Originally posted on Medium, 18 July, 2018
Unless you are a superhuman, or not human at all, you have been a procrastinator at some point in your life and the only thing that has spurred you out of a Netflix induced funk is a slowly but steadily rising sweaty panic induced by a looming deadline.
“I love deadlines. I love the whoooshing noise they make as they go by…” Douglas Adams
This was my exact experience in writing this post, and is a common occurrence in my life as a student juggling work at Applied, and doing assessments and a dissertation at the same time. You may even know that people tend to do more extreme things at the end of their decades (i.e. aged 29, 39, 49, so on), like running a marathon for the first time, or throwing yourself out of an aeroplane with a sheet strapped to your back (otherwise known as skydiving). Fortunately I’m not at that point yet…
We’re a curious bunch at Applied, so we wanted to understand procrastination on our platform and our candidates’ reactions to deadlines in the context of job applications. In particular, we wanted to use the data from our platform to answer the following:
- Are deadlines effective at boosting application numbers?.. and are there any other job characteristics which also help to boost applicants?
- Are there any demographic characteristics related to being late applicant?
- Does the moment in which an application is submittted relate to performance?
All the conclusions from these questions are useful not only for the sake of curiosity, but also for tailoring better hiring processes. We have oodles of data in the area so we split this topic into two — in this post we will un-pick question 1, and in part 2 we will tackle the rest with recommendations on how to use this data to improve job postings. Here goes….
1A. Are deadlines effective at boosting application numbers?
Yes, deadlines boost late applications
It seems really obvious right? Well yes, but what we found striking was the extent to which people seemingly need a deadline to spur them on.
Graph 1 summarises the behaviour of around 6,500 job applications that have been submitted through Applied. We can see on this graph that the majority of candidates submitted their applications during the last week (7 in 10), and there is a particular boost in applications on the day before the deadline (4 in 10 applicants submitted the application on the last day). This pattern is similar across job adverts with different durations, although the percentage of last-day submitters is higher for applications to jobs that were open for 4 to 5 weeks. In these cases, there were 5 in 10 last-day applicants.
7 in 10 applications were submitted during the week before the deadline; and 4 of those on the last day
We can see a similar pattern if we zoom into the last 24 hours and the last 60 minutes before the deadline. As we see in graph 2, 40% of last day applications were sent during the last 3 hours. And again, the last hour concentrated the majority of applications. In the case of the last-hour submissions, 21% of them were concentrated on the last three minutes before the deadline.
If you compare this data to a job with no deadline, you will simply see a steady stream of people applying for the job with some random variations — nothing like the ramp up seen in the above graphs.
So the question for you is: are you usually a last-week, last-day or last-minute procrastinator?
1B. What else could hirers try to boost application rates?
Try changing the closing day to Wednesday
Jobs closing on a Wednesday have the highest total submission rate, and also the highest proportion of people submitting just before the deadline. That means more people apply late, but more people apply overall.
For jobs that close on a Wednesday, 9 out of 10 candidates submitted in the final week, and 7 out of those were on the final day!
In second place are jobs closing on Sunday and Monday, for which around 8 in 10 candidates applied in the final week, and 5 of them applied on the last day.
The participation of late applicants decreases for jobs closing on Thursday and onwards, starting from 4 in 10 applicants who applied to jobs closing on a Thursday and finishing with 1 in 10 of last-day applications to jobs closing on a Saturday.
Tuesday presents a strange pattern that we still need to understand a bit more — do people get the Tuesday blues and not bother finishing applications? We’re not really sure yet.
In short, more late submissions can mean more overall submissions, so putting your job ad on a Wednesday is a good thing and will ensure that you get a large final boost before job closure.
Adding 4 or more sift questions makes candidates stop and think
We used the number of Applied Sift questions that were added to the job as rough proxy of complexity of job applications. Hiring managers increase their ability hiring better and more diverse talent by using our Applied Sift instead of looking at CVs . In this process, candidates provide answers to 3 to 5 questions as part of the application they submit in place of a CV and cover letter. After applicants submit their applications, hiring managers blindly and randomly score the answers for all candidates.
Given that the sift questions are asked before the job closes, we wanted to know if variations in the number of them would be related to late submissions, especially when we know that task-aversion affects procrastination. Knowing this can help organisations find a good balance between having sufficient information from candidates right from the start and making sure it is not overly onerous for the applicants.
Going from 3 to 4 sift questions increases the participation of last-day applicants and this participation maintains the same for additional questions (up to 6 questions).
As seen in graph 4, asking 3 questions resulted in 4 in 10 submissions being on the last day. When being asked 4 questions, the participation of late submitters moves to 5 in 10 candidates. In contrast asking 0 questions resulted in 3 in 10 submissions happening on the last day, although the majority of candidates in this case still send applications during the last week before the deadline (6 in 10 applicants).
So, as expected adding sift questions does increase late submissions but the effect plateaus after 4 sift questions. You would think adding ever more sift questions would decrease participation rates, however as job roles only very rarely go above 6 sift questions, this is not an effect we have seen on the platform yet.
Our takeaway from this is: ask your candidates the questions you need to understand if they have the skills and potential for the role, do not be overly concerned with the number of questions, as those that really want the job will answer one or two questions more.
When looking at the Applied data, we see some really interesting dynamics around procrastinators and the characteristics of a job that causes this behaviour. What we don’t really know yet is if this behaviour is a recruiter’s boon or bane?
Stay tuned for Part 2 of this post, where we answer the two remaining questions:
2. Are there any demographic characteristics related to being late applicants?
3. Does the moment in which an application is submittted relate to performance?
We’ll also summarise what we’ve learnt and the tips we’ve gleaned from this data that can be used to improve your recruitment.