Written by

In Pursuit of Perfection: Using Testing for Conversion Rate Optimization

Data-Driven Intelligence, Programmatic| Views: 2968

The ever-changing behaviors of job seekers, coupled with the evolving recruitment ecosystem, challenge recruiters to always be improving their hiring strategies. High competition for talent means that there is always more to do, and that success is highly dependent on data-driven testing.

Continuous improvement in recruitment focuses on an end goal: the measurable impact on the bottom line and the number of applicants or hires. The process is known as conversion rate optimization.

What is Conversion Rate Optimization?

A conversion happens when a user performs a desired action, such as clicking a link, submitting a contact form or completing steps in the application funnel. The specific action that constitutes a conversion depends on what the measurer is trying to achieve.

Conversion Rate Optimization is a structured and systematic approach to improving the conversion process and success rate – whether that’s via a website, mobile app or some other mechanism. The optimization process is defined by specific KPIs and continuously informed by data and insights delivered through analytics or user feedback.  

Conversion Rate Optimization in Recruitment: Applications and Techniques

In recruitment, a conversion can occur at both the job seeker and the job applicant stages. Conversions at the early stages of the application funnel means turning an active or passive job seeker into a potential job candidate. It occurs when job seekers take some kind of action that puts them into an application funnel (e.g. submitting an application).

A conversion further down the funnel – in the job offer and on-boarding stages – means that applicants turn into new hires. A recruiter’s goal is to maximize both types of conversions.

Recruiters should look specifically at the following areas when creating a plan for Conversion Rate Optimization. Each of these areas can be continuously improved: 

  • Job ad content. A/B testing is crucial at this top-funnel stage. Job ads must be relevant to specific candidate searches. Sometimes changing just one word or adding an additional piece of information can lead to a spike in click or application rates.
  • Sourcing channels. The many factors that go into candidate sourcing should be continuously monitored and optimized. Consider how experimentation with ad budgets/bids or traffic mixes could impact the bottom-line or number of conversions.
  • Application funnel. The order and length of steps in the application funnel, plus all associated touchpoints, combine to create a specific conversion rate result. Improving upon any of the small or large steps in the funnel can have a larger, relative impact on the final campaign outcome. 
  • Onboarding practices. Once a candidate is offered a job, the onboarding becomes an opportunity to increase productivity and subsequent Employee Lifetime Value (ELTV). Technically, this area goes beyond the two types of recruitment conversions discussed above, and further into converting new hires into productive hires.

Employers can use a four-step technique during any of these stages to achieve Conversion Rate Optimization:

  1. Identify friction points. Examples include poor branding, sourcing problems, usability issues, ineffective hiring processes and unattractive offers. Are there proper security measures in place throughout the application funnel? What about social proof and trust-building language? How does the written content and visual content flow together?
  2. Consider testable hypotheses based on these friction points. Metrics that can help inform these hunches include total conversions, conversion rate (conversions / total number of visitors), bounce rate or exit rate.
  3. Leverage all available tools for A/B testing and analytics. Think Optimizely, Google Analytics, Kissmetrics, Perengo, etc.
  4. Create variations of items for testing – such as landing pages, calls to action, application forms – and carry out A/B tests using your tool of choice.
  5. Analyze data from the results, apply changes to the application funnel and continue testing for further improvements.

With proper testing methods in place at each friction point, employers can be assured they are always optimizing with minimal effort involved.

Bottom line: The perfect recruitment campaign is a myth.

Continuous improvement is an important reality – one that can have a major impact on sheer hiring numbers, especially for large organizations.

When recruiters complete a successful hiring campaign, they should treat it not only as a milestone, but also as a source of valuable data for further optimization. A Demand-Side Platform can help set up valuable testing processes to achieve this while instilling the mindset of continuous improvement into teams who can benefit from it the most.

About Spencer Parra

Spencer Parra is the VP of Product Management for advertising and data products at Radancy. In that capacity, Spencer and his team of product managers, program managers, data scientists, and data analysts work to develop products in a data driven mindset. As the leader of Advertising products, he works to bring a holistic full funnel approach to Radancy’s advertising technology stack with Programmatic Jobs at its foundation. Through data products, he tells the story of media performance via Radancy’s Metrics Gateway and helps ensure data is democratized through Radancy’s unified platform. Spencer came to Radancy from the Perengo acquisition in mid 2019 where he served as Lead Product Manager and a member of the founding team. With Perengo, he worked towards the vision of leveraging the same rigor and concepts from ecommerce advertising technology to the recruitment advertising space. Prior to Perengo, Spencer launched and supported in-app advertising products at Criteo as a solutions engineer. Spencer holds a B.S. in Aerospace Engineering from the Massachusetts Institute of Technology.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to top