A £10,000 monthly Google Ads budget can look healthy in the platform while producing a pipeline that sales cannot close. The difference is rarely a lack of keywords. It is usually a bidding system trained to pursue the wrong conversion. The best bidding strategy for SaaS is the one that gives Google a reliable signal of commercial intent, then protects CAC as spend scales.
That sounds obvious. It is not how many SaaS accounts are set up. Form fills, pricing-page visits and generic leads are often treated as equal successes. They are not equal. A demo request from an ICP account with a real use case has a radically different expected value from a student downloading a template.
The best bidding strategy for SaaS starts with the revenue event
Google Ads bidding is only as intelligent as the conversion data it receives. If the platform is optimising towards low-quality lead volume, it will find more low-quality lead volume – often very efficiently.
For B2B SaaS, the most useful conversion hierarchy usually begins with a primary action such as a qualified demo, sales-qualified lead or opportunity created. A lower-friction action, such as a trial start or contact form submission, can still be measured, but it should not automatically be the signal that determines where budget goes.
The right event depends on the business model. A product-led SaaS company with a short path from trial to paid account may successfully optimise towards activated trials. A sales-led platform with a six-figure annual contract value should generally optimise towards qualified meetings and later-stage CRM outcomes. There is no universal event that works for every SaaS company.
The commercial test is straightforward: would you be happy if Google delivered 100 more of this conversion next month? If the answer is no, it should not be your primary bidding goal.
Start with Maximise Conversions, then earn the right to use targets
For most accounts with accurate tracking and enough recent conversion volume, Maximise Conversions is the sensible starting point. It lets the system use real-time auction signals without forcing an arbitrary cost target too early.
A target CPA can work well, but only when it reflects economics rather than wishful thinking. Set it too low and impression share collapses, volume dries up and the campaign loses the data it needs to improve. Set it too high and the account can buy pipeline at a cost the business cannot defend.
The practical sequence is usually this: establish clean conversion tracking, use Maximise Conversions to generate stable data, then introduce a target CPA once actual lead quality and sales feedback show a repeatable acquisition cost. Do not judge the result after three days. SaaS buying cycles, particularly in mid-market and enterprise segments, require patience and CRM validation.
Calculate target CPA from CAC, not a platform benchmark
A benchmark CPA is irrelevant if it does not fit your unit economics. Start with the maximum CAC your business can support, then work backwards through the funnel.
If a customer is worth £30,000 in gross profit over its lifetime and your acceptable CAC is £9,000, that does not mean you can bid £9,000 for a lead. You need to account for the conversion rate from lead to qualified opportunity, opportunity to closed-won, and any sales cost that sits outside paid media.
For example, if one in four qualified demos becomes an opportunity and one in four opportunities closes, you need 16 qualified demos to create one customer. A £9,000 paid acquisition allowance implies a maximum £562.50 cost per qualified demo before sales costs and attribution adjustments. That number may need to be lower if paid search is only one part of the acquisition mix.
This is why bidding conversations should involve the person responsible for revenue, not only campaign management. The platform can optimise towards a CPA. It cannot decide whether that CPA creates a viable business.
Use value-based bidding when lead values genuinely differ
Target ROAS and Maximise Conversion Value are powerful for SaaS businesses with meaningful variation in account value. They are also easy to misuse.
Value-based bidding makes sense when you can pass values that reflect likely revenue or gross profit. This could mean assigning different values by company size, product tier, geography, lead score or opportunity stage. An enterprise demo from a target account should carry more weight than a small-business enquiry if the expected contract values justify it.
The key word is expected. Inflated values do not improve bidding. They contaminate it. If a lead is marked as high value because someone visited the enterprise page, but sales consistently rejects that segment, Google will spend more aggressively on a false signal.
For mature SaaS teams, the strongest set-up is often offline conversion import from the CRM. Feed back qualified leads, opportunities and closed-won customers with accurate values. This closes the gap between ad click and revenue outcome. It also gives bidding models a chance to learn which searches, audiences and devices produce pipeline rather than merely enquiries.
That said, offline data is not a substitute for volume. If your sales cycle is nine months and you only close a handful of deals each quarter, waiting for closed-won revenue alone can make optimisation painfully slow. In that situation, use an earlier milestone with a proven relationship to revenue, such as a properly defined sales-qualified lead or opportunity.
Separate bidding decisions by intent and campaign maturity
A single bidding strategy across all campaigns often creates avoidable waste. Brand search, high-intent non-brand search, competitor terms and broad discovery activity behave differently. They should not be judged by identical rules.
High-intent non-brand campaigns usually deserve the clearest path to a qualified demo or trial. This is where a target CPA can become effective once data is stable. Brand campaigns may need less automation because they are capturing existing demand and can make performance appear better than it is. They still matter, but they should not mask weak non-brand acquisition.
Broad match can be highly profitable for SaaS when the conversion signal is strong and negatives are actively managed. It can also consume budget on vague research queries when tracking is weak. The question is not whether broad match is good or bad. The question is whether the account has enough quality data to guide it.
Campaign maturity matters too. A new campaign needs room to learn. A mature campaign with a dependable flow of qualified conversions can be held to a stricter target. Applying mature-account efficiency rules to a new segment usually suppresses growth before it has a chance to prove itself.
Protect bidding performance with better conversion design
Bidding cannot repair a weak landing page or an unclear offer. If paid clicks arrive on a page that asks too much, explains too little or gives no reason to book now, Google will simply pay more to find the few users who persist.
For demo-led SaaS, reduce friction without reducing qualification. Ask for the information sales genuinely needs. Make the outcome specific. A visitor should understand whether they are booking a product walkthrough, an implementation discussion or an assessment of fit. Vague calls to action create vague leads.
Also exclude noise from the bidding signal. Spam, duplicate submissions, existing customers, job seekers and unsupported regions should not count as successful conversions. These details sound operational, but they directly affect how budget is allocated.
What to monitor after changing strategy
Do not evaluate a bidding change on CPA alone. Watch qualified-demo rate, opportunity creation rate, pipeline generated, expected revenue, CAC and sales feedback alongside platform metrics. A lower CPA is a loss if lead quality drops far enough to reduce pipeline.
Look for the lag between click, conversion and qualification. If sales qualification happens two weeks after the initial lead, making daily CPA decisions can cause you to cut campaigns that were actually producing the best opportunities. Build reporting windows that match your real sales process.
The objective is not to make Google Ads look efficient. It is to make paid search a dependable source of commercially viable demand.
Book a 30-minute meeting to review the bidding signals, CAC targets and pipeline leakage in your Google Ads account.
Frequently asked questions
Should SaaS companies use target CPA or target ROAS?
Use target CPA when the primary outcome has a reasonably consistent value, such as a qualified demo. Use target ROAS when you can reliably pass different expected values from your CRM. If values are estimated badly, target CPA is usually safer.
How many conversions are needed before using target CPA?
There is no fixed threshold, but a campaign needs a consistent recent flow of meaningful conversions. If volume is sparse or highly volatile, Maximise Conversions without a restrictive target often gives the system more room to learn.
Should a free trial be the primary conversion?
Only if trial users reliably activate and convert to paid customers. If many trials never reach product value or match your ICP, optimise towards activated trials or qualified sales outcomes instead.
Why did CPA rise after switching to a smarter bidding strategy?
The system may be exploring more auctions, adapting to a new target or responding to a changed conversion signal. Check lead quality and pipeline before reacting. A higher initial CPA can be justified if qualified opportunity rate improves.
Can broad match work for B2B SaaS?
Yes, provided conversion tracking reflects genuine commercial intent and search-term hygiene is maintained. Broad match magnifies both good data and bad data, so it should not be used as a substitute for proper qualification.
How often should target CPA be changed?
Avoid frequent, dramatic edits. Change targets in measured increments, then allow enough time for the sales and conversion lag to become visible. Constant intervention prevents a clear read on performance.
The strongest bidding strategy is not the most automated setting in Google Ads. It is the one grounded in trusted revenue data, realistic economics and a clear definition of what a valuable customer looks like.