A recruiter’s cheat sheet on how to drive high RoI on hiring ads (based on a fictitious case study)
Recruiting and staffing teams have a wealth of data related to candidates’ profiles, sources from where they were hired, job categories, salary ranges and the time taken to hire/fill positions. Mining such data to provide a comprehensive view of the recruiting trends is key to improving hiring efficiencies and also, help reduce costs.
I’m Terra Carbert, Founder and CEO of SHEops Talent. As part of Freshteam’s Leadership Series, I’ve put together a fictitious case study (but based on similar data I’ve worked on in my previous stints) to show the narrative on how to make informed decisions to hire in the future. Companies that are in the scale-up phase could use this case study to build a framework for evaluating the data they collect as they make new hires and improve their recruiting process.
Premise of the case study:
- A company, let’s call it XYZ, needs to hire 10 people for a job profile it has never hired before.
- XYZ did not have an existing talent database for the job profile, no prior applicants, no prior interviewees and some locations were remote (with low population density).
- It had a short timeline to fill in these positions to meet a customer demand.
- Some of the positions were highly specialized and the recruiting staff was already at capacity.
How to carve out approach and spending categories
XYZ’s had to make decisions to leverage everything that it already had in place. The importance of researching each source cannot be overstated. Here are some of the ways XYZ decided to approach its urgent recruiting mandate:
Since the company was hiring for this profile for the first time, we had to discover associations where people who have this specific expertise would be a member and look at schools who were emphasizing education in the relevant areas of specialities.
We looked at public resources, meaning groups or forums both online and offline where individuals with those particular skills would meet, discuss and share ideas. If it were software development, Github could be an example of a public resource you might go to. We checked out agencies or headhunters who had experience in hiring for such profiles.
Then, we looked at competition – what were they doing, how many people did they have in these roles, how did the competitive landscape for talent look like and how many were competing with us to hire for the same profile as were at that time.
We decided to classify our advertising into four buckets:
We wanted to limit our investments and cut costs but were willing to spend on agencies if it meant we would hit our targets. Some people in the hiring team had this idea that we would spend money on a magical headhunting agency who would have everything we needed to hire if we spend enough money. However, we weren’t willing to spend $16,000-20,000 per hire because the total would have been a lot.
Since we knew we were going to hire for these profiles again in the future, the goal in creating these buckets was to group some of our resources together to get trend-wise observations – do we get better performance as theorized by some hiring managers from the high cost or do we get better performance from the low cost or the free assets?
The spending categories were decided in business partnership with the hiring teams. Like, how do we look at this in the future because ideally, in their dream world, it’s always going to be free. We’re never gonna have to pay for anything ever again. The idea was that if we invest properly the first few times we hire, we may be able to build a pipeline that we can continuously go back to or build a bigger presence on some of these free approaches.
Key challenges while hiring from various sources
Taking on this project threw the following challenges:
- Process Consistency
- Salary Constraints
- Budgetary Constraints
1. Scheduling: Knowing the date by which we need the hires in the door, we could work backwards. Then, we fixed the timelines with respect to when the offer letters should go out, by when the second and first interviews need to be completed and by when the phone interviews need to be done and reports submitted. After this, it makes sense to chalk out the time required for each step and the time commitments required by every single person involved in the recruiting process.
Despite doing all this, scheduling was a huge challenge, not only because of the volume but also because business leaders who were launching a new service and a new team and were travelling and in and out of meetings, had to commit time for this process.
2. Process Consistency: Making sure the interview teams were evaluating the same things and were committed to responding was another challenge. The experience of dealing with one hiring manager versus another might be completely different. We had a business partner with influence over the hiring leaders to emphasize to honor the timelines of the process.
3. Volume: I would like to highlight how much work goes into a 10-hire process.
Some of this data is on par with analysis that I’ve done on groups of hiring in the past. So, the 33 first interviews for a 10-hire approach is very realistic and the 5:1 ratio with respect to phone interviews and first interviews is pretty standard. We probably want to see that actually closer to 3:1 or 4:1 but there were so many applications, many of whom are qualified that we needed to complete the phone screen to make the determination who’s going to be recommended to go forward.
There were so many applications! In terms of time, even the tightest, cleanest phone interview with notes, uploading it into the system, making sure it’s all documented and then emailed to a hiring manager, it’s an hour of work.
Many don’t think about the application processing time and are just like – “Oh, get as many applications as possible!” Every single application takes at least a minute of time. And that’s just a cursory glance and not even a proper scrutiny. Not to mention the clicks on the database to move the applications forward and the communications that need to go out for each application.
4. Salary Constraints: We were at the lower end of the market for some of these positions.
5. Budgetary Constraints: It would have been great to just hand this all over to an agency but that absolutely was not something that we had the budget for.
Dive deep into your data and jot down trends
In the graph above, you might have noticed that most of the time and efforts spent were on applications that came from free or low-cost sources and the high-cost (or agency) certainly did not require any labor from us.
With the results out, we know who was successful in making it to the first interview. And I’ve not included the second interview because it’s an extra step in the process and doesn’t necessarily needs to be evaluated. What was interesting is that the trends weren’t the same for all the positions.
For the more advanced roles, our free sources didn’t yield any first interviews or hires. But flip it around and you’ll see that meeting the bar for our first level engineering position was easier and our free sources did yield first interviews and hires. In terms of number of hires, the optimal performance was from “medium” sources.
The following is a much deeper dive into the data and unearthing the story behind the numbers:
This data is for both positions (Engineer 1 and Engineer 2) combined and you can filter our for each separately. We got two hires from “Indeed Organic” – “a free source” but look at how much work my team had to do for that. A 191 applications had to be processed and I can add phone interviews to show more of that story.
Now, compare this to “Hiring Manager Referral”. This channel got us 11 applications but we got one hire from this pool. The labor costs associated with a single hiring manager choosing to reach out to their network is significantly lower. So how do we amp that up and turn out the ineffective sources?
Channels such as “LinkedIn Share” and “Career Site” created a lot of volume for us but got us no hires. The question to ask here is: for sources that have a relatively lower volume (Career Site), if we had done more to drive traffic to it in different ways, would that ultimately have been the source of some of the hires that we ended up paying to get?
The “Medium Cost” sources are defined as such because we need to pay to access the data and there are labor costs associated with it considering that my team has to spend time in looking for these databases. “Resume Database 1” got us no interviews or hires. “Resume Database 2” demanded a medium volume of work and got us a hire. Where am I going to go the next time? The “Specialty Job Boards” category has a similar story to tell.
Let’s take a look at the hires and where they were sourced from:
For Engineer 1, my low- and medium-cost sources have performed well. And for Engineer 2, I now know that we got a hire from “Indeed Sponsored” for which we were willing to go for high-cost “Agency” spend. I did get a hire from the “Agency” But the cost of that – astronomical! More expensive than “Indeed Sponsored”, which is a controllable budget and can be modified anytime with the resources allocated to a different job profile at any time.
However, my “Agency” channel is plugged in and the expenses are coming in, no matter what. So, maybe the next time I go to hire Engineer 2, we won’t use the agency but would likely invest more in “Indeed Sponsored”.
Key guideposts to optimize costs while hiring
Volume impact is costly: This includes not only labor costs but also in terms of burnouts of the team members involved in the recruitment process. Yes, looking at resumes and screening applicants is part of their jobs but how many resumes can you stare at before you want to throw the computer at the wall?
Stop low percentage success sources quickly: If we could do it again, I would look at the numbers and do away with the sources that do not get me enough phone/first/second interviews or hires. So for a resource where 1 in a 100 applicants make it to the phone interview, I’m done with that channel. I don’t want recruiters to be looking at resumes from places that aren’t even getting us interviews.
Communicate timelines early: The next time onwards, I would communicate the timelines earlier because some of that was reactionary with the hiring management team. There’s some guiltiness on my part as well with my team of not helping them to carve out the time for phone screens ahead of time because many of them came to me during the process and were like – “How am I going to get all this done?” I could have helped them plan ahead and assured them that I have their backs. I could have even done better by proactively telling other internal business customers about my team’s current projects/their available bandwidth, and moved the time to fill a certain job role or assign it to someone who can dedicate more time for it.
Establish performance targets ahead of time where possible: Maybe I could have (and also, did in the future) looked at the data ahead of time around what we were getting for some broad-stroke specialty positions even though we’ve never hired for these roles before. So, I couldn’t compare apples to apples. I could have looked at the data and asked what are the trends with respect to the different sources for a particular grade level or job category.
This could have helped us get early signals on how to control the volume because in this scenario, I didn’t want that many applications.
Subscribe for blog updates
Thank you for subscribing!
OOPS! something went wrong try after sometime