PLG onboarding analytics
Know how new users actually get to value.
The measurement surface PLG teams actually need: time-to-first-value, activation funnel drop-off, in-product feature adoption. Built around the question, not against the event stream.
47.3%
median activation rate (Q1 2026 cohort)
4.2 days
median time-to-first-value
63%
trial-to-paid conversion (top quartile)
Dashboard
Activation
Activation
47.3%
+3.2
TTFV
4.2d
-0.8
Trial-to-paid
63%
+1.1
Activation funnel
Recent cohorts
view all →Teams testing Threshold
fictional cohort
- Cohort Labs
- PRISM
- Hillcrest
- lattice.io
- Northbeam
- ATLAS
The gap
Onboarding metrics live in the cracks between general-purpose tools.
PLG teams use three categories of tool to measure activation today. Each is excellent at what it was built for. None was built to answer the question PLG teams actually ask.
Mixpanel and Amplitude
Built for the event stream. Onboarding analysis means writing custom queries against a general-purpose schema, every time. The PM asks a question; engineering writes SQL; the answer arrives a week later. Activation iteration takes longer than activation itself.
Pendo and Userpilot
Built for in-product onboarding execution: tooltips, tours, checklists. Excellent at running the experience. Measurement is downstream of execution; analyzing what actually moves activation requires a separate tool. The team ends up running both.
Custom dashboards in BigQuery
Engineering owns the dashboard. PMs file tickets to ask new questions. Iteration is slow; the dashboard goes stale; institutional knowledge calcifies in saved queries. The activation surface stops being a product surface and becomes a data surface.
What Threshold does
A measurement surface built around the activation question.
01
Pre-built activation funnels
Define the steps that matter, not just events. Threshold tracks them as a funnel from day one and surfaces drop-off where it actually happens.
Sample funnel
02
Time-to-first-value benchmarking
Compare against your cohort, not against an arbitrary baseline. The benchmark adapts as your data grows; PMs see TTFV trending in real time.
Median TTFV by week
03
In-product onboarding measurement
Pair with your existing onboarding tooling (Pendo, Userpilot, in-house). Threshold measures what actually moves activation; the execution layer stays where it lives.
04
PM-owned dashboards
Built for product managers, not data engineers. SQL access for engineers; opinionated UI for everyone else. New questions take minutes, not sprints.
Comparison
Focused, not just smaller.
Threshold builds dashboards around the activation question. General-purpose analytics tools build dashboards against the event stream and leave the question to you. In-product onboarding tools build the experience and treat measurement as downstream.
| Capability | Threshold | General-purpose analytics (Mixpanel, Amplitude, Heap) | In-product onboarding (Pendo, Userpilot) |
|---|---|---|---|
| Pre-built activation funnel | ✓ | · | · |
| Time-to-first-value benchmark | ✓ | · | · |
| Onboarding drop-off detection | ✓ | · | partial |
| In-product onboarding execution | · | · | ✓ |
| Custom event tracking | ✓ | ✓ | ✓ |
| Cohort comparison | ✓ | ✓ | · |
| SQL access | ✓ | ✓ | · |
| PM-friendly UI | ✓ | · | ✓ |
Comparison reflects vendor positioning, not feature exhaustiveness. Each tool is excellent at what it's built for; Threshold is built for measuring what happens between sign-up and first paid usage, which sits between the existing categories.
How it works
Three primitives. One question.
01
Connect
Pipe your existing event stream (Segment, Rudderstack, custom). No SDK required if you already track. New stack? Twelve-line client, four major platforms.
02
Define
Tell Threshold which events mark each onboarding step. The funnel builds itself; activation criteria stay editable as your understanding sharpens.
03
Measure
Time-to-first-value, drop-off, cohort benchmarks ship pre-built. Customize as you grow. New questions take minutes, not engineering tickets.
Early access
Join the waitlist.
Three quick questions help us match you with the right early-access cohort. We'll reach out when access opens for your profile.
- No sales calls until you're ready
- Beta cohorts grouped by team size and stack
- Founding team prices for the first 50 customers
What's your role?
Questions
Things people usually ask.
Why not just use Mixpanel or Amplitude?
Mixpanel and Amplitude are excellent general-purpose analytics. They track events well. For onboarding analysis specifically, you write custom queries against the event stream every time. Threshold pre-builds the activation funnel, time-to-first-value benchmarks, and drop-off detection so PMs can iterate without engineering.
What's the integration like?
If you already pipe events to Segment or Rudderstack, point those at Threshold's ingestion endpoint and you're done. If you don't, we have a 12-line SDK for the major platforms (JavaScript, Python, Ruby, Go). Most teams are productive within an afternoon.
What scale does Threshold work at?
Threshold is purpose-built for the 50K to 5M monthly user range. Below that, you don't need dedicated PLG analytics; spreadsheets and Mixpanel free tier handle it. Above that, Threshold still works but enterprise data warehouses become a better foundation.
What about privacy and data residency?
EU and US data residency available at launch. SOC 2 Type II in flight; expected before general availability. Threshold doesn't store PII beyond what you send; user-level identification is via your existing user IDs.
How is pricing structured?
Pricing isn't published yet. Founding-team customers (the first 50) get founding pricing for the life of their account. Beyond that, expect tiered seat-and-usage pricing competitive with category incumbents.
Is there a self-serve free tier?
Plans haven't fully shipped, but yes: a free tier for solo PMs and pre-seed teams is on the roadmap. Larger teams can join the waitlist for early-access pricing.