Making Decisions with Limited Data

How do you make decisions when you have limited data? What if your company doesn’t have a key piece of technology that most other companies have? How does this all affect the candidate experience?

(fair warning, I’m putting all buzz words in this article in quotes. That’s easier than thesaurus.com)

It’s easy to make “optimizations” when you have a full cycle recruitment marketing setup and the latest technology, but what if your company is missing one key component? What if that component is “actionable data” or *gasp* an Applicant Tracking System (ATS)? How do you still ensure a great “candidate experience” and smart decision making with 1000s of applicants each year?

Start with the data you do have

There’s a popular aphorism that states "Absence of evidence is not evidence of absence.” Simply put,

we may not have uncovered the right data yet but that does not mean that it doesn’t exist.

To understand a practical application of this, let’s look at Client X, a large retailer in the US.

• holds a lot of hiring events locally at stores but doesn’t have a way to apply online
• no ATS
• no clear CTA on job description pages
• PPC media and retargeting/prospecting programmatically

The challenge: what is a successful conversion for a page with no clear CTA?

Standard conversion KPI for a job description page is an apply click

What data did we have?

• Pageviews (clicks from media)
• Time on page
• Apply clicks

What is our ideal data set? All of the above, plus...

• Applications
• Quality applications
• Hires

It’s ok if your current state isn’t “industry best practice”

Without an ATS, we can’t get to ideal so let’s be smart with what we have. We decided to change our definition of a conversion to apply clicks and pageviews and found the following:

It’s no surprise that PPC provided more apply clicks as candidates are used to applying online for jobs. However, that wasn’t telling the whole story. As candidates didn’t have to click on apply, a successful conversion for Client X is a visit OR an apply click. As you can see, once we looked at a new conversion metric, we were able to see much more value in our programmatic efforts on influencing behavior. This has helped us to “optimize” our spending and make “data-driven” decisions, focusing our money on the areas that give us the conversions we need.

While I realize this is a very specific example, the point is that

there is data to help support any hypothesis you may have, but it may not be readily apparent

You need to look at the true goal of a page or campaign and create individualized KPIs that align with those goals. Don’t let your tech stack or the fact that HR Technology doesn’t like to talk to each other hold you back from making informed decisions.

Industry-standard metrics are great but if they don’t make sense for you, why not create a new metric?