Spindl
  • 👋Overview
    • Introduction
    • Attribution
    • Privacy
  • Your Spindl app setup
    • Managing team members
  • 🪄Features
    • 📈Onchain Attribution
      • 📉Plotting attribution
    • 📊Web3-native Analytics
      • Chart Builder
      • Event Selection and Filtering
      • Line Chart
      • Funnel
      • Cohort Retention
      • Sankey Diagram
      • Big Numbers
      • Pies and Donuts
    • 🧍‍♂️Audiences
      • Creating an audience
    • 🔗Short Links
      • 🖇️Custom Domains
    • 🤝Referrals
      • Management and reporting
      • ⏩Quick Start
      • ⛓️Technical Details
  • ⚙️Techncial
    • ⏩Start Here
    • Google GTM Guide
    • Javascript SDK / HTML Script Guide
      • ➡️Install
      • ➡️Wallet Connects
      • ➡️Page Views
      • ➡️Custom Events
      • ✅Verify That Events Are Sent Correctly
      • React, Next.js & Html Examples
      • Setup a Reverse Proxy
        • Using Next.js
        • Using Cloudflare Workers
        • Netlify
      • ➡️Referrals
      • Default vs Lite SDK Versions
      • Security
    • API
      • ➡️Short Links
      • ➡️Custom Events API
      • Data Exports
    • Android SDK
    • iOS (Swift) SDK
    • On-Chain
      • ➡️Rewards
  • Contact Spindl
  • Ads
    • ⏩Start Here
    • 📢Creative Specifications
      • Web Banner Ads
      • Discord Embeds
Powered by GitBook
On this page
  1. Features
  2. Web3-native Analytics

Sankey Diagram

PreviousCohort RetentionNextBig Numbers

Last updated 1 year ago

Arguably an even more complex chart type is , named after an Irish rail engineer plotting flows in a steam engine (though preceded by , which demonstrated the horrifying casualties suffered by Napoleon's army as it marched to and from Moscow). This one is a bit of a brain teaser, and the Spindl version is a pretty basic rendering of the full power of a Sankey, but it can be helpful to visualize certain flows.

Here's a Sankey showing the Page View entry points for a dex, filtered to traffic from Twitter (pretty high wallet connect rate!):

If you want to see what swaps get traded before users drop off, here's one view:

The configuration pane can be a bit confusing, and the output almost random-seeming (a lot of it depends on layout parameters, like depth of journey, that we don't expose yet).

Some definitions:

  • Event - The primary event to plot and focus on, though if users hit other events, those will appear too. If you filter, the entire chart will only be limited to events of that filtered type.

  • Direction - 'Entering event' means that generally you're looking at a plot of many origin points going to one final endpoint. 'Leaving event' means you're looking at the focus event as the source of the flow, and seeing where users go after that. If you reverse it in your head a lot, join the club; so do we.

  • Expanded Events - Here you can force events to expand into different sources (or sinks) based on whatever filter dimension you choose (e.g. session channel, AKA, where they came from).

It's hard to get a hang of it at first; play around the params until it sort of makes sense.

🪄
📊
the Sankey diagram
the much more famous Minard map