From $30K to $100K monthly Google Ads spend

The three patterns that cap accounts at $30-50K

Most ecom accounts that stall between $30K and $50K monthly spend stall for one of three reasons. The product is fine, the creative is fine, the bidder is fine. The structure is broken in a way that only shows up at scale.

Pattern one: branded search baked into PMax. PMax credits the brand-search side of the auction for conversions that were going to happen anyway. The dashboard shows a flattering ROAS that hides cold-funnel cannibalisation.

Pattern two: asset groups split by product category instead of by customer intent or routine. The audience signal can't seed because the conversions inside one group don't share a buyer pattern.

Pattern three: conversion value is gross revenue, not contribution margin. Smart Bidding compounds whatever you feed it. Feed it gross and it scales the worst-margin SKUs first.

30-minute diagnostic walk

1. 1

   Pull the 30-day search-terms report from PMax

   Filter to terms containing your brand name. Sum the conversion value attributed to those queries. That's the brand-search bleed.

2. 2

   Compare branded ROAS vs blended ROAS

   If the gap between dashboard ROAS and blended (Shopify) ROAS is wider than 25%, attribution is overstating PMax. Brand-search baked-in is the most common cause.

3. 3

   Inspect asset group structure

   Open the campaign and look at the asset groups one level deep. Are they named by product category, by routine, or by margin tier? Category-only naming is a structural smell.

4. 4

   Pull the conversion-value setup

   Open Tools and Settings, then Conversions. Look at the primary action's value. Is it dynamic (transaction-aware) or fixed? Fixed values cap scale.

5. 5

   Score the account

   Three structural signals fixed = ready for scale. Two of three = structurally on the edge. One or zero = the account will hit a wall inside the next 60 days regardless of spend lift.

Diagnostic decision matrix

Run the table top to bottom. Each row that ends in an action becomes part of the rebuild plan in chapter 3.

Reference reading inside Ad-Lab

Why the rebuild order matters

Most agencies fix accounts in the wrong order. They start with bid strategy because it's easy to change. They end with structure because it's politically harder. The result: bid changes that compound the wrong signal because the underlying structure is still broken.

The right order is structure first, signal second, bid last. Branded search comes out of PMax before anything else because every subsequent change is contaminated by the wrong attribution baseline. Asset groups get rebuilt next so the audience signal seeds correctly. Conversion values get re-mapped third so the bidder has the right number to optimise against. Only then does the bid strategy get touched.

Most rebuilds we run finish in 14 days. The brand stays live the entire time. We mirror the existing campaigns, build the new structure in parallel, switch traffic over once the new campaigns have learned, and tear down the old structure last.

Rebuild week 1: branded-search split

1. 1

   Create dedicated Search campaign

   New campaign, exact-match branded keywords, all variants. Bid strategy: target POAS at the brand-search benchmark for your vertical (typically 8-15x ROAS depending on AOV).

2. 2

   Add brand terms as negatives in PMax

   Account-level negative keyword list applied to the existing PMax. This is the moment branded queries stop double-counting.

3. 3

   Watch the dashboard for 7 days

   Dashboard ROAS will drop because the inflated branded-attribution number is gone. Blended (Shopify) ROAS should hold steady. That's the proof the bleed was real.

4. 4

   Adjust the new branded campaign

   Tighten POAS as the bidder learns. Reduce ad-group splits if any are starved. Add ad-extensions: callouts, structured snippets, sitelinks per branded query intent.

Rebuild week 2: asset group split

1. 1

   Map customer intent to asset groups

   For skincare: cleanse / treat / moisturise / sun. For supplements: sleep / focus / energy / recovery. For fashion: drop / evergreen / sale. The pattern is intent-led, not category-led.

2. 2

   Audience signals per asset group

   Each asset group gets its own custom segment. Use search-history segments + your CRM list filtered to that intent's buyer pattern.

3. 3

   Search themes per group

   Three to five tight themes per asset group. Themes are what the model thinks the group is for. Don't catch-all.

4. 4

   Migrate creative

   Move existing assets into the new groups based on what they best serve. Identify gaps. Schedule a 14-day creative shoot to fill the gaps for groups starting empty.

Rebuild verification checklist

• Branded search campaign live, exact-match, POAS-bid
• Brand terms in PMax negative-keyword list
• Asset groups renamed by intent, not by category
• Each asset group has 3-5 search themes
• Each asset group has at least 1 audience signal
• Conversion value in pixel is transaction-aware
• Primary conversion is set to enhanced-conversions on
• tROAS or tCPA target matches account economics
• Old PMax structure paused, not deleted, for 7 days
• Dashboard delta documented before/after

Frameworks that drive this chapter

What new constraints appear past $50K

The structural fixes from chapters 1 and 2 unlock spend up to roughly $50K-$70K monthly on most accounts. Past that, three new constraints surface: creative supply, YouTube readiness, and attribution defence.

Creative supply becomes the bottleneck because the same six creatives that worked at $30K stop converting at $80K. The audience saturates faster at higher spend. The fix is a creative cadence that ships 18+ ad units every two weeks, mixed across YouTube, Demand Gen, and PMax video.

YouTube readiness matters because the cold-funnel demand needed to support $100K of monthly spend doesn't come from Search alone. YouTube + Demand Gen become the demand factory. The accounts that scale into six figures monthly without ROAS collapse are the ones that build the YouTube layer before they need it, not after.

Attribution defence matters because at $100K monthly spend the CFO starts asking for blended numbers, MER targets, and incrementality proof. If you can't answer those questions, the spend gets rolled back. Reporting maturity is the third constraint.

Stage-specific scaling priorities

Build the YouTube layer in parallel

1. 1

   Pick the six scripts

   Two demonstration, two transformation, one founder-story, one customer-proof. Each gets three hook variants for an 18-unit creative batch.

2. 2

   Run YouTube and Demand Gen in parallel on the same scripts

   Same creative, two distribution surfaces. Whichever one delivers lower cost-per-lead at the comparable view-quality bar gets the bigger share next batch.

3. 3

   Wire view-through into the value signal

   Enhanced Conversions for Web + the Conversion API on the page-view event. View-through assists become measurable, then become defendable to the CFO.

4. 4

   Compound on the winner

   Each two-week batch ships 18 new units. Top 4 by efficiency get re-cut into 6 derivative variations next batch. The compounding effect is visible by month three.

$100K-readiness checklist

• Branded search lives outside PMax
• Asset groups split by intent / margin
• Conversion value reflects margin or LTV, not gross
• Six-scripts-three-hooks cadence shipping every 14 days
• YouTube and Demand Gen running in parallel
• Server-side tagging deployed on conversions
• Enhanced Conversions live, EMQ above 8
• POAS, MER, and blended ROAS reported weekly
• One incrementality test scheduled in the next 30 days
• Account has a documented 90-day rebuild log

Deeper reads to layer in