Google Ads Account Structure for B2B with Long Sales Cycles
When months pass between the first click and a signed contract, standard account structure stops working: there aren't enough conversions for the algorithm, and an averaged CPA tells you nothing. Here's how I build accounts for international B2B — using a software house selling into DACH, the Nordics and the USA as the working example.
01 — Problem
Where Standard Structure Breaks on a Long Cycle
In e-commerce, minutes pass between click and purchase. In B2B with a complex service it's weeks or months, several people in the decision, and dozens of touchpoints. Structure that works beautifully in the first case falls apart in the second in four predictable ways.
One campaign for every market
DACH, the Nordics and the USA have different click prices, different competition and a different language of value. In a shared campaign the expensive region eats the cheaper one's budget, and all you see is an averaged number that tells you nothing.
All services in one ad group
A B2B buyer searches for a specific solution, not "software development". When three verticals are blended into one ad group, the ad can't be relevant to any of them.
Optimizing only for the final form fill
If a deal closes in 3–6 months and you get 5–10 conversions a month, the algorithm simply has nothing to learn from. The campaign sits in learning mode for months.
Remarketing as an afterthought
In a long cycle, bringing people back is the core mechanism, not a bonus. When remarketing lives inside a general campaign, it can neither be measured nor scaled on its own.
02 — Structure
The Four Layers of a B2B Account
The core principle: each layer owns its stage of the cycle and is judged on its own metric. A single account-wide CPA in B2B is a number that hides failures and successes equally well.
Segmentation: region × vertical
The first cut is not campaign type — it's the buyer. On the Acropolium account, structure was built by region (DACH, Nordics, USA, Oceania) and by vertical (Subscription, Hospitality, Logistics). Every "region + vertical" pair gets its own budget and its own message.
Search — demand that already exists
High-intent queries with carefully built keyword sets. This is the layer that delivers volume: on Acropolium, Google Search produced the bulk of leads at a $0.21 CPC — remarkably low for B2B IT.
PMax and Display — demand not yet formed
Separate campaigns, separate budget, separate success metric. Their job isn't to deliver form fills directly but to fill the audience for the next layer. Judging them on last-click CPA is a guaranteed route to the false conclusion that "Display doesn't work".
Remarketing — returning across the cycle
The same principle applied at Holiday Bar: Search, Display and Remarketing split into separate campaigns, each with its own metric. In B2B with a months-long cycle, this layer often has the best economics — the person already knows who you are.
03 — Signals
Micro-Conversions: Teaching the Algorithm When Leads Are Scarce
This is the piece without which everything above stays a pretty diagram with no fuel. On the Acropolium account, micro-conversion tracking was built on top of hard conversions precisely to read intent earlier — long before someone submits a form.
Scroll depth
Shows whether a visitor actually reads the service page — an early signal of real interest.
"Contact Us" / "Get in Touch" clicks
Intent to reach out, captured before the form is actually submitted.
Chat opens
Often precedes an enquiry by weeks — and happens several times more often than the enquiry itself.
Key page views
Case studies, pricing, process — the set of pages that separates a buyer from a casual visitor.
The logic is simple: a form fill happens once in hundreds of visits, while a chat open or a fully-read page happens many times more often. That's the difference between an algorithm that learns in a week and one that never learns at all.
04 — Proof
What This Looks Like in Practice: Acropolium
Acropolium is a B2B software-development house selling complex services to international buyers across DACH, the Nordics, the USA and Oceania. Narrow audiences, a long buying cycle and three channels that each behave differently — exactly the case where "one campaign for everything" cannot work.
What matters here isn't the total — it's the division of roles. Google Search delivered volume at a $0.21 CPC, remarkably low for B2B IT. LinkedIn delivered quality through precise seniority and industry targeting, where the high click price is justified. Meta held the middle of the funnel. Judging all three on a single metric would have meant switching two of them off as "ineffective".
Read the full Acropolium case →05 — Mistakes
Four Mistakes That Kill a B2B Account
✕ Waiting until "the data builds up"
At 5–10 conversions a month it never will. That's exactly why micro-conversions aren't "extra analytics" — they're the condition under which automated bidding starts working at all.
✕ Judging Display by direct enquiries
The demand-generation layer is measured by how it fills remarketing audiences and by assisted conversions, not by last-click CPA.
✕ The same ad for every region
What works in the USA often fails in DACH — different expectations around formality, guarantees, and how a service is described in the first place.
✕ Skipping the landing page test
At Acropolium, testing two landing page versions lifted on-site conversion — and the winning message was then scaled across all channels.
06 — Checklist
Structure Checklist — Save This
07 — FAQ
Frequently Asked Questions
How many campaigns does a long-cycle B2B account need?
What if there aren't enough conversions for automated bidding?
Can I merge regions to gather data faster?
Google Ads or LinkedIn Ads for B2B?
How long before this structure shows results?
Does Performance Max work for B2B?
Free Audit
Will Your Structure Survive a Long Cycle?
I'll look at your account and tell you whether its structure is built for a months-long B2B cycle — free, no obligation.