Skip to main content

Setting a Google Ads budget for SaaS

The right Google Ads budget depends on what you're trying to do. A budget sized to learn looks different from a budget sized to scale. This article covers three common frameworks, when each applies, and the math.

The three frameworks​

In rough order of sophistication:

  1. Test budget. Sized to produce enough data to learn whether a campaign or cluster works.
  2. Target CPA budget. Sized to deliver a specific number of conversions at a specific cost.
  3. Percent of revenue budget. Sized as a fixed percentage of revenue, with paid search competing against other line items.

Most teams use a mix as they mature.

Framework 1: test budget​

For a campaign you've never run before. The goal is data, not conversions.

How to size it​

Estimate two things:

  • Average CPC for the cluster (Google's keyword data, weighted toward the high end to be conservative).
  • Click volume needed for Google's algorithm to learn, roughly 50 to 100 clicks per week minimum.

Multiply: CPC × clicks per week × 4 weeks ≈ monthly test budget.

For B2B SaaS clusters with $5 to $20 CPCs and a 75 click per week target, that lands in the $1,500 to $6,000 monthly range.

When to stop testing​

After 2 to 4 weeks of data. By then you should know:

  • CTR (acceptable at 2 to 4 percent for most B2B SaaS).
  • Conversion rate (acceptable at 1 to 5 percent for sign ups; lower for demos).
  • CPA (compared to your target).

If the campaign is hitting numbers you like, transition to scale. If not, kill it and try a different cluster, copy, or landing page.

Don't over invest​

A common mistake: running a test campaign at $20,000/month "to give it a real chance". You don't need that to learn. You need just enough volume to see clean numbers. Save the bigger budget for once you have signal.

Framework 2: target CPA budget​

For a campaign you understand well enough to size against a specific outcome.

The math​

Let:

  • C = your max acceptable CPA (cost per conversion).
  • CPC = the cluster's average cost per click.
  • CR = your conversion rate (conversions divided by clicks).
  • N = conversions you want per month.

Then: monthly budget = N × C, given that CPA = CPC / CR.

If your CPA target is $200 and the cluster delivers $200 CPA at scale, every $200 spent produces one conversion. Want 50 conversions per month? Budget $10,000.

What if CPA is above target​

If observed CPA is higher than your target, you have three options:

  • Lower CPA through better keywords, copy, or landing page. Often the right move first.
  • Accept higher CPA if downstream economics support it. A $400 CPA can be fine if customers are worth $20,000.
  • Walk away from the cluster. Some clusters are just expensive. Move budget to a cheaper one.

Budget alone won't fix CPA. Throwing more budget at a campaign with bad CPA produces more bad CPA.

When this framework breaks​

  • You don't know your CPA target. Common for early stage products. In that case, default to test budget framework.
  • Conversion data is too thin to estimate CR. Run test budget first, accumulate data, then switch to CPA targeting.
  • Conversions don't directly correlate with revenue. A sign up CPA is meaningful only if some predictable percentage become customers. If sign up to customer rate is wildly variable, optimize on something further down the funnel.

Framework 3: percent of revenue budget​

For mature operations where paid search is one channel among many.

How it works​

Pick a target marketing spend as a percentage of revenue (often 10 to 30 percent of ARR for early stage SaaS, lower for mature). Within that, allocate a percentage to paid search based on its proven efficiency relative to other channels.

If you do $5M ARR, target 20 percent on marketing ($1M total), and paid search is 30 percent of that ($300K, or $25K/month), you have a Google Ads budget of $25K/month split across campaigns.

When this works​

  • You have multiple proven channels and need to allocate among them.
  • You're managing to a CAC payback or LTV/CAC target at the company level.
  • Finance wants stable, predictable spend.

When this breaks​

  • You're earlier stage and don't have stable revenue. Percent of revenue budgets ratchet down with revenue dips, which is when you need acquisition most.
  • Your channels have different maturity. Spending 30 percent of marketing on Google Ads when Google Ads isn't yet proven is investing in unknown.

The starting point​

If you're starting fresh:

  • Run test budget for 6 to 8 weeks across 3 to 5 clusters.
  • Identify the one or two clusters that are converting at a reasonable CPA.
  • Move to target CPA budget on those, scaling up to your target volume.
  • Add percent of revenue framing once paid search is one of several proven channels and you're managing portfolio level.

This trajectory takes 3 to 6 months for most B2B SaaS. Don't try to skip steps.

Common budget mistakes​

Daily budget too low to win auctions​

A daily budget of $20 in a cluster with $10 CPC means you'll get 2 clicks on a good day, often zero. Either accept the campaign won't generate meaningful data, or raise the budget.

Daily budget too high for a learning phase​

Running $500/day on a brand new campaign means you'll burn $15,000 before the algorithm has learned what works. You'll learn from $5,000 just as well, with $10,000 left for the campaign that actually performs.

Changing budgets too often​

Google's algorithm needs days to recalibrate after a budget change. Frequent changes (weekly or more) prevent it from ever stabilizing. Set budget, leave it for at least 2 weeks, then evaluate.

Treating budget as the lever​

Budget sets ceiling. Performance is determined by keywords, copy, landing page, and bidding strategy. If your campaign isn't working at $1,000/month, it won't work at $5,000/month either.

How Hero Marketer suggests budgets​

The campaign wizard suggests a monthly budget when you reach the budget step. The suggestion is sized for "test budget" framework: enough volume for Google to learn, not optimized for a specific CPA target.

You can override the suggestion. See Set your budget for the in product details and How budget suggestions are calculated for the math.

Going further​