The SaaS Ad Problem Is Not the Ad
Most SaaS founders who come to an audit have already done the obvious things. They have A/B tested their ad copy. They have refined their audience. They have tried different landing page layouts. The cost per trial signup is still too high and the payback period on customer acquisition is stretching longer every quarter. The agency sends a report showing improving click-through rates and they wonder why the bank account does not reflect it.
The problem is almost never the ad. The ad is the part everyone looks at because it is visible. The problem is the infrastructure the ad sends people to, and the infrastructure the campaign uses to make decisions about where to send the next person.
Performance Max and the Algorithm Tax
Performance Max is Google's fully automated campaign type. You give it a budget, some creative assets, and a conversion goal, and it decides everything else — where to show the ad, who to show it to, how much to bid, which creative combination to use. Google's pitch is that the AI optimises better than a human can. For SaaS campaigns this is often true in theory and expensive in practice.
The reason is conversion data quality. Performance Max optimises on the conversion signals you give it. If your conversion tracking captures 70 percent of your real signups because 30 percent happen on browsers or devices where your client-side tag does not fire correctly, the algorithm is training on a partial dataset. It learns which audiences and placements appear to convert based on the 70 percent it can see. The 30 percent it cannot see distorts the picture in ways that compound over weeks of campaign learning.
The result is a campaign that has confidently learned the wrong lessons. It over-invests in audiences that happen to convert in ways the tracking can measure and under-invests in audiences that convert in ways it cannot see. This is the Algorithm Tax at scale. A SaaS company spending $20,000 a month on Performance Max with 25 percent tracking gaps is potentially wasting $5,000 to $6,000 a month on decisions the algorithm made with incomplete information.
Trial Signup Pages and the Three-Second Rule
SaaS landing pages tend to be better built than most because the founding team usually has technical people involved. But better than most is not the same as fast enough for paid traffic. The benchmark for a paid traffic landing page is not a good desktop experience. It is a mobile load time under two seconds on a throttled 4G connection, which is the condition Google uses when evaluating landing page experience for Quality Score.
A one-second improvement in mobile load time increases SaaS trial conversion rates by seven to twelve percent in most studies. For a company spending $20,000 a month on Google Ads with a 2.5 percent trial signup rate, a one-second improvement means roughly 175 to 300 additional trial signups per month from the same budget. At a 20 percent trial-to-paid conversion rate and a $100 monthly subscription, that is $3,500 to $6,000 in additional monthly recurring revenue from an infrastructure change that costs a fraction of that to implement.
The reason this does not get fixed is that it requires the engineering team's time and the engineering team is building the product. The reverse proxy approach solves this without engineering involvement. A Cloudflare Worker deployed in front of the trial signup landing page delivers the performance improvement at the infrastructure level. The product team never needs to be involved.
Signal Loss and What It Does to Your Retargeting
SaaS companies typically have sophisticated retargeting setups — visitors who did not sign up get followed across the internet with ads designed to bring them back. The effectiveness of this retargeting depends entirely on the accuracy of the audience data feeding it.
If your pixel or your GA4 tag is missing 25 percent of trial page visitors because of browser privacy settings, iOS restrictions, or ad blocker interference, your retargeting audience is 25 percent smaller than it should be. You are not retargeting everyone who showed intent. You are retargeting everyone who showed intent and happened to be using a browser that did not block your tag. That selection bias means your retargeting audience over-represents a specific type of user and under-represents others, which skews the retargeting performance data and the campaign decisions that follow.
Server-side event tracking captures the full audience regardless of browser behaviour. The retargeting pool becomes complete. The algorithm has accurate data about who visited and who did not convert. The retargeting campaigns make better decisions and the cost per reactivation comes down.
The Burn Rate Calculation
For VC-backed SaaS companies the conversation about ad infrastructure is a burn rate conversation. Every dollar of ad spend that goes to waste because of slow landing pages, incomplete tracking, or miscalibrated Performance Max campaigns is burn that does not produce MRR. At Series A burn rates, a 20 percent improvement in cost per trial signup from infrastructure fixes rather than budget increases extends runway without requiring an additional funding round.
The Technical Tax Audit for SaaS campaigns goes through Performance Max settings and conversion data completeness, landing page speed on the specific trial signup URLs, tracking verification across the full conversion flow, and a calculation of the monthly budget impact of each issue found. The calculator on the homepage gives a rough estimate in 30 seconds based on your monthly spend and current load time. The full audit goes deeper into the specific mechanisms driving the waste.
Ready to Fix This on Your Site?
Run your site through PageSpeed Insights and send me the score. Free diagnosis. No pitch. Just the numbers.