Repost from "Jobs Best Practices & Assisting Employers"
Handshake Community -
Thank you for all of your thoughtful feedback about the new job search experiments that we launched recently. We wanted to take some time to provide more context and address your feedback with next steps.
Over the past 6 months, our product teams have been launching new features, backed by data from experiments, to improve the job search experience for students - these experiments have yielded over 10% more applications weekly in aggregate. Most recently, we’ve experimented with the job search page and filters.
Why? In our regular user research sessions with students, we consistently heard a common theme: the job search experience - the actual use of search filters, keyword search, etc. - could be much simpler, easier to use and more intuitive.
Here are some of the quotes from students.
- “Hard to use, filters are difficult to set, doesn't fit needs”
- “Needs a better user interface, especially when choosing filters”
- “One thing I'd like for Handshake to improve in would be the structure of selecting filters for the job search.”
- “The filters aren't very good when you search and it's really not easy to find jobs that are good for me without sifting through a lot of information.”
So we aimed to simplify this experience, to make it easier for students to find the most helpful job filters without overwhelming them with too much choice. This is a key design paradigm used across sites from job search to travel search and beyond - show the most helpful and widely used features, and then group other low-volume filters in an “advanced” filter area.
To figure out which filters to prioritize, we:
- Analyzed which search filters were best at helping students find jobs by looking at the data. How often are certain filters used? Which filters are most helpful in driving students to apply to jobs?
- Experimented with reducing certain filters which were less effective and/or not used.
- Regrouped and restructured remaining filters into more intuitive categories, based on data and user feedback.
As a concrete example, we found that the job type filter (full-time/part-time/on-campus), had been used in over 50% of successful searches, or those leading to application. To help more students successfully search for jobs, we decided to promote this filter to its current featured location below the keyword box.
On the job search page we also introduced Suggested Keyword Searches that are derived from each student's major, as well as from the inventory of jobs. We found that searches generated using suggested keywords led to applications at twice the rate of other searches.
Importantly, results have shown a 2-3% lift in applicants indicating that this new design results in more students applying to jobs.
That said, we have heard your feedback, particularly about filters for job function, major and employer preferences (major, school year, etc.) and want to share an update given our data-driven learnings around this experiment.
The job function and major filters are a good example of how our product experimentation works. Our original data analysis showed that very few students used either of these filters to find jobs. In the spirit of simplifying, we experimented with excluding these as an option. However, the data and feedback from that experiment has shown that, while few students used these filters, those who did were much more likely to apply to a job, meaning that each filter provided very strong signal. Over the next two weeks, we will be re-adding job function and major filtering back to the job search UI.
Regarding employer preferences filters, our goal is to provide students the ability to quickly filter jobs that meet employer preferences which comprise multiple fields (major, GPA, work auth, school year). Over the next few weeks, students will also be able to filter on each individual preference, should they choose to.
Thank you for your continued feedback. We appreciate your partnership in thinking through innovative ways to better help students find jobs on Handshake!
- Handshake Team
Please sign in to leave a comment.