# Google Ads for SaaS

> How to run Google Ads for a subscription product: the campaign types that survive, CPC ranges by category, target CPA derived from ACV, and honest tracking.

Source: https://saas-marketing.net/guides/google-ads-for-saas/
Topic: SaaS PPC and Paid Ads
Type: guide
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/google-ads-for-saas/

## Short answer

Google Ads works for SaaS when Search carries the budget, bids are derived from ACV and win rate rather than from a lead cost target, and closed-won revenue is imported back into the platform so smart bidding optimises to pipeline instead of form fills. Demand Gen and Video are retargeting channels, Display rarely pays for itself, and Performance Max should stay off until brand search is protected by a separate campaign.

## Key takeaways

- Search carries almost all profitable SaaS spend; Demand Gen and Video earn their place only as retargeting.
- Max CPC equals your target CAC multiplied by the click to customer rate, and nothing else decides the ceiling.
- Optimising to raw form fills is the single most common reason a SaaS account loses money at scale.
- Offline conversion import with GCLID is the first change to make, before any ad copy is rewritten.
- Category terms in CRM, security, HR and project management routinely clear $30 to $80 per click.
- Smart bidding needs roughly 30 conversions a month per campaign, which most SaaS accounts never reach without consolidation.

---

Most SaaS accounts that lose money are not badly written. They're badly wired. The bid strategy is optimising toward a number the finance team has never seen, usually a form fill, and every efficiency gain the algorithm finds makes the leads slightly worse.

Fix the signal first. Everything below assumes you have an account open, some spend history, and a sales cycle long enough that the platform cannot see the outcome without your help.

## Which campaign types survive in B2B SaaS

Search carries almost all profitable SaaS spend. Everything else is either a retargeting surface or a way to spend money quickly.

The Performance Max question deserves a straight answer. PMax can work for a self-serve product at $19 to $99 a month with hundreds of monthly signups, because it has enough conversion volume to learn. It goes wrong the same way every time: the campaign quietly serves against brand queries, claims the conversions that brand search would have produced anyway, and posts a CPA that makes it look like the best performer in the account. Run a brand exclusion list and keep a dedicated brand Search campaign before PMax gets a dollar.

Display gets a harsher verdict than most agencies will give it. Across B2B SaaS accounts the placement report is consistently dominated by inventory that does not contain buyers, and the conversions it does report are usually view-through. If you want display inventory, buy it through remarketing lists against people who already reached a pricing page.

## The four intent tiers, and what each one is worth

SaaS search demand splits into four tiers, and the economics of each are different enough that they should never share a budget.

| Tier | Example query | Typical CPC | Typical click to conversion | Intent quality |
| --- | --- | --- | --- | --- |
| Brand | your product name, your product pricing | $1 to $4 | 10% to 25% | Highest, partly cannibalised |
| Competitor | competitor alternatives, competitor vs | $8 to $25 | 2% to 6% | High, low volume, expensive per win |
| Category | project management software, CRM for startups | $20 to $80 | 1% to 4% | Medium, the volume tier |
| Problem or job | how to track employee onboarding compliance | $4 to $15 | 0.5% to 2% | Low, cheapest clicks, longest cycle |

Brand is defensive. Competitors bid on your name, the clicks cost almost nothing, and the conversion rate flatters every dashboard it appears on. Count it separately and test incrementality by pausing it for two weeks in one geography, because a meaningful share of those clicks would have come to you free.

Competitor terms are where most SaaS advertisers get emotionally attached. They convert, they feel strategic, and at $18 a click with a 3 percent conversion rate and a 20 percent close rate you are paying $3,000 per customer before anyone has looked at a demo. That is fine at $25,000 ACV and a disaster at $3,000. Competitor bidding is a price point decision, not a competitive one, and the teardowns in the [SaaS Google Ads examples](/examples/saas-google-ads-teardowns/) show how differently the same keyword behaves across ACV bands.

Category terms carry the volume and the pain. In security compliance, HR, CRM and project management, the top of page bid routinely sits between $30 and $80 because the incumbents are well funded and the ACVs are large. Vanta, Rippling and Monday operate in exactly those auctions, and a $3,000 ACV product cannot win a bidding war against them on the same keyword.

Problem and job-to-be-done queries are the cheapest clicks in SaaS and the most frequently misused. They work when the landing page answers the question and offers something smaller than a demo, and they fail when they point at a pricing page.

Broad match on category terms without a negative list is the fastest way to spend a budget on jobs, free alternatives and definitions. Start from a prebuilt [SaaS negative keyword list](/templates/saas-negative-keyword-list/) and add to it weekly for the first quarter.

## How to derive max CPC from ACV instead of guessing

Max CPC is not a preference. It falls out of two numbers: what you can afford to pay for a customer, and how many clicks it takes to get one.

Target CAC comes from gross profit and payback tolerance. Take annual contract value, multiply by gross margin, divide by twelve to get monthly gross profit, then multiply by the number of payback months your board will accept. A $25,000 ACV product at 78 percent margin generates $1,625 a month in gross profit, so a twelve month payback allows a CAC near $19,500 and a conservative seven month payback allows $12,000.

Click to customer rate is the chain, not a single number. Click to primary conversion, primary conversion to qualified, qualified to closed won. Multiply all three.

Then the formula is short: **max CPC = target CAC × click to customer rate**. Bid at 60 to 70 percent of that ceiling, because the ceiling is break even and you need margin for the months when win rates drop.

**Working out your own ceiling**

## Worked numbers at $3K, $25K and $120K ACV

Same formula, three very different accounts. The right campaign structure, conversion action and bid ceiling all change with the price point.

| | $3,000 ACV self-serve | $25,000 ACV sales-assisted | $120,000 ACV enterprise |
| --- | --- | --- | --- |
| Gross margin | 85% | 78% | 75% |
| Payback period accepted | 5 months | 7 months | 8 months |
| Target CAC | $1,000 | $12,000 | $60,000 |
| Primary conversion action | Trial start | Demo request | Demo request |
| Click to primary action | 6% | 3% | 2% |
| Primary action to closed won | 18% | 9.9% | 7% |
| Click to customer | 1.08% | 0.30% | 0.14% |
| Break even max CPC | $10.80 | $35.60 | $84.00 |
| Recommended bid ceiling | $7.00 | $24.00 | $57.00 |
| Target CPA on the action | $180 | $1,188 | $4,200 |
| Monthly spend for 30 conversions | $5,400 | $35,640 | $126,000 |

The last row is the one that ends most conversations. Smart bidding wants roughly 30 conversions per campaign per month to perform, and at enterprise ACVs that volume is unreachable at any sane budget. Enterprise accounts should optimise to an earlier action, usually a qualified form submission or a pricing page visit with a value attached, and accept that the bid strategy is learning from a proxy.

**$84** Break even cost per click at $120,000 ACV, a 2 percent click to demo rate and a 7 percent demo to won rate

The $3,000 ACV column explains why so many small SaaS products fail at paid search. A $7 bid ceiling cannot compete for category terms priced at $30, which leaves brand, long tail problem queries and a narrow set of competitor names. That is a real business, but it is a $6,000 a month business, not a growth engine, and you can test your own inputs against it in the [max CPC calculator](/calculators/saas-max-cpc-calculator/) before committing budget.

## Conversion tracking that keeps bids honest

This is the section to implement first, ahead of any ad copy work. If Google only sees form fills, it will find you more form fills, and the cheapest form fills come from people who will never buy.

The chain has four parts. Capture the GCLID from the URL into a hidden field on every form. Write it to the lead record in Salesforce, HubSpot or whatever holds the pipeline. When the record reaches a stage that matters, push an offline conversion back to Google Ads with that GCLID, a timestamp and a value. Then point the bid strategy at that imported action rather than the on-site form.

Values matter as much as the import. Send pipeline value at opportunity creation and actual revenue at closed won, and the algorithm will start preferring the traffic that produces large deals over the traffic that produces many small ones. Sending a flat value of 1 for every conversion is the same as running target CPA with extra steps.

With a 90 day sales cycle, closed-won import arrives too late to steer bids. Send a weighted value at an earlier stage instead: an opportunity worth $25,000 at a historical 22 percent win rate imports as $5,500. Reconcile to real revenue monthly and adjust the weights quarterly.

Two honest failure modes. GCLID is lost whenever a visitor clicks an ad, leaves, and returns organically before converting, which is common in a 60 day cycle; enhanced conversions for leads using a hashed email address recovers part of that but not all. And the import itself creates a reporting lag that makes week-to-week optimisation feel blind for the first two months. Both are worth the cost, and the mechanics are covered further in the offline tracking sections of the [PPC tools guide](/guides/ppc-tools-for-saas/).

## A 90 day ramp with weekly checkpoints

New accounts fail by doing everything in week one. The order below front loads measurement and holds spend back until the signal is trustworthy.

**Weeks 1 and 2.** Conversion tracking, GCLID capture, negative keyword list, brand campaign live with a hard daily cap. Nothing else runs. Checkpoint: a test lead appears in the CRM with a GCLID attached.

- **Weeks 3 and 4.** Competitor campaign and one category campaign, exact and phrase match only, manual CPC at the bid ceiling from your model. Checkpoint: search terms report is 80 percent relevant after two rounds of negatives.

**Weeks 5 to 8.** Add the problem tier with a lighter offer than a demo. Move campaigns to maximise conversions with a bid cap once each has passed 15 conversions in 30 days. Checkpoint: cost per primary action within 30 percent of target.

**Weeks 9 to 12.** Offline conversion import goes live, bid strategy switches to target CPA against the imported action, then to target ROAS once 30 days of values exist. Checkpoint: the pipeline value reported in Google Ads is within 20 percent of the CRM figure for the same period.

Budget allocation across the tiers should shift over that quarter, and the split that works at week 12 is rarely the split you started with. The [budget allocation guide](/guides/saas-ppc-budget-allocation/) has the tier by tier percentages by company stage.

## Where SaaS accounts actually leak money

Three leaks account for most of the waste I see, and none of them are about ad copy.

The first is optimising to the wrong action. An account running target CPA on newsletter signups will reliably drive the cost per signup down and the cost per customer up, because those two things are in tension whenever the cheap conversions come from unqualified traffic.

This second is campaign fragmentation. Fifteen campaigns with two conversions each give smart bidding nothing to learn from, and the platform will spend the whole month in a learning phase that never resolves. Consolidation is uncomfortable and usually correct, which is the argument running through the [account structure guide](/guides/saas-google-ads-campaign-structure/).

The third is unmanaged match type drift. Broad match plus smart bidding is Google's recommended setup and it works only when the conversion signal is accurate. Turn broad match on before offline import is live and you have handed the algorithm a fast way to find the wrong people. The full catalogue of these sits in [SaaS PPC mistakes](/guides/saas-ppc-mistakes/).

One tradeoff worth naming honestly: everything above makes the account slower to optimise. Offline import delays feedback, consolidation reduces granular control, and stage weighted values introduce modelling error. You are trading reporting comfort for a bid signal that matches the business, and the first six weeks feel worse than what you had.

## What to change in your account this week

Open the search terms report, add negatives, and check whether a GCLID is present on your last twenty CRM leads. If it is not, that is the whole week's work, and it matters more than any keyword expansion.

Then run the max CPC calculation on your own median ACV and compare the answer to the actual top of page bids in your category. If your ceiling is below the auction price, the honest conclusion is that paid search is a brand and competitor channel for you rather than a category one. Once the ceiling is set, the ad copy work has a constraint to write against, and the tested patterns are in the [SaaS ad copy swipe file](/templates/saas-ad-copy-swipe-file/), with the wider channel context on the [SaaS PPC hub](/saas-ppc/) and current performance ranges in the [PPC benchmarks](/research/saas-ppc-benchmarks/).

## Frequently asked questions

### Does Google Ads work for B2B SaaS?

Yes, on Search, for products where people already search for the category or a competitor. It works badly for products defining a new category, because the demand does not exist yet. The deciding factor is whether your target CAC divided by the click to customer rate produces a max CPC above the auction price in your category. Run that calculation before opening an account.

### What is a good CPA for SaaS Google Ads?

There is no universal number because CPA depends on what you count as a conversion. A demo request at $1,200 is excellent at $25,000 ACV and ruinous at $3,000. Derive the target from allowable CAC and the conversion rate between your primary action and closed won. At $25,000 ACV with a 10 percent demo to customer rate, roughly $1,200 per demo is break even against a $12,000 CAC.

### Should SaaS companies bid on their own brand terms?

Usually yes, in a separate campaign with a capped budget. Competitors bid on your name, and the click is cheap, typically $1 to $4. The honest caveat is that a portion of that traffic would have arrived organically, so treat brand as defensive insurance rather than acquisition and measure it on incrementality tests rather than on its flattering ROAS.

### Is Performance Max worth running for B2B SaaS?

Conditionally, and never as the first campaign. PMax will absorb brand search and report it as new conversions unless brand terms are excluded through a brand exclusion list and a dedicated brand Search campaign. It performs best for self-serve products with high conversion volume and poorly for enterprise motions where the conversion signal is thin and delayed.

### How do you track SaaS conversions properly in Google Ads?

Capture the GCLID in a hidden form field, store it on the lead record in the CRM, then import the offline conversion when the deal reaches a stage you care about. Send a value with it, ideally pipeline value at opportunity and revenue at closed won. Enhanced conversions for leads using hashed email covers the cases where the GCLID is lost.

### How much should a SaaS company spend on Google Ads to start?

Enough to generate roughly 30 conversions a month in the campaign you want smart bidding to run, which is the real minimum. At a $180 cost per trial that means about $5,400 a month; at a $1,200 cost per demo it means $36,000. If the budget cannot reach that, run manual CPC or maximise clicks with a bid cap and accept slower learning.

### What is the difference between target CPA and target ROAS for SaaS?

Target CPA optimises for a count of conversions and treats every lead as equal. Target ROAS optimises against a value you send, so a demo from a 500 person company can be worth ten times one from a solo user. ROAS is the better setting once offline import is live and values are accurate, and it is actively harmful before that.