The single most consequential setting in a B2B SaaS Google Ads account is also one of the least discussed: which conversion action you tell Smart Bidding to optimize toward. For most SaaS the choice comes down to two options — a free-trial (or product) signup, or a demo request — and it is not a copy decision or a landing-page detail. It is the objective the algorithm spends your entire budget pursuing. Choose wrong and every other optimization is downstream of a machine chasing the wrong outcome.
The honest answer is that neither is universally better; they win on different axes. As MarTech's analysis of free trial vs demo in B2B SaaS ads found, free-trial CTAs drove close to twice the click-through of demo CTAs — but demos attract higher-intent, better-qualified buyers. The job is to match the conversion action to your economics, your product, and how much data Smart Bidding can actually get.
ACV decides the motion first
Start with average contract value, because it caps what you can afford to spend on human sales effort. A demo is a person's time, and below a certain deal size that time cannot be recovered. The working rule in the GrowthSpree free-trial-vs-demo decision framework is blunt: under roughly $15K ACV, demo time is uneconomic. At that price point a self-serve free trial is the efficient motion, and your Google Ads should bid to signups, with activation and paid conversion measured downstream.
Above ~$15-25K ACV, the math flips. Deals are large enough to justify a salesperson, buyers expect a guided evaluation, and a demo request becomes the conversion worth paying for. This is the same logic that separates a product-led growth motion from a sales-led one, applied specifically to the bidding signal. Get ACV clear before you touch the campaign, because it determines everything that follows.
The real trade-off: volume vs intent
Free trials and demos sit at opposite ends of a commitment spectrum, and that spectrum is the whole trade-off. Low friction means more people act: free trials reliably generate several times more qualified leads per dollar of spend, because signing up costs a prospect almost nothing. High friction means fewer but better people act: a demo request is a stronger buying signal, and those leads convert to revenue at a much higher rate — MarTech's data points to demos generating on the order of 11x the revenue per qualified lead even as free trials produce many more leads overall.
Neither figure is the answer on its own; together they define the choice. A high-volume, low-intent signal (trials) gives Smart Bidding lots of data but risks optimizing toward users who never pay. A low-volume, high-intent signal (demos) points straight at revenue but can starve the algorithm of the conversions it needs to learn. This is exactly the tension behind optimizing for SQLs rather than raw leads — you want the signal closest to revenue that still fires often enough to train the model.
Smart Bidding needs enough conversions to learn
The most common reason to reconsider bidding to demos is not strategy — it is data volume. Smart Bidding is a learning system, and it needs a steady flow of conversions to bid well; a practical working floor is roughly 15-30 conversions per campaign per month. Demo requests, especially on non-brand terms or at launch, are frequently too sparse to clear that bar. When you bid to a signal that fires a handful of times a month, CPA gets volatile and the algorithm never stabilizes.
When demo volume is thin, do not force it. Bid to a higher-volume upstream action — a free-trial signup or a qualified micro-conversion — and import demo requests, SQLs, and closed-won deals as offline conversions from your CRM so Google still learns which clicks lead to real pipeline. This is the same problem I cover in running Google Ads on thin conversion data: the fix is to move the bidding signal up the funnel until it fires often enough, then let offline data re-anchor it to revenue.
Use one primary signal, measure both
You do not have to choose only one action to exist — you have to choose only one to bid to. Google lets you mark conversion actions as "primary" (used by Smart Bidding) or "secondary" (tracked but not optimized to). The right setup for most accounts is a single primary conversion per campaign — trial or demo, whichever matches your ACV — with the other action and downstream events set to secondary so you still see the full picture without splitting the algorithm's objective.
Running free trial and demo as co-equal primary conversions is a quiet but expensive mistake. A trial signup and a demo request arrive at completely different volumes and are worth completely different amounts, so treating them as equal tells Smart Bidding to average two incompatible goals. If you genuinely need to optimize to both — say a PLG product with an enterprise motion — separate them into different campaigns with different primary signals rather than blending them, following the same intent-tiering discipline as my B2B SaaS campaign structure guide.
Tie the choice back to CAC and revenue
Whichever action you bid to, remember what Smart Bidding actually does: it finds more of whatever you tell it to value. Bid to raw trial signups and it will hunt the cheapest signups, which are often the least likely to pay — efficient on cost-per-conversion, ugly on real cost per SQL. Bid to demos with too little volume and CPA spikes on a noisy signal. Either way, the number that matters is CAC against the customers who actually close, not the cost of the form fill.
The durable fix is the same regardless of which action you pick: feed closed-won revenue back to Google so bidding optimizes to customers, not conversions. My guides on the conversion value ladder and conversion tracking for SaaS cover wiring that up. Decide the primary action from your ACV and data volume, keep it singular per campaign, and anchor it to revenue — that is the whole decision. If you want a specialist to set this up, see how I run SaaS Google Ads management, or start with a Google Ads audit to check whether you are bidding to the right conversion in the first place.