02 · Paid Advertising

Spend that answers for itself.

Paid media is the fastest way to learn what your market wants, and the fastest way to burn a budget finding out. The difference between those two outcomes is almost never the bidding strategy.

We treat every account as a measurement problem first. When the tracking is wrong, the optimization is confidently wrong — the algorithm will spend efficiently toward a signal that does not correspond to revenue, and the reports will look fine while it happens.

What usually goes wrong

Most accounts are optimizing toward a lie.

Conversions counted more than once

Duplicate tags, form loads firing as submissions, and every platform claiming the same sale. Add up the platform reports and they exceed the revenue in your accounting system — often by a lot.

Brand traffic sold back to you

Campaigns capture people already searching your name and report a spectacular return. That demand existed before the spend. Stripping it out is usually the single most clarifying change in an account.

Signal loss nobody accounted for

Browser restrictions and consent choices mean a meaningful share of conversions never reach the platform. Bidding runs on partial data, and the gap is invisible unless someone goes looking for it.

Creative refreshed on a calendar

New assets shipped monthly because the contract says so, with no record of which hook or format actually moved anything. Nothing accumulates, so every quarter starts over.

What we do

Measurement first, then scale.

Conversion tracking

This comes before any campaign work. If we cannot trust the numbers, nothing downstream is worth doing — and this is where most inherited accounts turn out to be broken.

GA4 and Google Tag Manager — events and conversions defined to match how your business actually books revenue.
Server-side tracking — where client-side signal loss is measurably costing you bidding accuracy.
Deduplication and validation — platform numbers reconciled against your own system, with the gap explained.

Google Ads

Structured around margin rather than around whatever the platform defaults to, with branded and non-branded demand kept separate so you can see what the spend is genuinely creating.

Search — intent-led structure, negative keyword hygiene, and query mining as an ongoing habit.
Performance Max — used deliberately, with exclusions and asset group structure that stop it claiming credit for demand you already had.
Shopping — campaign structure driven by product margin and velocity, not by category.
YouTube — where the product benefits from demonstration rather than description.

Meta Ads

Prospecting and retention treated as different jobs, with creative built for each placement instead of one asset resized five ways and hoped for.

Audience strategy — broad targeting where the algorithm performs, tight where it does not.
Creative-led testing — on Meta, creative is the targeting; that is where the testing budget goes.
Conversions API — implemented properly so the signal survives browser restrictions.

Feed management

For e-commerce, the product feed is the campaign. Cleaned and enriched data does more for Shopping performance than most bid adjustments ever will.

Data quality — titles, attributes, and categories written for how people search, not how your PIM exports.
Enrichment — the fields that decide whether you appear for a specific query at all.
Monitoring — disapprovals and stock issues caught before they quietly drain a campaign.

Creative testing

A running queue of hooks, angles, and formats, each with a stated expectation before it launches — so a result teaches you something instead of just replacing the previous asset.

Structured hypotheses — what we expect to move and why, written down in advance.
Retired on evidence — losers cut on data, winners iterated rather than left to fatigue.
A documented library — so the learning survives staff changes on both sides.

What we measureCost per qualified acquisition and contribution after ad spend, reconciled against your own revenue data. Impressions, clicks, and platform-reported ROAS are diagnostics we use to find problems — they are not what we ask you to judge us on.

First 30 days

What happens before we spend.

Week 1

Account audit

Read-only access to your ad accounts and analytics. We reconcile platform-reported conversions against your actual revenue and quantify where the two disagree.

Week 2

Fix the signal

Tracking rebuilt where it is wrong, branded demand separated out, and targets set against real margins rather than against last quarter's reported ROAS.

Week 3–4

Restructure and launch

Campaigns rebuilt around intent and margin, first creative tests live, and a weekly optimization rhythm with a standing call in your time zone.

FAQ

Paid media, answered.

Is there a minimum ad budget?+

No fixed floor, but below roughly a few thousand US dollars a month in media spend, paid search rarely gathers enough data to optimize meaningfully, and your money usually does more in conversion work and content. If that is your situation we will say so rather than take the retainer.

Who owns the ad accounts?+

You do, always. We work inside your Google Ads and Meta Business Manager as added users. We never ask for ownership transfer, and if we part ways your accounts, spend history, and algorithmic learning stay with you — that history is an asset and it is yours.

Do you charge a percentage of ad spend?+

No. Percentage-of-spend pricing pays an agency more for spending more of your money, which is precisely the wrong incentive. We charge a fixed monthly retainer scoped to the work, so recommending that you cut spend on a channel costs us nothing.

How do you handle Performance Max?+

Carefully. It hides placement detail and will happily absorb budget that your brand search would have captured for free. We separate branded traffic, use asset group structure and exclusions to keep it honest, and judge it on incremental results rather than on the ROAS it reports about itself.

How quickly can campaigns go live?+

Usually two to three weeks, and most of that is fixing conversion tracking before launch. Going live against unreliable data means paying the platform to optimize toward the wrong signal, which costs more than the wait does.