Dataford
Interview QuestionsInterview GuidesExperiencesMock InterviewsPricing
Get started

Evaluate SMB Onboarding Activation Lift

EasyMetrics00:00
Practice interviewer
In session
5 left
00:00

Your question is Evaluate SMB Onboarding Activation Lift. Take a moment with it on the right.

Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).

You need to log in / sign up to chat or submit.

Problem

Business Context

LedgerFlow is a SaaS bookkeeping platform for small businesses. The product team launched a new onboarding flow for SMB signups and wants to know whether it is improving activation without hurting downstream retention or paid conversion.

Metric Scenario

Before the launch, LedgerFlow averaged 12,000 new SMB signups per month. Under the old flow, 68% completed account setup, 52% connected a bank account, 41% invited at least one teammate, and 34% generated their first invoice within 14 days. The team currently uses "generated first invoice within 14 days" as the working definition of activation. After rolling out the new flow to all new SMB users last month, the same funnel is 74% setup completion, 58% bank connection, 39% teammate invite, and 37% first invoice within 14 days. Early 30-day retention moved from 28% to 27%, and trial-to-paid conversion is still pending for most of the cohort.

Stakeholders are asking whether the new onboarding flow is truly better, which step is driving the change, and whether the company should keep, iterate, or roll back the experience.

Requirements

  1. Define the primary activation metric precisely for SMBs, including numerator, denominator, time window, and edge cases.
  2. Explain how you would evaluate whether the new onboarding flow improved activation versus just shifting users between funnel steps.
  3. Decompose the activation change by key dimensions and identify the most likely drivers.
  4. Recommend guardrails and related metrics to ensure the onboarding flow does not improve activation at the expense of retention, monetization, or user quality.
  5. State what additional analysis you would do if the rollout was not randomized.

Data Available

  • signup_events: account_id, signup_date, company_size, acquisition_channel, country, device_type
  • onboarding_events: account_id, step_name, step_completed_at, flow_version
  • product_actions: account_id, event_name, event_time, invoice_id, bank_connected_flag, teammate_invited_flag
  • subscriptions: account_id, trial_start_date, paid_start_date, plan_type, cancellation_date
  • account_health_daily: account_id, active_users_7d, invoices_sent_30d, retention_day_30_flag