can read in one sitting.

  • 01

    Cross-platform paid audit

    Account archaeology across every platform you spend on, where money leaks, what compounds, what to kill.

  • 02

    Account architecture rebuild

    Naming, bidding, audience, exclusions, conversion tracking, the unglamorous plumbing everyone skips.

  • 03

    Platform-native creative, made in house

    Owner on camera, motion design, customer footage, doc ads, built to platform, refreshed weekly.

  • 04

    Measurement model

    One cost per booked job, holdout tests, a 'how did you hear about us' field, honest weekly reporting.

  • 05

    Weekly experiment register

    Every channel ships one named experiment per week, with the outcome metric agreed before it starts.

  • 06

    Live-buy operations + pacing

    Campaign builds, daily pacing, pause-rules, exclusions audit, the buying desk runs every day.

  • 07

    Monthly review + plain-English report

    45-minute review. Three numbers. What changes for next month, in writing, by Friday.

The stack

Five surfaces. One buying desk.

The platform screens are the easy part. The stack that holds it together, conversion tracking, server-side events, creative QA, holdout tooling, and the question on the quote form, is where the gains come from.

  • GoogleSearch · PMax · Demand Gen · YouTube · Display
  • MetaFB · IG feed · Reels · Stories · Advantage+
  • LinkedInSponsored · Doc · Conversation · Thought Leader
  • TikTokIn-feed · Spark · Smart+ · TopView
  • YouTube + ProgrammaticShorts · Bumper · DV360 · Reddit · X · Pinterest
  • MeasurementGA4 · server-side CAPI · GeoLift · in-form survey
Questions we get

What owners ask on the second call.

01

We already have a Google Ads guy, why run paid as one system?

Single-platform agencies optimise for their own platform's number, not yours. Google counts last click, Meta counts view-through, LinkedIn counts something else again. Add up what all three claim and you have booked more jobs than you actually did, while paying twice for the same enquiry. One team pacing one cost per booked job across the surfaces, with proper holdout tests, tells you which channels are carrying their weight.
02

Do you do creative or just buying?

Creative is half the job. Plenty of agencies farm it out to a freelance pool, ship a Canva template, and call it platform-native. We run an in-house creative team: the owner on camera, real crews on real jobs, before and afters, motion design, doc ads, refreshed weekly, because creative is the ceiling on every paid motion past month two. If you already have someone doing creative, we work with them.
03

How long before the cost per job comes down?

The first month is diagnosis and rebuild. You'll see structural fixes early, wasted spend switched off, exclusions, naming, tracking, before you see the cost per job move. It usually moves in the second and third month on accounts that had never been operated this way. After that the gains come from creative discipline and honest holdouts: slower, and they compound.
04

How honest is your reporting?

Brutally. The monthly review is one 45-minute call with three numbers: what you spent, what it cost per booked job, and what those jobs were worth. What worked, what didn't, what changes for next month, in writing by Friday. If a channel quietly stopped working in week three, you hear about it in week three.
05

What's the minimum spend?

There is a floor, and the experiments set it, not us. Below a certain monthly budget there isn't enough volume on any one channel to test anything, and our fee ends up bigger than what the platforms get. If you are under it, one platform run well by one person will serve you better. Send us your budget and we'll tell you straight whether the model fits, and when to come back.
A clean conversion event arriving on the war-room screen, five surfaces, one number.
Ready when you are

Audit the spend.
Rebuild the model.
Run the system.

First: the full audit and the kill list, signed. Then: account architecture rebuilt and conversion tracking live. Then: a weekly experiment cadence shipping creative across every surface. Every month: a three-number review, in writing, by Friday.

Want a paid-account audit across Google, Meta, LinkedIn?

Send read-only access. Written audit plus a prioritised list of what to fix first, back within a week.

By submitting you agree to our privacy policy.
SEO page libraries for service businesses

Want the phone to ring?

A 30-minute call. Tell us the work you want more of and the areas you serve. We’ll show you the searches happening there now, the pages it would take to cover them, and what that costs at $50 a page. You keep that whether or not you hire us.

Metro Vancouver · Greater Toronto · Indianapolis