Project Casper
Changing the way employers sponsor their jobs on Indeed.com
First, a little context
Indeed generates revenue by charging employers for “sponsoring” their job posts, which gives them greater visibility within relevant search results. It’s a cost-per-click model. The more you pay, the more “clicks” you get.
But when we interviewed them, over 80% of employers said their biggest issue with Indeed was the time they wasted on applicants without the necessary skills, or in the wrong location for the job.
More clicks means more eyes on your job, but that also means more resumes to sort through.
But what if we charged for results?
Our new pricing model would charge for applications. Because why pay for clicks if they don’t result in applications, interviews, or in hiring someone great?
The cost per application model. Or, internally, “Project Casper.”
We set ambitious goals, and we tried everything. Over a three-month period, we iterated constantly, and conducted 14 separate tests. See some of our preliminary designs below.
We launch, but conversion drops.
Employers don’t care about pricing models. They just want the applications they receive to meet the expectations that we set.
Our main value prop for employers was that they would only be charged for applications that met the conditions of the screener questions they marked as Deal Breakers. But they still thought they were missing out on
great candidates.
Were employers saving time, or was Indeed
withholding information?
If we aren’t meeting our clients’ expectations, let’s let
them decide.
We deployed a “reject and replace” functionality, which gives employers 48 hours to reject any applications that don’t align with their expectations. Rejected applications are replaced, free of charge.
Meanwhile, Indeed learns from employers’ preferences, so it can deliver candidates that align more closely with
employer expectations.
We introduce the concept on the sponsorship page.
And then reference the functionality throughout the experience.
Driving engagement through multiple touchpoints.
Once fully ramped, conversions increased by something like 4.5 percent.
My UX Director at Indeed