Google Ads keyword economics for SaaS.
The keyword with the cheaper click is often the more expensive customer. SaaS search is a close-rate problem wearing a CPC costume.
Ecommerce search has a merciful property: the click is close to the revenue event. You can be crude about keyword value and still land near the truth, because the cart closes the loop in days. SaaS search does not have that property. The click is a trial, a demo, or a content download. The revenue event is a contract months later, inside a CRM the ad account cannot see. Ranking keywords by CPC, or even by cost per lead, is how a SaaS account funds the cheapest curiosity and starves the queries that actually close.
This piece is the keyword layer of the SaaS paid acquisition playbook. The playbook owns cash, payback, and channel sequence. The Google Ads for SaaS hire page owns how we run the account. What this page owns is the pricing of a query: how to decide which search terms deserve budget once you admit that CPC is an input, not a cost.
The unit of value is a closed customer, not a click
Take a category term at $18 CPC and 8% lead rate, against a problem-space term at $6 CPC and 12% lead rate. The cheap term looks like the winner on the dashboard. Now attach a close rate. Category terms in a working SaaS account often close at two to four times the rate of problem-space terms, because the searcher already named the category. Problem-space traffic is earlier, broader, and full of people who wanted a PDF.
Category term
CPC $18, lead rate 8% -> $225 per lead
SQL rate 40%, close 25% of SQL -> 10% of leads become customers
Cost per customer = $225 / 0.10 = $2,250
Problem-space term
CPC $6, lead rate 12% -> $50 per lead
SQL rate 15%, close 12% of SQL -> 1.8% of leads become customers
Cost per customer = $50 / 0.018 = $2,778The cheap click is the more expensive customer. That inversion is the whole essay. If your CRM loop is closed, you will see it in 60 to 90 days. If it is not, you will keep buying the $6 click until sales tells you the leads are junk, which they will, late.
Cost per lead is not a SaaS KPI. It is a volume dial. The number that prices a keyword is cost per customer, or cost per qualified opportunity if the sales cycle is long, computed from the actual close rates of that intent tier. Anything else is a proxy that the auction will game.
Three intent tiers, three prices
We price SaaS search in three tiers, each with its own budget, target, and landing experience. Mixing them in one campaign with one tCPA is how the cheap tier eats the expensive one.
- Category terms: the product class, named. Highest close rate, highest CPC, usually the first tier worth scaling. Landing page should be the product, not a guide.
- Competitor terms: the buyer already has a shortlist. Intent is real, close rates are lower, CPCs are ugly, and the landing page has to be comparison-honest or you are paying to lose. Fund as an experiment inside a working category tier, not as the first campaign.
- Problem-space terms: the job to be done, before the category has a name. Cheap clicks, weak close rates, useful for education if you measure pipeline rather than leads. Easy to overfund because volume looks like progress.
Brand is a fourth bucket and it does not belong in any of the three. Brand search in SaaS is often a late-stage assist, not acquisition. Mixing it with non-brand is the same measurement error as in ecommerce, just with a longer delay. Separate it, report it, and do not let it flatter the CAC of the account. The argument is in brand vs non-brand.
The competitor-term test
Competitor keywords are the place SaaS teams waste the most money with the most conviction. The searcher is in-market. The CPC is a badge of seriousness. The landing page is a feature table. Then sales reports that those leads already had a favorite, wanted a discount, or were doing vendor due diligence for a deal that was going to close anyway.
The test is arithmetic, not opinion. Price the customer from that tier alone. If cost per customer on competitor terms clears payback at your ACV and close rate, keep them in their own campaign with their own target. If they only look efficient when blended with brand or category, they are a parasite. Comparison-honest pages help; they do not repeal the close-rate gap.
Volume is a trap when the funnel is long
Smart Bidding will find you cheaper leads if you ask it to. In SaaS that is often a request to find worse leads. The machine cannot see the contract. It can see the form. Give it a conversion event of "demo booked" with no offline import of qualified stages, and it will hunt volume at the quality the form allows.
That is why lead-gen bidding belongs next to this page. Keyword economics without a qualified conversion event is theater: you will price the tiers correctly on a spreadsheet and the account will still buy the wrong traffic, because the bid strategy is optimizing a different object.
The practical sequence is: close the CRM loop, then split the tiers, then let bidding optimize to the qualified event on each tier. Doing the keyword architecture first, on a raw lead conversion, just builds a more elaborate way to buy junk.
What a good SaaS keyword still has to earn
A term stays funded if, on a trailing window long enough for your sales cycle, cost per customer (or per SQL, early) sits inside payback. It gets paused if volume is high and qualified rate is structurally worse than the tier next to it, not because CPC ticked up. CPC inflation is weather. Close-rate collapse is the climate.
If you do not yet have close rates by query, use SQL rate by campaign as the standing proxy and treat it as provisional. Do not invent a close rate from a blog benchmark. The SaaS metrics piece is the scoreboard; this page is the bid.
Where this advice stops
PLG products with a free trial as the only conversion, and no sales team, have a different object: activation and conversion-to-paid, not SQL. The tiers still apply. The prices change. Self-serve search that cannot see activation will overbuy problem-space the same way a sales-led account overbuys cheap demos.
This is also not a hire-page substitute. If you want the operating system (offline imports, value-based bidding, the engagement), that is Google Ads for SaaS. If you want the cash constraint that sits above any keyword, that is the playbook. This page is only the query.
Should SaaS companies bid on cheap high-volume keywords?
Only after you have priced them in customers, not in clicks. High-volume problem-space terms often produce the cheapest leads and the most expensive customers. Fund them when SQL or close rates support the CAC. Pause them when they only look efficient on cost per lead.
How do I decide if competitor keywords are worth the CPC?
Run them in their own campaign and compute cost per customer from that tier alone. If they only look efficient when blended with brand or category terms, they are borrowing credit. Comparison-honest landing pages are required; they do not replace the close-rate test.
What is a good CPC for Google Ads in SaaS?
There is not one. A $22 category click that closes at 10% of leads can beat a $5 problem-space click that closes at 2%. Price the customer from your ACV, margin, and payback, then back into an allowable cost per lead per tier. The CPC is whatever the auction charges inside that ceiling.
Why does my SaaS account get leads that sales ignores?
Usually because bidding is optimizing a form fill, and the cheapest fills are the least qualified. Close the CRM loop so qualified stages come back as offline conversions, split intent tiers, and stop reporting cost per lead as the governing number.
Written by Sophie Mills, performance marketing strategist. If this resonated and you want to apply it to your own account, you can book a strategy call or run a free audit.
How we research, source figures, and handle corrections: editorial policy.