At Greenhouse, we’ve been spending a significant amount of time on elevating reporting in 2020, from conducting customer research to releasing enhancements to our Greenhouse Reporting (GHR) tool. All this work will culminate in a pretty exciting overhauled reporting experience for our customers toward the end of this year.
But, before we get there, we want to share some pretty eye-opening learnings the team uncovered in their research, and hopefully dispel a few myths about reporting in Greenhouse. A few themes kept rearing their head in our conversations with customers and we wanted to address them – head on.
Challenge 1: Some users weren’t confident in the data being presented
In general, reports are a lagging indicator, which means they are giving you information on what has already occurred. Yes, they provide trends and signals that yield action, but the very nature of data is that it summarizes things that have already happened. It’s imperative that your workflows and actions lend themselves to clean data so your business can measure and iterate based on accurate data.
Over dozens and dozens of customer research sessions we continually heard that some customers were not seeing accurate data in the reports they were pulling from Greenhouse. But based on the very nature of reports outlined above, we knew that making the data in Greenhouse reports more trustworthy would actually require that we start much further up the behavior pipeline to ensure there was good data coming in before we got to business of dealing with any data coming out.
Solution 1: Good data in, good data out
And so we have spent, and continue to spend, a large portion of our efforts focused on driving healthier usage and proper product adoption in order to produce more trustworthy data to yield accurate reports. We pinpointed the cause of these data issues earlier on in the process and established trainings and feature releases to help increase the likelihood of gathering clear data upfront.
A few recent feature enhancements to help drive better data quality include:
- Marking jobs as official templates to mitigate the errors when repetitive data is being re-entered over and over again, such as desired candidate attributes and common interview questions.
- Dependent offer fields give admins the ability to set up dependent custom offer fields for a cleaner, easier, and faster way for recruiters to generate offers with only the relevant information displayed, removing the potential for human error and ensuring cleaner data.
- Scorecard upgrades including grouped focus attributes and rich text make it easier for the hiring team to submit their scorecards, which in turn makes it easier to report on these decisions and a company’s structured hiring practices.
- Simplified interviewing permissions make it easier for recruiters to assign interviewers per job, further reducing any concerns about data leakage in reports.
Helping customers to ensure their data is clean and accurate is a continued focus for us. As we diagnose “bad” data and determine where it is coming from, we’ll continue to offer best practices, educational programs and in-product enhancements to catch it early and course correct.
Challenge 2: People were working outside of the system, making cohesion a challenge (or impossible)
Through our research, we also discovered that some customers were running key workflows outside of Greenhouse. Sometimes this was because a specific piece of data wasn’t readily available in the system for reporting, but often it was because users were performing entire tasks outside of Greenhouse – recruiters would keep their scheduled interviews in an excel file, or a hiring manager would keep track of all of their favorite candidates in a Google doc -- and so it was impossible for us to have accurate data in the system about these activities in the first place.
You cannot manage what you cannot measure, and our reporting tools simply didn’t have access to the data we needed. This leads to all sorts of sub-optimal outcomes, and often meant users were cobbling together data from Greenhouse and other disparate data sets, which led to easily avoidable human error, inconsistencies and overall confusion.
Solution 2: Consistency is key
The first step in correcting this issue was to make sure that the data people want to report on and keep track of is available in the system, and then making sure that the tools themselves actually let you report on those things!
Over the last few months, we’ve enhanced our reporting tool with:
- Essential candidate reports
- More data fields options in our custom reports
- and more…
In our next Solving the challenges blog installment, we’ll address other challenges we’ve heard around customization limitations, discoverability, permissions and performance. Get excited, because we have some upcoming announcements in this realm that will make it much easier to answer the right questions all within the Greenhouse reporting tool.