Sequel
Product Analytics

Signup funnel drop-off analysis

Find the step where the most users abandon signup or onboarding, and how that varies by segment.

Install into your agent
  • Where do users drop off during signup?
  • How many people who signed up finished onboarding?
  • Which step in the funnel is leaking?

The playbook

Use whichever product analytics source is connected — the shape of the analysis is the same, only the call differs. Do not query all three and merge; they define sessions and users differently and the numbers will not reconcile.

Steps

  1. Establish the real step sequence. Don't assume it. List the available events first:

    • PostHog: posthog.list_events
    • Mixpanel: mixpanel.list_events
    • Amplitude: amplitude.list_events Confirm the ordered step names with the user if they're ambiguous. A funnel built on guessed event names produces confidently wrong output.

  2. Run the funnel.

    • Amplitude: amplitude.funnel_report with the ordered events array and an explicit conversion_window_seconds.
    • Mixpanel: mixpanel.list_funnels first, then mixpanel.funnel_report with the funnel_id. Mixpanel funnels are saved objects — you generally cannot define an ad-hoc one here.
    • PostHog: posthog.query with a HogQL query over the events table, or posthog.list_insights to find an existing saved funnel.

  3. Compute step-to-step conversion, not just overall. The largest single drop between adjacent steps is the finding. Overall conversion tells you there's a problem; step-to-step tells you where.

  4. Segment the worst step. Re-run split by a meaningful dimension (group_by on Amplitude, on for Mixpanel, a GROUP BY in HogQL) — device, plan, signup source, or country. Drop-off is rarely uniform, and a step that's fine on desktop but collapses on mobile is a much sharper finding than the average.

  5. Report. A step-by-step table: step name, users entering, users continuing, step conversion rate, drop-off count. Call out the worst step, quantify what fixing it to parity would be worth in absolute users, and note the segment where it's worst.

Gotchas

  • The conversion window drives the result. A 1-hour window and a 7-day window on the same funnel give very different answers. State the window you used; if the product has a long consideration cycle, a short window manufactures drop-off that isn't real.
  • Funnels are ordered. Users completing steps out of order may be excluded entirely depending on the tool's mode. Check whether the funnel is strict or any-order before interpreting.
  • Instrumentation gaps look exactly like user drop-off. If a step shows near-total loss, verify the event is actually firing before concluding the UI is broken.
  • mixpanel.export_events is expensive — it streams raw events. Use mixpanel.segmentation or a saved funnel unless raw rows are genuinely required.
  • Don't compare funnel numbers across tools. If two are connected, pick one, and say which.

Install

Install Sequel skills into your agent

Pick your app and Sequel installs itself: a one-click link where the client supports it, that client’s own few steps where it doesn’t. The playbooks come with the server: your agent sees that this playbook exists and reads it when a question matches.

Claude, ChatGPT, Cursor, VS Code, Claude Code, Codex, Grok, Muse, dots and any other MCP app

Already have an MCP client?

https://api.sequel.sh/mcp

Add it as a remote (streamable HTTP) server and sign in when prompted. Skills come with it; nothing else to install. Manual setup per client

Choose your app

  • Claude

    One click, opens in the app

  • Cursor

    One click, opens in the app

  • VS Code

    One click, opens in the app

  • ChatGPT

    Turn on developer mode, then add Sequel. About two minutes.

  • Claude Code

    Run one command, then sign in with /mcp.

  • Codex

    Run one command, then sign in.

  • Gemini CLI

    Run one command, then sign in with /mcp auth.

  • Grok

    Add Sequel as a connector and sign in.

  • Grok Bot

    Paste this to the bot. It registers Sequel and asks you to sign in.

  • Muse

    Muse writes its own connector. One message is the whole setup.

  • dots

    Add Sequel in ChatGPT; your dot picks it up with no extra step.

  • OpenCode

    Add Sequel as a remote server from the OpenCode CLI.

  • Other MCP app

    Any client that speaks remote MCP can connect with this URL.

More like this

Other Product Analytics skills

Can we trust our product metrics

Two tools show different numbers for the same metric and nobody knows which to trust. Reconciles the disputed figure against a system of record such as Stripe or the application database, then finds the instrumentation defect behind it: duplicate and near-duplicate event names, events that silently stopped firing, missing properties and system-fired events counted as user actions.

Feature launch impact readout

Measure whether a feature we just shipped actually landed: adoption against an eligible denominator of users who could even reach it, breadth, depth and sustained use after the release, a pre and post impact estimate with or without a control group, and a keep, iterate or sunset call.

Find our activation moment

Discover the aha moment: which early in-product action, taken in the first session or first week, most separates users who stay from users who vanish, how to score candidate first actions by retention lift, and what share of new signups ever reach it. Covers activation metric definition, activation rate and time to value.

Is our retention curve flattening

Do users keep coming back, or do they drift away after the first week or month? Builds the cohort retention curve for your key action at its natural frequency and judges whether it flattens to a plateau or decays to zero, covering N-day versus unbounded retention, repeat usage, leaky bucket diagnosis and how the plateau compares to published benchmarks.

Read out an A/B test honestly

Is this A/B test result real, or statistically significant only by accident? Runs the validity gates before reporting anything (sample ratio mismatch, peeking, power and minimum detectable effect, novelty), then gives absolute and relative lift with a confidence interval, a ship or kill call, and the winning variant extended through to revenue instead of stopping at the proxy metric.

FAQ

Frequently asked questions

What is signup funnel drop-off analysis?
Identifying which steps in your signup or conversion flow lose the most users and why. For each step, you measure how many users entered and how many proceeded to the next, then investigate the largest drop-offs with session data to find whether the cause is friction, confusion, or a technical issue.
How do I set up a funnel in GA4?
Use Explorations in GA4 and select the Funnel exploration template. Define each step by the event or page view that represents it. GA4 shows open and closed funnels: closed only counts users who completed every step in order; open counts anyone who reached that step.
What is a normal signup conversion rate?
It varies enormously by product, audience, and funnel length. A freemium B2C tool might convert 40% of visitors who reach the signup page. A complex B2B demo request form might convert 5%. Benchmark against your own historical data before comparing to industry averages.
How do I investigate why users drop off at a specific step?
Watch session replays for users who dropped at that step. Check for form errors in GA4 error events. Check page load time for the step. Look at the form fields required: each additional required field reduces completion rates. Test the step on mobile yourself.
Can an AI agent analyze signup funnel drop-off?
Yes. With GA4 connected, the agent defines the funnel steps, computes drop-off rates at each, and returns the biggest friction points with hypotheses for each. Sequel provides the GA4 connection over MCP and the signup-funnel-drop-off playbook.

Put this playbook to work

Connect a source, ask the question, and the agent follows these steps. Free to start.

Get started freeAll skills