Founded in 2010, ZipRecruiter is the fastest-growing job search site in America. Every month, some 25 million job seekers come to ZipRecruiter.com to search and apply to the site’s more than 9 million active job postings. The company leverages Artificial Intelligence (AI)-powered smart matching technology to connect these employers and job seekers. In fact, 80% of employers who post a job on ZipRecruiter get a quality candidate within the first day.
The company has been a Mixpanel customer since 2015.
Before adopting Mixpanel, ZipRecruiter relied solely on a SQL-based tool, but they wanted a second, fast-and-easy analytics solution for the product team. The complexity of a two-sided marketplace means that ZipRecruiter is a data-heavy culture. “We test almost everything we launch, and we constantly challenge our assumptions,” says Akio Bandle, Senior Product Manager. The company required a tool that could quickly clarify and streamline the learnings from A/B testing.
Without Mixpanel, I'd be pretty bogged down and frustrated, says Bandle. I'd be spending a lot more time writing and processing SQL queries. I'm able to move as quickly as I do because of Mixpanel. Mixpanel makes analytics fun. I spend less time trying to figure out how to get the right data, and more time looking at it.Akio BandleSenior Product Manager at Ziprecruiter
“As a long-time Mixpanel customer at previous companies, I recommended the service to ZipRecruiter,” says Bandle. “The mobile team was the first to really embrace using Mixpanel; once other product managers saw us using it with great success, the rest of the product teams followed suit.”
Mixpanel is heavily used by all consumer-facing product and engineering teams at ZipRecruiter, with more than 25 people across the organization using it monthly. The Job Seeker product team often uses Mixpanel for quick ad hoc analysis, A/B test tracking, and dashboarding. “How likely are users to click on a job when they use the salary filter? Which variant drives a higher apply rate? For users who completed the onboarding flow, how likely are they going to be satisfied with their job matches versus those who don’t? These are the types of questions we’ve used Mixpanel to help us answer,” says Bandle.
“As soon as I launch an A/B test, I start building out my Mixpanel dashboard. I love how easy and seamless it is to consolidate charts and funnels into a clean-looking dashboard and share it with my team. Even for the largest projects we launch, it doesn’t often take me more than 20 minutes to set up a dashboard. If I want to dig deeper, tools like Cohorts are fantastic for segmenting my events by different user properties.”
“We’ve made so many exciting strides with the job-seeker product over the past year,” says Bandle. “We completely redesigned our signed-in homepage experience, offering users a more customized view of job matches. On top of that, our new post-apply experience identifies weak areas in a job seeker’s profile, and prompts the user to fill out more information for the employer. We also completely revamped our search experience to better assist users with their search results. Mixpanel was used every step of the way to help launch these features successfully.”
In terms of mobile, the company initially ran events through Mixpanel’s SDK but, in an effort to better consolidate the event pipeline, they are now running a server-to-server integration. The result: the same events can now be accessed between Mixpanel and their SQL tool, providing great flexibility.
“Without Mixpanel, I’d be pretty bogged down and frustrated,” says Bandle. “I’d be spending a lot more time writing and processing SQL queries. I’m able to move as quickly as I do because of Mixpanel. Mixpanel makes analytics fun. I spend less time trying to figure out how to get the right data, and more time looking at it.”
- Cohorts. “I’ll frequently use Cohorts to look at users who reply to one of our customer-satisfaction surveys and get bucketed into an A/B test,” says Bandle. “For instance, I can send an email survey and see if users who request ‘Add skills’ in a new post-apply flow are more likely to say they are satisfied with their matches than those who don’t make that request. I can use that as an extra data point when evaluating the effectiveness of a new feature.” Bandle also notes that the Cohorts feature is a great way to segment data in Insights and Funnels.
- Funnels. “We just launched a feature that guides users to job matches that increase their chances of getting a Thumbs Up from an employer. Seeing how we were able to dramatically shift applies between Test and Control was really gratifying. By using Funnels and making consistent optimizations we were able to improve the conversions for users getting matched to jobs.”
- Insights. “Our CSAT (customer satisfaction) email is our most basic and most useful report–it gives us an excellent high-level overview of job-seeker user experience across different dimensions. We’re able to use the data from this survey to benchmark our performance with user experience overall in Insights and identify specific pain points, and to cross-reference with other data points, such as A/B tests.”
- Flows. “Completed applications is an important engagement metric for us. We find that desktop web users have a much easier time completing an application than mobile web or mobile app users. The reason is simple: uploading a resume is easier from a laptop than from an iPhone, and so is filling out long, question-riddled ATS forms. As more users rely on their phones for their job-search needs, this will only become a bigger area of focus for us. We will look to Mixpanel to help us make this the most efficient and seamless process for job seekers.
Mixpanel best practices
Bandle has some last words of advice for companies regarding data collection. “First, before you do anything else, establish your event/property framework. Ensure that all your stakeholders–both in web and mobile–are aligned on how to name those events and properties. Finally, make certain that everyone is bought in on maintaining best practices.”