How do I measure SaaS trial-to-paid conversion by channel?
Save each sign-up's first source on the account, then divide paying customers by trials for every channel separately. Compare only after 30 days and 20+ trials per channel.
An overall trial-to-paid rate hides the most useful thing you could know. A blended 18% can be a 30% channel and a 9% channel averaged together, and the cheap channel may be the weak one. To see the split, every trial needs a source stored at sign-up, not reconstructed later from analytics sessions that expire or get blocked by cookie banners.
Expect real differences between channels. In our model of typical SaaS conversion rates, about 22% of trials from paid search convert, while trials from a targeted email sequence convert at around 30%, because those people already knew why they signed up. A worked example: 140 PPC trials bring 29 customers (20.7%) and 37 email trials bring 11 (29.7%). The blended 22.6% says nothing about where to spend next.
Pick one attribution rule and keep it. First touch, where the person first came from, suits early SaaS with short cycles; last touch suits products that people revisit many times before trying. For sales-assisted trials, add a field for the source of the first sales conversation, which keeps the line between MQLs and SQLs visible in the same report.
Setting up a channel report
- 1Capture UTMs at sign-upStore utm_source, utm_medium and the referring domain in hidden form fields and save them on the account, so the source survives cookie loss and device switches.
- 2Ask one self-reported question'Where did you hear about us?' with five options catches podcasts, word of mouth and AI chat recommendations that UTMs never see.
- 3Group into five to seven channelsPaid search, paid social, organic, outbound email, referrals, partners and direct. Fewer groups give each one enough trials to compare.
- 4Count by trial cohortAssign each customer to the month their trial started, not the month they paid, and close a cohort only when the trial plus 30 days has passed.
- 5Give calls their own sourceIf prospects reach you through sales conversations, for example short intro calls set up on Meetcatcher, tag them separately. Call-led trials often convert very differently from self-serve ones.
Once the table exists, put cost per trial next to trial-to-paid for each channel. A channel with a lower rate can still win if its trials are cheap enough, which is the logic behind a typical B2B customer acquisition cost.