The Deadline Effect: Nothing Ever Happens Until The Last Minute

The graph above shows the number of proposals for talks submitted for Linux.conf.au graphed against the deadline for submissions. This reaffirms what we've long known: if you set a deadline, most people won't do anything about it until the last possible minute.

While LCA 2013 (which hits Canberra next January) had a record number of proposals, the same pattern occurs every time. There's a frantic last-minute rush of proposals submitted right on deadline day. Then the deadline date gets extended, everyone does nothing for two weeks, and there's another gush of submissions on the last day. Only a tiny fraction of people get organised and submit ahead of time. (I'm guilty of that too, but I did submit before the original deadline got extended.)

The lesson? If you want something quickly, set a tight deadline. No-one will doing anything until the day before anyway, so you may as well choose tomorrow as your due date.

So Close! [Linux.conf.au]


Comments

    Hello to Engineering

      You ain't kidding there...

    This is an unfair assumption. I am a uni student and I consistently submit my work as close to the deadline as possible, regardless of when I started. Even if I finish an assignment a few days or even weeks before it is due, I will hold on to it just in case I change an answer or would like to make any further revisions. A submission is final and finalizing early can be detrimental to my performance.

      And there are many like you (including myself), however this type of person seems to generally be forgotten. There are some interesting ways people are trying to stop this happening, but as long as this mindset exists (so, forever) it will keep happening.

      That only applies if you can't revise a submission after submitting. In the case of most conference submission systems, you can continue to revise your submission right up until the deadline.

    It's one weakness of the GTD paradigm. If a task has a deadline, you're likely to put it on your calendar, rather than your next action list. Then you don't tend to notice it until the date draws nigh.

    This graph doesn't affirm anything of the sort IF it is to scale. Not most but only 1/3 of people seem to be waiting until the deadline actually...

      Hi. Terrible graph making guy here. You made me curious, so I pulled the numbers. 60% of papers were submitted on the two due dates. So, I'd say that Angus' theory holds.

      Ummm... Are you sure you're reading it right? No. Proposals on the y axis...

    I'd be more concerned that about a quarter of the folks had submitted something before the time was half over... Surely it could have been better if a few more weeks of effort were put into their presentation.

      Chris -- this might hold true for other conferences. However, we only require a couple of paragraph abstract, so there isn't much to gain by polishing it for several weeks. We're also experimenting this year with trying to provide feedback to those who will benefit from it during the review period, which means that people have a chance to address our concerns if needed.

      I'd guess that this only looks at the initial submission date, not any further revisions. linux.conf.au, like many conferences, allows revision of submissions until the deadline. So, you could submit a draft a few weeks in advance, then revise it a few times.

    You need to plot an activity which takes 30 seconds or less in order to guage whether people are putting it off, rather than maximising their use of time available.

    What I'd like to see is whether there is a correlation between the date of submission and the quality of the presentation (if accepted). For instance, if someone is putting their submission in at the last minute, are they finishing their slide deck 10 minutes before getting on stage? And is that a problem if they're an excellent presenter? My point is, does extending the submission period positively or negatively influence the quality of linux.conf.au presentations?

      That sounds like a brilliant statistic to track! After the announcement of the accepted papers, I'd love to see a graph correlating submission time with acceptance. For a simple version, having a graph just like this one but for accepted papers only, scaled to the same scale and overlaid on each other, would provide some interesting insights.

Join the discussion!

Trending Stories Right Now