SaaS Google Ads account structure

Open a poorly performing SaaS search account and you’ll usually find the same thing: thirty campaigns named after keyword themes, brand terms buried inside a campaign called “Core”, and a shared budget that quietly funds the cheapest clicks in the account. The keyword research was fine. The architecture decided the outcome.

Structure in 2026 is a budget control problem, not a relevance problem. Google’s matching and bidding systems handle relevance now. What they do not handle is your need to spend exactly 4,000 dollars on competitor terms and not a dollar more, and that need is what campaign boundaries exist to serve.

Why should intent tier define your campaigns?

Because budget lives at the campaign level, and intent is the dimension you most need to control budget along. Brand clicks cost two dollars and convert at 15 percent. Category clicks cost 30 dollars and convert at 3 percent. Put them in the same campaign and the auction will happily spend your whole day’s budget on the cheap ones.

Five tiers cover almost every SaaS account.

Tier	Match types	Bid strategy	Landing page	Success metric
Brand	Exact plus phrase on your name	Manual CPC or maximise clicks with CPC cap	Homepage or a brand specific page	Impression share above 90 percent at a capped budget
Competitor	Exact and phrase on rival names	Manual CPC, reviewed weekly	Comparison page, never the homepage	Cost per qualified lead, not CPL
Category	Exact plus phrase on 'x software', 'best x tool'	Target CPA once above 30 conversions per month	Category or product page with demo form	Qualified leads and pipeline
Problem	Phrase and broad on pain point queries	Target CPA with a looser target	Solution page or a gated resource	Cost per lead and downstream MQL rate
Retargeting	Audience based, no keywords	Target CPA or maximise conversions	Case study, pricing or a direct demo page	Assisted conversions and view through discipline

The tier tells you the bid, the copy, the page and the metric. That’s the whole point: a structure where knowing which campaign a click came from tells you everything you need to decide what to do about it.

Why are single keyword ad groups now actively harmful?

Because they fragment your conversion data below the volume automated bidding needs. Google’s documentation puts the working threshold at roughly 30 conversions per campaign per 30 days. A typical B2B SaaS account produces 40 to 80 demo requests a month in total.

Split those across 25 SKAGs and every ad group carries one or two conversions. The algorithm has nothing to learn from. It falls back to broader account level signals, which is exactly the outcome SKAGs were invented to avoid.

SKAGs made sense in 2016. Match types were literal, you bid manually per keyword, and isolating a keyword gave you real control over its bid and its ad. Close variant matching ended that. Today an exact match keyword already pulls in dozens of query variations, so your “single keyword” ad group is serving a cluster whether you like it or not.

The replacement is intent themed ad groups: four to eight keywords that share a query intent and can honestly use the same ad and the same landing page. Ten ad groups per campaign is plenty. If you cannot write one ad that serves every keyword in an ad group, the ad group is wrong, not too big.

When do you split by geography or language?

When pricing, currency, support hours or the landing page actually differ, or when a market has enough volume to carry its own budget. Not before.

Three practical cases where the split earns itself. You price in euros with local VAT handling and the landing page is different. You run EMEA support hours and want ad schedules that match. Or Germany is 30 percent of your pipeline and you need to protect its budget from a US spending spike.

The case against splitting is always the same: conversion volume. Three campaigns at eight conversions each perform worse than one at 24. If you need geo visibility without geo control, use location targeting on one campaign and read the locations report, which gives you the numbers without costing you the signal.

Language targeting is separate and more often mishandled. Targeting German language ads at Germany excludes the substantial share of German professionals browsing with English browser settings. If your product’s interface is English, target English language in non English markets too, and let the ad copy do the qualifying.

What naming convention makes CRM reporting possible?

One that encodes tier, geography, match type and goal in a machine readable order. Pick it before launch, because renaming campaigns later breaks every historical join in your warehouse.

A convention that survives contact with Salesforce and HubSpot:

Search_US_Category_Exact_Demo Search_DACH_Competitor_Phrase_Demo PMax_US_Signup Retarget_Global_PricingVisitors_Demo

Five segments, underscore delimited, no spaces, no free text. Your ValueTrack parameters then carry campaign name into the CRM, and a single report tells you which intent tier produced the closed won deals. Without this, someone rebuilds attribution by hand every quarter.

Labels handle the dimensions the name cannot. Use them for test cohorts, quarter launched, and owner. Labels can change without breaking reporting joins, campaign names cannot.

How do portfolio bid strategies rescue thin campaigns?

They pool conversion data across campaigns that individually have too little. If you’ve got four campaigns each producing eight conversions a month, a shared portfolio target CPA strategy gives the algorithm 32 to work with while you keep separate budgets.

This is the single most useful setting for mid sized SaaS accounts, and it’s underused because it lives in the shared library rather than the campaign screen. Use it when campaigns share a conversion goal and a roughly similar acceptable CPA. Do not use it to pool brand with category, because their CPAs differ by an order of magnitude and the portfolio will simply chase brand.

Two constraints worth knowing. A portfolio strategy applies one target across all members, so campaigns with genuinely different economics need their own. And the learning period resets when you add or remove a campaign, so batch your changes rather than adding one campaign a week.

What does this look like at $15K a month?

Here’s a worked skeleton for a 2 million dollar ARR B2B product selling a 12,000 dollar ACV tool, running 15,000 dollars a month with a 25,000 dollar target cost per closed won deal.

Six campaigns. Brand capped and left alone. The two category campaigns share a portfolio strategy so they clear the 30 conversion threshold together rather than starving separately. Competitor stays manual because its conversion rate runs a third of category and automated bidding will chase the wrong clicks. Problem tier runs to a content offer with its own lower target, because demo requests from problem queries convert badly and pretending otherwise wrecks the category target.

Notice what’s absent. No SKAGs. No campaign per product feature. No broad match experiment until the account has a reliable conversion signal to feed it. The benchmarks will tell you whether your CPCs are in a normal band before you conclude the structure is the problem.

What does this structure cost you?

Control has a price and it’s worth stating honestly. Five to eight campaigns means five to eight conversion pools, and in a small account that fragmentation is real. If you’re spending 3,000 dollars a month and generating 12 conversions, three campaigns is the correct answer, not six, and you should run manual or maximise clicks until volume arrives.

The second cost is maintenance. Every campaign is a set of negatives, ad copy, extensions and a budget someone has to review. Accounts sprawl because adding a campaign is easy and deleting one feels risky. Run a quarterly consolidation pass and kill anything under 300 dollars a month that isn’t a deliberate test.

The third is that intent tiers blur. A query like “best project management software for agencies” is category and problem at once, and you’ll spend real time deciding where things live. Pick a rule, write it in the account notes, and stop relitigating it. Most of the common PPC mistakes that waste real money start as a structural decision nobody documented.

What to do next

Export your current campaign list and tag each one with an intent tier. If a campaign contains two tiers, that’s your first split. If two campaigns share a tier, a geography and a goal, that’s your first merge.

Then fix naming before you touch anything else, verify one lead’s GCLID reaches your CRM, and set the brand budget cap. Once the skeleton is right, the ad copy work matters, and our ad copy swipe file plus the negative keyword list will save a week. For the wider channel picture, start at SaaS PPC, work through the Google Ads guide, look at real accounts in the teardowns, and pick your stack from the PPC tools guide.