How much is your service?

Our pricing model is based on the performance we achieve, ensuring our goals align with yours. We do not charge retainers.

What level of communication and reporting can I expect?

Weekly Loom walkthroughs with a written summary, plus a dedicated Slack channel with the team available seven days a week. No portal you have to log into to find your numbers.

How do you stay updated on industry trends and best practices?

Through our close relations with Google, we receive ongoing training from high-level representatives on new features often before public release. Our scale across high-spending accounts gives us real-time data on what's actually working.

What sets your agency apart from the competition?

Our specialists are among the top professionals globally. At Ad-Lab we enforce stringent conditions to ensure each team member handles a minimal number of accounts, allowing unparalleled focus on each client.

Do you offer free account audits?

Yes. We conduct in-depth audits and provide strategic recommendations. We identify areas where your previous or current agency is missing optimization opportunities.

How long are your contracts?

Month-to-month. We are confident that within the first month we can demonstrate the Ad-Lab difference, eliminating any need for long-term commitment.

How does the pricing work?

Performance-priced. Month-to-month. No retainer, no annual lock-in, no exit fee.

What's a typical Ad-Lab client size?

$30,000+ in monthly Google + YouTube ad spend, or $100,000+ in monthly revenue. Most are doing both. Below that range the audit doesn't surface enough leverage to justify our model. Above $1M/month is where the system around the ad starts compounding hardest. If you're not sure where you fit, apply for the free audit anyway and we'll tell you straight.

Why aren't you on Meta or TikTok?

Because the work that compounds Google + YouTube spend at scale (landing pages, creative, feed engineering, tracking, AI ops, weekly audits) is full-time work. Spreading thin across Meta and TikTok would mean less depth in the channels where we already have $150M+ of pattern recognition. We partner with Meta and TikTok specialists for clients who need both, but we won't pretend to be the best at every channel ourselves. That said, the Google + YouTube work we ship affects your Meta and TikTok performance every week (branded search volume, AOV across the shop, blended CPA), and we report that cross-channel context as part of the standard cadence.

All case studies

EcommerceIllustrative exampleOne quarter

$520K to $1.06M a month by adding YouTube on top of a clean Search, Shopping, and PMax stack

The harvest layers were maxed. We built a demand layer to feed them.

Monthly revenue

$520K → $1.06M

Blended ROAS

3.0x → 2.8x

Branded search demand

+44%

Monthly spend

$173K → $378K

Chapter 01

The demand ceiling nobody talks about

There are two jobs in a paid account: capture demand that already exists, and create demand that doesn't yet. Search, Shopping, and Performance Max are harvest channels. They're brilliant at scooping up people already looking. But there's only so much existing demand at any moment, and once you own most of it, pushing more budget into harvest just bids up the price of the same clicks. ROAS falls and you blame the platform.

This account had hit that wall. Impression share on the core terms was above 70%, branded search was flat, and every budget increase on Search and PMax dropped the ROAS. It wasn't a budget ceiling. It was a demand ceiling. The fix isn't to harvest harder, it's to plant.

Chapter 02

The full stack: every campaign and what it feeds

This is the entire account architecture at the end of the quarter. Three layers, each with a job, each feeding the next.

We kept the harvest layers exactly as they were, because they were working. On top we added a demand layer (YouTube and Demand Gen) whose job is to put the brand in front of people who weren't looking yet. Demand created there shows up days or weeks later as branded searches, direct visits, and a higher Shopping conversion rate. The harvest layers then capture it cheaply. The two halves are not separate experiments, they're one machine.

The complete account map

Layer	Campaign	Funnel role	Bid strategy	Budget share
Harvest	Brand Search	Capture owned demand	Target impression share	3%
Harvest	Non-brand Search (high intent)	Capture ready buyers	tROAS	14%
Harvest	Non-brand Search (research)	Capture early demand	Max conversions → tROAS	8%
Harvest	Performance Max (brand excluded)	Scale catalogue + remarketing	tROAS	30%
Harvest	Standard Shopping (priority funnel)	Long-tail + query control	Manual / tCPC	7%
Demand	Demand Gen	Mid-funnel discovery	Max conversions / value	13%
Demand	YouTube in-stream (video action)	Create demand, drive action	tCPA / max conversions	18%
Demand	YouTube reach	Top-funnel awareness	Target CPM / efficient reach	7%

Budget shares of the end-of-quarter $378K. Roughly 62% on harvest, 38% on demand creation. Early on, demand was a sliver and grew as it proved out.

How we sequenced it without blowing up ROAS

1
Confirmed the harvest layers were genuinely saturated first
Demand creation is wasted money on an account that still has cheap harvest left. We only added YouTube once impression share, flat branded volume, and falling marginal ROAS all confirmed the ceiling.
2
Started YouTube small with a conversion-led video system
Skippable in-stream built for action, not a brand film. Strong hook in the first three seconds, product and problem shown early, one clear offer, several cuts from the same script so the system could find the winner.
3
Held it through the learning window
Three to four weeks before judging anything. Demand layers look bad on day-three last-click numbers and good on week-four blended numbers. Cutting early is the most common way teams kill a working YouTube campaign.
4
Measured on the right scoreboard
New-customer acquisition, the trend in branded search and direct traffic, and blended ROAS. Not the in-platform view-through number, which over-credits.
5
Scaled demand and watched harvest get cheaper
As YouTube and Demand Gen spent, branded search climbed 44% and the Shopping conversion rate rose. We fed that back into harvest budgets. The machine compounded.

Chapter 03

What happened to the numbers

The mindset shift that made it work

Last-click thinking

Every channel judged on its own last-click ROAS
YouTube cut the moment it showed a 1.x in-platform number
Budget piled into Search until CPCs ran away
Revenue stuck at the ceiling of existing demand

System thinking

Harvest measured on ROAS, demand measured on new customers and brand lift
YouTube held through the learning window and scaled on blended results
Spend balanced across creating demand and capturing it
Branded search up 44%, the proof the demand layer worked

Start to end of quarter

Metric	Start	End
Monthly revenue	$520K	$1.06M
Monthly spend	$173K	$378K
Blended ROAS	3.0x	2.8x
Branded search volume	Baseline	+44%
Channels live	Search, Shopping, PMax	+ YouTube, Demand Gen

The 0.2x of blended ROAS we gave up bought a doubling of the business. Demand you create is demand competitors can't outbid you for.

We'd been stuck around half a million for months and assumed we'd hit the lid. We hadn't hit a budget lid, we'd hit a demand lid. The day YouTube started spending, branded search started climbing.

Founder

Questions we get asked about this

When should I add YouTube?

When your harvest layers are saturated, not before. The tells: search impression share above roughly 70% on your core terms, branded search volume flat, and ROAS dropping every time you raise Search or PMax budgets. That means you've bought the existing demand. YouTube and Demand Gen create new demand instead of competing for the same clicks. Add it early and you're paying for awareness an account with cheap harvest left doesn't need yet.

Won't YouTube tank my ROAS?

On a last-click view, yes, and that's the wrong view. YouTube's job is to create demand that shows up later as branded searches, direct visits, and a higher Shopping conversion rate. Judge it on blended ROAS, new-customer rate, and whether branded search climbs. Ours dipped from 3.0x to 2.8x blended while revenue doubled. That trade is a win every time.

How do I measure YouTube if it's mostly view-through?

Three signals together: Google's new-customer acquisition reporting, the trend in branded search and direct traffic in GA4, and a geo or time-based holdout if you want to be strict. Don't lean on the in-platform view-through number alone, it over-credits. Watch whether the whole account grows when YouTube spends.

What video should I actually run?

A conversion-led structure, not a brand film. Strong hook in the first three seconds, the product and the problem it solves shown early, one clear offer, and several cuts from the same core script so the system can find the winner. Run skippable in-stream for action and keep awareness formats to a small slice until the action campaign proves out.

Frameworks behind this case study

Track: YouTube + Demand Gen as a demand machine

Glossary: YouTube Ads

Track: Performance Max done right

About these numbers

Illustrative example. The campaign structure shown is the exact build we deploy. The figures are representative of what it produces for accounts in this spend band over a quarter, not an export from one client's account.

More breakdowns on the daily newsletter

Jun 14, 2026

Want this run on your account?

Free 15-minute audit. No deck. No pitch. We screenshare your account, run the same audit playbook these case studies came out of, and walk away leaving you the doc.

Book the call