Self-serve SaaS is a different game to sales-led B2B, and Google Search campaigns built for one rarely work well for the other. I have spent time recently setting up search acquisition for a self-serve product line inside a larger B2B business, and the gap between "how we run search for our enterprise motion" and "how we should run it for people who sign up and pay with a credit card" is bigger than most teams expect.
In a sales-led funnel, the conversion event is usually a demo request or a qualified lead. In self-serve, it is a signup, and ideally a paid signup or a trial that converts. If you point your self-serve campaigns at a "book a demo" conversion action, or worse, inherit the account's existing conversion tracking from the enterprise motion, you are optimising the algorithm towards the wrong outcome entirely. It will happily find you people who want to talk to a salesperson, which is precisely the group least likely to complete a self-serve signup.
The right setup tracks signup as a mid-funnel event and trial-to-paid or first payment as the event you actually optimise towards once you have enough volume to support it. Early on, before payment volume exists, optimise towards signup and accept that you are buying a slightly noisier signal in exchange for having any signal at all.
Self-serve products tend to share category keywords with enterprise alternatives, but the buyer intent behind the same search term can be completely different. Someone searching your category term might want a five-minute self-serve setup, or might want an enterprise deployment with SSO and a dedicated account manager. Serving both groups the same ad and the same landing page loses one of them.
Negative keyword lists matter more here than in most accounts. Terms that signal enterprise intent, "enterprise," "on-premise," "SSO," procurement-related terms, industry compliance terms, should be excluded from self-serve campaigns and routed to whatever landing experience serves that audience, if one exists. Without this, self-serve spend quietly subsidises interest from people who were never going to complete a self-serve signup.
Enterprise landing pages sell trust: logos, case studies, security credentials, a form. Self-serve landing pages need to sell the actual product experience: what it does, how quickly someone can get value, pricing that is visible rather than hidden behind "contact us." If your landing page still asks for a work email and a company size before showing a price, you have built an enterprise page and pointed self-serve traffic at it.
Self-serve products, particularly ones that are new or serving a narrower use case within a larger company, often do not have the conversion volume for Target CPA or Target ROAS to work well from day one. Starting on Maximise Conversions with a realistic budget, and graduating to a value-based bidding strategy once you have weeks of consistent signup data, avoids the erratic early performance that comes from asking the algorithm to converge on too little information.
Paid traffic is more exposed to free trial abuse than organic traffic because it is actively being targeted rather than arriving with existing intent. If your trial-to-paid conversion rate on paid channels is meaningfully worse than organic, check whether you are attracting genuine evaluators or people signing up opportunistically for a free tier with no intent to pay. This shows up in the data as decent signup volume and poor downstream conversion, and it is worth diagnosing before you conclude the channel does not work.
If you are trying to stand up Google Search for a self-serve motion inside a larger business, or separate it properly from an existing enterprise account, get in touch.