PLG Analytics Without Consent Banners: Trial-to-Paid Measurement in 2026

If you run growth at a European SaaS company, you have one version of this conversation every quarter. Finance asks what the CAC is. You quote a number from your standard analytics tool. They ask how confident you are. You shrug. You know that, in our experience with clients, between 40% and 60% of traffic doesn't accept cookies, that your Google Ads spend converts to signups the attribution model can't see, and that the number you just quoted is optimistic by some unknown factor.

Product-led growth depends on a measurement stack that works. When the marketing-site layer is broken, every downstream CAC calculation, every paid-channel ROI decision is made on partial data. Cookieless analytics fixes the top of the funnel — not by tracking harder, but by counting events in aggregate and attributing each signup last-click at channel level. For the cross-vertical category overview, see the cookieless analytics pillar; this post is the SaaS-specific reading.

Why the SaaS measurement stack breaks at the top

A standard PLG measurement stack has three layers:

1. Marketing-site attribution (typically the standard market analytics tool): traffic source → landing page → signup.
2. Product event analytics (typically your product analytics tool): in-product actions → activation → retention cohorts.
3. Warehouse / revenue (typically BigQuery or Snowflake): subscription tables, MRR, churn.

The top layer is the one that breaks. After consent rejection and ad blockers, the standard market analytics tool doesn't see part of your traffic, and how much depends on the site; on Incapto's Shopify store, measured over 48 days, it did not record 29% of visits. Even the signups it sees are attributed poorly, because the cookie that would have preserved the source was rejected or expired. The in-product analytics tool works fine — the user has authenticated, cookies are not the problem, and user-level tracking is compliant on explicit consent inside the product. But the marketing-to-signup layer is unreliable.

What cookieless analytics replaces (and doesn't)

Cookieless analytics for SaaS replaces the standard market analytics tool at the marketing-site layer. It does not replace your product analytics tool anywhere. The two systems have different jobs:

• Cookieless analytics (marketing site, aggregated). Counts visits and signups by channel, campaign and landing page. No user identification, no per-visitor journey.
• Product analytics tool (authenticated product, user-level). Tracks what each logged-in user does inside the app. Retention cohorts, feature adoption, in-app funnels.

What cookieless analytics covers:

• Marketing-site visit counts by source
• Campaign-level signup conversion counts
• Landing-page test measurement without sampling
• Paid-channel ROI at the trial-signup event
• BigQuery / Snowflake export for joining against in-product data

What it does not cover:

• In-product feature adoption
• Retention cohort analysis
• Per-user behavior inside the app
• Session replay

The PQL measurement pattern

A working PQL definition for European PLG SaaS requires joining three datasets in the warehouse:

1. Aggregate marketing attribution from cookieless analytics: acquisition channel, campaign, UTM — stored in aggregate with signup event metadata.
2. Trial signup records from the product database: email, signup timestamp, plan selected (user-level once authenticated).
3. In-product activation from your product analytics tool or your own event log: activation event, time-to-value (user-level).

The join happens at the warehouse, not in the marketing tracker. Because cookieless analytics exports signup attribution to BigQuery or Snowflake at full resolution (no sampling), you get clean channel metadata on every signup. You then join on your own user/email column to add the activation side. A PQL definition like “signed up from paid search + activated within 7 days” becomes a single SQL query.

Implementation for a European PLG SaaS

The setup that works for most teams:

1. Install cookieless analytics on the marketing site only. Leave your product analytics tool alone in the app.
2. Configure BigQuery / Snowflake export.
3. Run side-by-side with your standard analytics tool on the marketing site for 30 days.
4. Write the join query (channel attribution + signup + product activation).
5. Switch channel-level CAC reporting to the new numbers when the side-by-side comparison stabilises.

No cookie banner for the marketing-site analytics itself (keep the banner for advertising pixels if needed). No migration for in-product analytics. No change to the product team's workflow. If you want to check which plan includes warehouse export before you start, the pricing page lists what's included at each tier.

Questions SaaS teams ask