Paysa Profile Builder

Paysa (formerly CompGenome) is a compensation prediction product that helps high tech employees discover and achieve their "Market Salary." Paysa takes a user's entire employment history into account when making predictions, including education, current and former positions, companies, location, and skills. Paysa users can in turn use this information to determine if they are being compensated fairly, if they are worth more or less than they're currently being paid, to discover new opportunities that match their predicted Market Salary, and to research top and trending companies and available jobs.

1. I took Paysa’s Profile Builder through almost a dozen iterations, as it was the major gating (but crucial) step to obtaining a Market Salary. We needed users to enter a minimum number of items, including at least one job and or education (school), a location, and a few skills. Ideally, we needed users to enter their entire employment history, all of their degrees, skills, etc. This process could be tedious depending on how much experience the user had so we enabled them to import data from either Facebook, Google, or LinkedIn. The challenge here was that LinkedIn had a very low limit on API calls for this data, so low that we in fact could not use the direct import feature and had to rely on a manual user download of the PDF “LinkedIn resume.” Further complicating this process was the fact that we could not currently merge data once the user had chosen to import from a specific site. So for example, once a user chose to import from Facebook, s/he could not later import data from LinkedIn. The flowchart below shows the logic for this behavior.


2. Per the flow above, the first three mocks below show the process where the user chooses to import from LinkedIn but then skips the import process. This makes the Import from Facebook/Google buttons available. Early versions of the Profile Builder were straight text fields, either all open or some collapsed, with no particular order in which to complete the fields. After testing, I discovered that people felt much more comfortable following ordered steps, even if there was no requirement for them to proceed in order. This design shows a vertical stepper (wizard) that enabled users to return to steps they’d already completed but not proceed until the current step is completed. In addition, we added a “Confidence” score, with the explanation that the more information the user entered, the more confidently or accurately Paysa could predict Market Salary. I used Material Design visuals to speed up implementation for engineering.