Normalize email addresses for Google Enhanced Conversions

[tcgilbert]

Rocket Mortgage is experiencing an issue with low match rates for Google Enhanced Conversions. Although the data is being sent successfully, the match rate is below 1%.

During troubleshooting, we identified that email addresses are not being normalized according to Google’s specifications when using the endpoint. While this discrepancy is unlikely to fully explain the low match rate, it is an issue worth addressing since normalization is outlined as a requirement in both our documentation and Google’s documentation.

Testing

hashing currently

  input: '[email protected]',
  hash: '87924606b4131a8aceeeae8868531fbb9712aaa07a5d3a756b26ce0f5d6ca674'


  input: '[email protected]',
  hash: '60b7750a8ee074c6a8b1a2e8178e4da4c5cffc5309b5236a12355b7b21a985d6'

  input: '[email protected]'
  hash: '06bfc6aa38674253530e62f2b585d63e3786cbb759b81b73df34ae80894d8813'

  input: '[email protected]'
  hash: 'f60f8555710483f6d082157013ca06779173aa5454a6f0255dcf2258f007a834'

hashing with this PR

  input: '[email protected]',
  hash: '87924606b4131a8aceeeae8868531fbb9712aaa07a5d3a756b26ce0f5d6ca674'

Stage Test

  // converts to lowercase before hashing
  input: '[email protected]',
  hash: '87924606b4131a8aceeeae8868531fbb9712aaa07a5d3a756b26ce0f5d6ca674'

Stage Test

  input: '[email protected]'
  hash: '06bfc6aa38674253530e62f2b585d63e3786cbb759b81b73df34ae80894d8813'

Stage Test

  // removes period before @ for google domains
  input: '[email protected]'
  hash: '06bfc6aa38674253530e62f2b585d63e3786cbb759b81b73df34ae80894d8813'

Stage Test

• Added unit tests for new functionality

Note

There is currently a failing test for userLists. I didn't touch any of the functionality here, as this is also failing in main.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 78.41%. Comparing base (1225556) to head (58e9ff5).
Report is 15 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2668   +/-   ##
=======================================
  Coverage   78.40%   78.41%           
=======================================
  Files        1032     1032           
  Lines       18564    18570    +6     
  Branches     3517     3519    +2     
=======================================
+ Hits        14555    14561    +6     
  Misses       2833     2833           
  Partials     1176     1176           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[joe-ayoub-segment]

Hi @rvadera12 PR deployed.
Are you OK verifying this in Production please (as Thomas has left the company)?

[rvadera12]

Hi @rvadera12 PR deployed. Are you OK verifying this in Production please (as Thomas has left the company)?

Everything is working as expected in production