PPC For SaaS: The Campaign Structure And Metrics That Work

Most SaaS accounts I audit are built like an ecommerce account and measured like a local lead-gen one. That combination is why they plateau at the same number every quarter no matter how much budget goes in. The fix is a structure split by buyer intent rather than by product feature, a conversion signal that predicts revenue rather than one that counts form fills, and a scorecard built on CAC payback instead of cost per lead.

I run paid search for software businesses where a single closed deal can be worth more than a month of media spend, and the accounts that work all share those three traits. Below is the structure, the conversion hierarchy to bid against at each stage of your data maturity, how to push closed revenue back into Google, and the numbers that tell you whether any of it is paying.

Structure Search by intent, not by feature

The most common SaaS structure I inherit is one campaign per product module. It feels tidy on a whiteboard and it wrecks your bidding, because the person searching your brand name and the person searching a category term have nothing in common except that they both ended up in your account.

Split by intent instead. Four buckets cover almost every SaaS account.

Campaign	What sits in it	What to expect
Brand	Your name, misspellings, name plus “pricing”, “login”, “review”	Cheapest clicks, highest conversion rate, defends against competitor bids
Category	The job the software does: “helpdesk software”, “invoice automation tool”	Your main growth engine and your most expensive clicks
Competitor	Rival names, “vs” and “alternative to” queries	Low volume, low conversion rate, high strategic value
Problem-aware	The pain before the category is known: “how to stop double-booking staff”	Cheap traffic, long lag to revenue, needs its own success measure

Keep brand in its own campaign and its own budget. Never let it share a budget with category terms, because brand will always look better on a cost-per-conversion chart and Smart Bidding will happily drain your growth budget into traffic you were going to get anyway.

The other structural trap is over-segmentation. Single-keyword ad groups are still doing the rounds in SaaS because sales teams like the reporting granularity, but they starve the bidding algorithm. Google’s own guidance is that Target ROAS typically needs at least 15 conversions in a 30 day period at the conversion tracking level to be eligible, and that hitting a Target CPA consistently is more realistic with around 30 conversions a month per ad group. Split a 40-conversion account into twelve ad groups and none of them ever learn anything. Consolidate until each ad group clears a volume bar, then split.

Bid to the signal that predicts revenue

Here is where most SaaS paid search quietly fails. The account is optimising to “trial started”, the trial button is easy to press, and Smart Bidding does exactly what you asked: it finds you the cheapest possible people to press it. Six weeks later the cost per trial is down 30% and the sales team is complaining that nothing converts.

Pick the deepest signal in your funnel that still clears the volume bar.

Signal	Bid to it when	The risk
Demo request / trial start	You are under roughly 30 conversions a month	Optimises for form-fillers, not buyers
Qualified lead (SQL)	You can mark qualification within a week or two	Sales admin becomes a media-buying dependency
Opportunity created	You get 15 or more a month and CRM data is clean	Longer feedback loop, slower learning
Closed won	Rare in SaaS, only at real volume	Almost always too thin and too delayed to bid on

If you close four deals a month, you cannot bid on closed won. That is not a failure of ambition, it is arithmetic. Bid to the closest upstream signal that has volume, and carry the revenue information in the conversion value rather than the conversion count. Score a lead at the point it is created using what you already know (company size, domain type, plan selected) and feed that score in as the value.

Google’s value-based bidding guidance is explicit that you need to report two or more different values, and those can be real economic values like revenue or proxy values like a lead score. It also recommends uploading values for three weeks or one to two conversion cycles, whichever is longer, before you actually activate value-based bidding. Do not flip the switch the same afternoon you start sending values.

Feed real outcomes back with enhanced conversions for leads

A SaaS funnel that lives in a CRM and never talks back to Google Ads is bidding blind. Enhanced conversions for leads is the upgraded version of offline conversion import: it uses first-party data such as email addresses to supplement the imported offline conversion data, so the match rate and the reporting accuracy both improve. Google has published a median 10% increase in conversions for advertisers who sent first-party data alongside GCLIDs, compared with standard offline conversion import.

Two platform changes from this year matter if you set this up a while ago and have not looked since:

• Since 15 June 2026, offline conversion import and enhanced conversions for leads uploads have migrated to the Data Manager API and are blocked in the Google Ads API. If your CRM integration or a bespoke script still pushes through the old route, your offline conversions have stopped arriving. Check the conversion action’s “last imported” date today.
• Since April 2026, enhanced conversions for web and for leads are a single on/off setting rather than two separate implementation choices.

The sequence I run is simple. Capture the GCLID on the form and write it to the CRM record. Send the email address as user-provided data. When the deal stage changes, upload the outcome against that GCLID with its value. Google prefers conversion delays under seven days, which is another reason to send a scored lead event quickly rather than waiting ninety days for a closed-won that may never land.

Stop paying to reach people you already serve

Every SaaS account I open has budget going to existing customers searching for the login page or a support article. Build a Customer Match list from your CRM and exclude current customers from prospecting campaigns, then build a second list of churned accounts and free users to target deliberately with different messaging.

Two mechanics to plan around. A list needs at least 100 members added or updated within the last 540 days to stay eligible, and memberships age out at 540 days, so the upload has to run on a schedule. And while exclusions are open to all policy-compliant advertisers, using these lists for targeting, observation and manual bid adjustments requires 90 days of Google Ads history plus more than USD $50,000 of lifetime spend. Start with the exclusions. Everyone can run them and that is where the immediate saving sits.

The metrics that actually judge the account

Cost per lead is a diagnostic. It is not a target, and a board deck built on it will get you defunded the first quarter that sales pushes back on quality. These are the numbers I report on.

Metric	What it tells you	How often
Cost per qualified lead	Whether the traffic is the right traffic	Weekly
Pipeline value created per £ spent	Whether paid is feeding the number sales care about	Monthly
CAC by campaign	Which buckets deserve more budget	Monthly
CAC payback in months	Whether growth is fundable or just expensive	Quarterly
LTV to CAC ratio	Whether the unit economics hold at scale	Quarterly
Win rate by campaign	The one that kills or scales a campaign	Quarterly

Win rate by campaign is the metric almost nobody builds and the one that changes decisions fastest. Competitor campaigns usually show a terrible cost per lead and a strong win rate, because someone comparing you to a rival is already in market. Problem-aware campaigns show the reverse. Judge each bucket on the number that suits its job, not on one account-wide average. If the unit economics side of this is new to you, mastering unit economics in paid ads sets out the maths, and why throwing money at paid ads won’t scale your B2B SaaS covers what happens when you skip it.

What to do in the first 90 days

1. Weeks 1 and 2. Split brand out into its own campaign and budget. Build the CRM-sourced customer exclusion list. Fix conversion tracking so the GCLID reaches the CRM on every form.
2. Weeks 3 to 6. Define lead scoring, then start uploading values through Data Manager without changing the bid strategy. Let the data accumulate.
3. Weeks 7 to 12. Move campaigns with enough volume onto value-based bidding. Build the win-rate-by-campaign report. Review search terms weekly and cut the job-seeker and existing-customer queries that always turn up.

Then settle into a rhythm: search terms and cost per qualified lead weekly, pipeline and CAC monthly, payback and win rate quarterly. Resist the urge to make bid strategy changes faster than your sales cycle can report back on them.

Get a second read on your account

If you are running SaaS paid search and the leads arrive but the pipeline does not move, the problem is nearly always in the signal you are bidding to rather than the keywords you picked. That is the first thing we look at in a free Google Ads audit, alongside structure, wasted spend and whether your offline conversions are still landing after this year’s API changes. If you would rather hand the whole thing over, our paid search management work is built around exactly this: tie the spend to pipeline, then scale what closes.

FAQ

How much should a SaaS company spend on PPC? Work back from CAC and payback rather than picking a percentage of revenue. Take your target number of new customers, divide by your realistic lead-to-close rate to get the leads you need, then multiply by your cost per qualified lead. If the resulting CAC pays back inside a period your cash flow can carry, the budget is defensible.

Should I bid on my own brand name as a SaaS business? Yes, in a separate campaign with its own budget. Competitors bid on your name, comparison sites rank for it, and the clicks are cheap. Keep it separated so its strong metrics never flatter your growth campaigns or absorb their budget.

What conversion should I optimise Google Ads towards for a long SaaS sales cycle? The deepest signal that still produces enough volume for the bid strategy to learn from, which for most SaaS accounts is a qualified lead rather than a closed deal. Carry the revenue information in the conversion value through lead scoring, and import outcomes back with enhanced conversions for leads.

Is Performance Max worth running for B2B SaaS? Only once your conversion signal is trustworthy. Point Performance Max at raw trial starts and it will find the cheapest ones available anywhere. Point it at a scored, CRM-validated signal with a customer exclusion list applied, and it can work as a complement to Search rather than a competitor for the same clicks.