search

OpenSourceFellows / amplify

Open Source Fellow Sandbox
https://amplify-app-production.herokuapp.com/
MIT License
77 stars 59 forks source link

Create tracking to understand accessibility #405

Open manishapriya94 opened 1 year ago

manishapriya94 commented 1 year ago

Use case: we want to better accessibility and education by understanding which steps users are getting stuck at, ultimately how can we create a graph of where users drop off?

Today: we can estimate points of fall off based off of API calls in the user engagement accordion

Research prompt: Which analytics can we embed in frontend around clicks that haven't already been covered?

manishapriya94 commented 1 year ago

@paramsiddharth before assigning devs, lets work to research what it would take. @tried-again can help point out which screens need what tracking?

strawberrie68 commented 8 months ago

What if we implemented Google Analytics? That should tell us which page the user dropped off.

If we can't implement it, we could test it with Heat Maps. Heat Maps can help us see where users drop off or get lost. However, that does require some user testing.

Let me know what you think!

manishapriya94 commented 8 months ago

It seems like vue-analytics is set up well to work with Google analytics and we can discuss the above use cases and whichever fit for any custom events if needed?

Vue heatmapJS is also available

Sparky338 commented 1 month ago

Hey @DietBepis1. I'm curious if this would be able to fit into the control panel that you're working on here and if it's something that seems feasible. If so, do you have a preference for Google Analytics or Vue analytics?

DietBepis1 commented 4 weeks ago

@eunicey Adding you to this discussion based the expertise you mentioned in pair hours. Do you think Google Analytics will give us enough info about user drop-off and friction? Should we be thinking about other tools as well? I see @strawberrie68 also mentioned heatmaps.

eunicey commented 3 weeks ago

@DietBepis1

Google analytics (and most other tools) track URL changes out of the box. Google analytics will automatically measure where users drop-off based on url .

Clicks require some dev work. The preferred method is sending events through a data layer object or using google tag manager to track click events for buttons with custom IDs or classes.

The first thing we should discuss is whether the questions you want answered can be tracked this way.

  1. whether Lob displayed letter - Does the url differ depending on whether the letter fails or succeeds in loading?
  2. whether user put in input - I would recommend sending a custom event when users click the 'Next' button and all input fields are filled. OR conditionally enable the 'Next' button based on the fields and send an event when it's clicked.
  3. whether Stripe API was called - Is this accompanied by a url change?
  4. whether letter was created - Is this accompanied by a url change?