Marketing Attribution Sankey Diagram: Editable Template
Trace visits from acquisition channel to landing page to signup to paid customer, and see which channels and pages carry the volume and which convert. Replace the counts.
An attribution Sankey shows how visits from each channel spread across landing pages and how many of them end up as signups and paying customers. It is a view of where traffic goes, not a model of which channel deserves the credit.
Four channels send 25,000 visits to three landing pages. Each page turns a share of its visitors into signups, and signups split into paid and free. The page, not the channel, is where the signup counts are taken.
That is a real limitation. Once visits merge at a landing page, the diagram no longer says which channel a signup came from. To keep channel identity to the end you would need one chain per channel, or a separate chart. The counts here are invented.
How to read this diagram
- 25,000 visits produce 2,169 signups (8.7%) and 260 paying customers.
- The pricing page converts 15% of its visitors to signup, the home page 10% and the blog 4%.
- The blog receives 46% of all visits but produces 21% of signups.
- Paid search sends two thirds of its visits straight to the pricing page, which is why it looks efficient at this stage.
The data behind it
Invented visit counts for one period. Not data about any real site.
| From | To | Amount (visits) |
|---|---|---|
| Organic search | Home page | 3000 |
| Organic search | Pricing page | 2000 |
| Organic search | Blog | 7000 |
| Paid search | Home page | 1000 |
| Paid search | Pricing page | 4000 |
| Paid search | Blog | 1000 |
| Social | Home page | 1500 |
| Social | Pricing page | 500 |
| Social | Blog | 3000 |
| Referral | Home page | 600 |
| Referral | Pricing page | 800 |
| Referral | Blog | 600 |
| Home page | Signed up | 610 |
| Home page | Left without signing up | 5490 |
| Pricing page | Signed up | 1095 |
| Pricing page | Left without signing up | 6205 |
| Blog | Signed up | 464 |
| Blog | Left without signing up | 11136 |
| Signed up | Became paying | 260 |
| Signed up | Stayed free | 1909 |
The same data as CSV
source,target,value
Organic search,Home page,3000
Organic search,Pricing page,2000
Organic search,Blog,7000
Paid search,Home page,1000
Paid search,Pricing page,4000
Paid search,Blog,1000
Social,Home page,1500
Social,Pricing page,500
Social,Blog,3000
Referral,Home page,600
Referral,Pricing page,800
Referral,Blog,600
Home page,Signed up,610
Home page,Left without signing up,5490
Pricing page,Signed up,1095
Pricing page,Left without signing up,6205
Blog,Signed up,464
Blog,Left without signing up,11136
Signed up,Became paying,260
Signed up,Stayed free,1909How it works
Choose the channels and pages
Group sources into four to six channels and landing pages into a few types.
Count visits from each channel to each page
Use the first page of each session and the session's source.
Count signups and payers
Add the signup and paying counts for each landing page.
Check and annotate
Check balances, and state on the chart that it is landing-page based and not multi-touch.
Frequently asked questions
Is this multi-touch attribution?
No. It shows one path per visit, from channel to landing page to outcome. Multi-touch models need data on every touch and their own rules for sharing credit.
Can I keep the channel all the way to the end?
Only by drawing a separate chain for each channel, which gets wide quickly. Many teams chart one or two channels of interest at that level.
Edit this attribution Sankey
Open the example as your own chart and enter your channel, page and signup counts.