Funnelsight is a fictional company. This site was built as a training project and does not offer a real product.

Start tracking your PLG funnel

Updated

To start tracking a PLG funnel, three metrics are enough: sign-ups, activation and trial-to-paid conversion, each read by source. A channel can bring plenty of sign-ups and few paying customers, and only the three together show it. The other metrics in the funnel metrics cheat sheet can wait until these are in place.

The three metrics to start with

Three metrics, each read by source

  1. Sign-upsThe CRM or the sign-up database, with the source of each userReadable from day one
  2. ActivationNeeds a link to the sourceProduct analytics, on the event your team choseReadable after the first week
  3. Trial-to-paidBilling or the CRM, in customers and in new MRRReadable once a full trial has ended

Sign-ups and paying customers usually sit in the CRM, next to the source of each user. Activation sits in the product, which rarely knows where a user came from. That makes it the step that takes the most work.

The activation event is also a decision your team has to agree on before reading anything. What counts as activation in PLG walks through how to choose it.

Why read each metric by source

Take September's sign-ups, read in mid-October once every 14-day trial has ended. A user counts as activated once they complete a first project within 7 days.

September's sign-ups by source, followed to the end of their trial. Illustrative data
SourceSign-upsActivatedPaying customersNew MRR
Paid search450171 (38%)45 (10%)$9,000
Organic search380198 (52%)68 (18%)$16,320
LinkedIn ads12040 (33%)10 (8%)$3,000
Referrals12078 (65%)34 (28%)$9,520
All sources1,070487 (46%)157 (15%)$37,840

Paid search brings 42% of the month's sign-ups and 24% of its new MRR, slightly less than referrals, which brought about a quarter as many sign-ups. Ad platforms learn to find the people most likely to sign up, who aren't always the ones who go on to use the product. A report that stopped at sign-ups would move next month's budget toward paid search.

Activation shows the gap within a week, long before the first payments. Trial-to-paid confirms it once the trials have ended.

Linking activation to the source

The source is recorded when a user signs up. Activation is recorded in the product, a few days later. To read one by the other, both records need a key they share.

Sign-ups with no recorded source stay in the view, labelled as such. If their share grows, a tag is missing somewhere upstream.

Without a product analytics tool, the product's own database already records what each account creates, connects or configures. An extract with the user ID is enough to read activation until the product sends clean events.

Reading trial-to-paid in customers and in revenue

Trial-to-paid is the share of trials that become paying customers. Read it only once every trial in the period has ended. With a 14-day trial, users who signed up on 30 September decide in mid-October. Read earlier, the rate looks lower than it is, and lowest for the sources that grew late in the month.

Count new MRR next to customers, because plans differ from one source to another. In the example, LinkedIn ads bring the fewest customers, on plans worth $300 a month on average against $200 for paid search. Per sign-up, they bring in more revenue than paid search. Whether that makes them the better channel depends on what each sign-up costs, the next metric to add.

What you can do alone, and where you'll need help

You can start alone, but the data belongs to several teams and comes with rules, so part of the work needs someone else.

You can do yourself

  • Choose the sources to compare
  • Propose an activation event
  • Connect the CRM, product analytics and billing once you have access
  • Read the view and compare sources

You'll need someone else for

  • Access to the CRM, product analytics and billing, often granted by an admin
  • Agreement on the activation event from the teams that use it
  • Saving the campaign tags and the user ID with each sign-up, if the form doesn't yet
  • Recording an event the product doesn't capture yet
  • The data rules: user records hold personal data, so bring in only the fields the view needs and limit who can see it

From a first view to a weekly one

A first view can come from a spreadsheet. It stops being enough once several teams report on the same funnel every week, and a few signs point to that moment:

A dedicated tool doesn't fix the data it receives. Campaigns still need consistent tags, and the activation event still needs the teams' agreement. What changes is that those rules are written once and applied to every report. In Funnelsight, the activation and retention dashboard follows the three metrics by source every week. Once spend is connected or imported, cost per paying customer comes next.

See sign-ups, activation and trial-to-paid by source in one view.

Start free trial