Talent Radar
XING - 2019
Problem
Our users could perform complex candidate searches in XING TalentManager, but they were asking to know more about their potential candidates. To do this, they had to visit candidate profiles one by one, which was a time-consuming task. It was clear we needed a quick way of visualizing more insights about candidates.
Goal
Offer a visualization of the candidate market. Help our users to optimize their recruiting strategies. Allow active scouring of recruiters to explore specific job titles in their desired locations, such as cities, countries, or worldwide.
Insights and visualization research
One of my initial tasks I prepared was a quick scenario for a remote and unmoderated user test, mostly with heavy active sourcing and internal recruiters. Our goal with this was to gain some generic first impressions and also learn how interesting the list of insights would be for recruiters.
I had to tweak most of the visualizations, not because the insights weren’t interesting for recruiters, but rather because of reading difficulties or because it took too much time to understand the meaning. In general, the most interesting insights concerned location, such as top locations, radius, or candidates who are relocating to other cities, amongst others.
During the second research phase, users generally understood that the page’s purpose was to gain insights on candidates, and improve recruiting tasks by saving time and effort in contacting the right candidates. With the results on hand, we knew that we’ll have to work better on writing, item locations, hierarchy, and visuals.
Above: Some of the iterations and new proposals we made during the second phase of user tests. This time I prepared a small prototype where our users could see the entire Talent Radar dashboard with all the widgets.
Input fields
I designed two input fields that would auto-suggest results for our users. They can enter cities, countries, and job titles. To search for all job titles, city, or country, I added the entries: All job titles and Worldwide as fixed footers in the auto-suggest dropdowns.
One type of functionality I was curious to test with our users was which parts from the suggestions were more helpful to highlight: the matching characters (that the user entered) or the different characters that complete the result inside the dropdown? Surprisingly, the most helpful way was to highlight the different characters. This helped users read different results in the list more quickly.
Recent discoveries
We tried different solutions to allow our users to select or run recent searches they made. Since the search was always composed of two terms: job title and location, we could either try single recent searches (per input field), add the entire query in both inputs (job title and location), or place the entry point of those just outside of the input fields. The final decision was to place them outside, making it more accessible for our users — especially because the fields were already too technically complex to use.
