Skip to content
Academy
Marketing Academy · Field Work●Growth Marketing
All projects·From: Product-Led Growth
CoreAudit· 35 minutes

The First 5 Minutes: Auditing a Self-Serve Signup Funnel for Activation Gaps

Coinbase

Objective: Given a real-shaped funnel export (signups, verified, first action, PQL trigger), find exactly where the activation flywheel leaks and recommend the highest-leverage fix.

You're a growth analyst at Coinbase reviewing last month's retail signup cohort: 50,000 signups, but only a fraction ever complete the actions that predict they'll become active traders.

Use the lesson's activation-rate and PQL-trigger framework to read a funnel export, find the biggest leak, and size the fix.

Before you start

What you'll need

Free path (everything below is enough to finish)

Google Sheets
FreeImport and segment the funnel export

Free, no account needed, handles pivot-style segmentation easily

Google Analytics 4(optional)
FreeConfirm stage-to-stage drop-off matches the exported funnel

Free funnel exploration report validates the CSV against live event data

No access? Skip and trust the CSV export if GA4 isn't wired to the product

Paid upgrades (optional, faster/deeper)

Amplitude's free tier covers this exact exercise; the paid tier only matters once you're tracking millions of monthly events.

Amplitude(optional)
FreemiumBuild a persistent activation funnel dashboard with cohort retention curves

Purpose-built product-analytics tool for PQL scoring at scale

The process

2 steps

Step 01 of 02

Activation Rate

The lesson defines activation rate as the % of signups completing the action that predicts retention: 20-40% is average, 50%+ is excellent.

Of 50,000 signups, 41,000 verified identity but only 9,800 completed a first funded trade within 7 days. What's the activation rate, and is it healthy?

Google Sheets— Import funnel-export.csv, compute completed/signups per stage.

Procedure

  1. Import funnel-export.csv and freeze the header row
  2. Add a column dividing each stage's count by total signups
  3. Flag any stage below the lesson's 20% floor
Sample output
Stage counts (50,000 signups)
  Verified identity      41,000   82.0%
  Funded first trade      9,800   19.6%   <- activation rate
  Traded on 3+ days       6,100   12.2%

Healthy

Activation rate lands at or above 20%, with most drop-off between signup and identity verification, a known friction point smoothed by better UX copy.

Unhealthy

Activation sits at 19.6%, just under the lesson's 20% floor, and the steepest drop is AFTER verification: users who cleared identity checks still don't fund a trade.

What this means

The leak isn't awareness or trust (82% verify), it's translating a verified account into a completed action. That points at onboarding friction inside the app, not top-of-funnel messaging.

So what do I do about it?

SymptomActionEffort
Verified users stall before funding a first tradeShip a guided 'fund $10 and place your first trade' prompt inside onboardingdev ticket
No segmentation of where verified users drop offAdd stage-level event tracking between verification and first tradehalf day
YouYou can do this yourself, no engineering access required.

Step 02 of 02

Product-Qualified Leads (PQLs)

PQLs are free/self-serve users who hit usage signals predicting conversion; the lesson notes PQLs convert 3-5x better than MQLs.

Which of these three usage signals should trigger a PQL flag: traded once, traded on 3+ days, or invited a friend?

Google Sheets— Cross-reference the funnel export's behavior columns against the 7-day activity window.

Procedure

  1. Filter the export to users who traded on 3+ distinct days
  2. Compare their week-4 retention rate against the 'traded once' segment
  3. Flag the higher-retention segment as the PQL trigger
Sample output
Week-4 retention by segment
  Traded once only        14%
  Traded 3+ days          58%   <- PQL trigger
  Invited a friend        31%

Healthy

The team picks 'traded 3+ days' as the PQL trigger because it has by far the highest downstream retention, a real predictive signal.

Unhealthy

Picking 'invited a friend' as the PQL trigger because it feels like the most engaged action, even though it retains worse than the 3+ day trading signal.

What this means

A PQL trigger has to be chosen by what it predicts, not by which action sounds most impressive.

So what do I do about it?

SymptomActionEffort
Sales/lifecycle team has no PQL definition to work fromShip the '3+ trading days in 7' flag as the PQL trigger and route it to a lifecycle email30 min
YouYou can do this yourself, no engineering access required.

Final deliverable

A one-page activation audit memo: the measured activation rate, where the biggest leak sits in the funnel, the chosen PQL trigger with its supporting retention data, and one dev-ticket-sized fix.

See a reference example
Sample output
Wise, Q2 self-serve signup audit (excerpt)

Activation rate: 34% (above the 20% floor)
Biggest leak: 'first transfer' step, 22% drop after account funding
PQL trigger selected: '2+ transfers in 14 days' (61% week-4 retention vs 18% for 'transfer once')
Recommended fix: add a second-transfer nudge email at day 3

Success criteria

You're done when you can:

  • Correctly calculates activation rate from the raw stage counts
  • Identifies the funded-but-not-trading leak, not the identity-verification stage
  • Selects the PQL trigger with the highest supporting retention data, not the most 'active-sounding' behavior