Home
Today (BTDY)
Top Stories (TOP)
News Read Story (NSN)
My News (MYN) Equities (WEI)
Economic Events (ECO) Worksheets
[+ any other widgets on home which has a "More >" action]
Search
IB
Start/Find Chat
Read Chat Start/Find/Read (iPad mode uses split view so user sees both things together)
New Chat
New Blast
Settings
What's new tour
View Chat Details
Participants
Forward Post
Reply to Post
Alerts
People
Front Page
Securities
Research
TV
Radio
TW
FILE
NOTE
For basic graphs we could use https://developers.google.com/chart/interactive/docs/gallery but it can be fiddly to get exactly what you want. if the examples are good enough and you just want a different dataset it is easy though.
Google Analytics https://www.orbitmedia.com/blog/user-flow-google-analytics/ https://digitalcommunications.wp.st-andrews.ac.uk/2016/11/15/using-the-users-flow-report-in-google-analytics/ https://yoast.com/users-flow-in-google-analytics/
This seems a clone of GA https://mixpanel.com/blog/user-flow-analysis/
Good example showing how loops are represented in GA User Flow analytics https://digitalcommunications.wp.st-andrews.ac.uk/files/2016/10/users-flow-step1.png
Note how "/" and "/index.html" appears in multiple columns
Note also the (>100 more pages) bucketsI think the way this works is user picks a start page (in our case Activity), or a session attribute (e.g. country), and then we group every session based on first activity, second, third, etc. such that we have a tree (non a graph - there is no recombining paths) where activities are represented as nodes (and can appear repeated at multiple levels) and the tree depth corresponds to first activity, second, third, etc.
Chart style is called Sankey https://d3-graph-gallery.com/sankey.htm https://developers.google.com/chart/interactive/docs/gallery/sankeyl
based on https://digitalcommunications.wp.st-andrews.ac.uk/files/2016/10/users-flow-step1.png
See also https://mixpanel.com/blog/user-flow-analysis/ (Starting=Open Message, Ending=Any Event, Users In 2 Cohorts, Time Last 30 days)
https://www.w3schools.com/graphics/svg_grad_linear.asp https://developer.mozilla.org/en-US/docs/Web/SVG/Tutorial/Gradients https://developer.mozilla.org/en-US/docs/Web/SVG/Tutorial/Paths
for each stage "1st interaction,..."
X sessions, y drop-offs
for each activity, ranked by count
add catch all bucket (other, 272,332)
height of box proportional to count (compute scaling factor based on total sessions / imageHeight)
make whitespace constant between each
if text fits in box draw there, else in whitespace
out flows, multiple, one per connection, with proportional
red gradient drop off, width proportional to drop off rate
Interactivity
Filters on sessions - time, device type, device, UUID, cohort,