June 28th, 2011 by Anna Lewis

On Recruiting Some Statistics on Recruiting

Fog Creek Software’s summer internship has always been the cornerstone of our strategy for recruiting top developer talent. And, to be honest, we’ve never felt the need to track applicant data too carefully. Thanks to the Joel Factor and our reputation for being a phenomenal place for developers to make great software, young coding whizzes have always just seemed to find us and apply.

But, in recent years, our intern class has doubled in size, from around five wunderkinds to around ten wunderkinds. Last year, as we saw our internship recruiting needs increase, we decided to attend on-campus career fairs for the first time. In September and October, we visited four schools: Carnegie Mellon, Princeton, Columbia, and Rutgers. As applications started pouring in, we realized that, in order to compare career fairs to our other recruiting strategies, it would help to know the real numbers behind our recruiting process.

So, over the course of the past year, our intrepid recruiting team has tracked a big ol’ bunch of data. When possible, we’ve compared it to relevant data from previous years. We’ve learned a lot not just about the impact of career fairs but about internship recruiting trends in general. And while we make no promises that you’ll find this information useful in any way whatsoever, we foolishly think it’s too fascinating to keep to ourselves. We just can’t help sharing!

Chart #1

Number of Applications Received This Year and Last Year

A few things to note:

1. We received way more applications this year than we did last year. In fact, we received way more applications than we did during each of the past two years. Here’s a table with the direct comparison:

2. See the two gigantic spikes – first in the fall and again in the spring? That happened both years. It seems safe to say that students just don’t think about internships while they take their fall semester exams or celebrate Festivus.

Chart #2

For the first time ever, we actually asked all our applicants how they found out about our internship. Here’s how they answered:

A few things to note:

  1. Besides attending career fairs, we used a few other methods of marketing our internship: a) in late January, we sent emails to computer science departments and posted ads on college job boards; b) throughout the recruiting year, we sent brochures and a letter to computer science departments; c) we posted a link on Hacker News.
  2. More applicants found us on the web or through the career fairs than in any other way.
  3. Applicants may have lied. In which case, our information is completely meaningless.
  4. 11% of applicants failed to answer the question, “How did you find out about our internship?” Some applicants claimed they couldn’t remember. Others answered, quite simply, “Yes.” To which we answer, “Wow. Just wow.”

Chart #3

Chart 3 combines the two previous data sets. It shows the number of applicants over the course of the year, sorted according to how the applicants heard about our internship.

A few things to note:

  1. The candidates who caused a huge application spike in the fall found out about us from our fall career fairs. No surprise there.
  2. Most of the spring applicants learned about our internship via the departmental emails we sent or job ads we posted in late January. Also no surprise.
  3. But, also in the early spring, we saw an increase in applicants who claimed to find out about us via “word of mouth” — i.e. informal oral (and, of course, digital) communication. In-ter-est-ing. Maybe all those emails and job postings got kids buzzing about us on campus, which is always a good thing. At least there seems to be a correlation.
  4. On the other hand, as we know from Chart #1, more students tend to apply in the spring no matter what. In fact, a handful of students who learned about us from fall career fairs waited until the spring to apply. We chose to attend fall career fairs in order to get in on the recruiting action as early as possible. But since it seems that many students aren’t even thinking about internships until the spring, it might be worth attending spring career fairs instead of or in addition to fall career fairs.

Chart #4

But were all these candidates any good? Is there any correlation between how they found out about us and how well they did in our screening process?

Here’s a table that shows how many applicants reached each stage of our screening process and how they found out about our internship.

And here’s a chart that shows the same thing in pretty colors.

A few things to note:

1. Our screening process has evolved since Joel wrote his Guerilla Guide to Interviewing. This past year, our screening process has consisted of:

  • the résumé review, in which our developers sort résumés and cover letters
  • the “code screen”, a phone interview that requires the candidate to write code
  • the “phone screen,” a second phone interview that requires the candidate to tackle a high-level design question (no coding required); and
  • a full day of at least six in-person interviews that require the candidate to write code

2. Of the career fairs we attended, Rutgers was the only one that did not lead to an offer. CMU, Princeton, and Columbia each led to two offers. In other words, six of fifteen offers originated in career fairs. Not bad.

3. Which origins led to no offers at all whatsoever? Zero, zilch, nada, not one iota of an offer?

  • “Word of mouth” resulted in no offers – which is interesting because many recruiting resources claim that generating buzz among students is one of the most valuable recruiting strategies.
  • Job postings through career center sites also resulted in no offers.
  • Departmental emails also resulted in no offers.

In other words, all those extra applications that contributed to our springtime spike in application numbers? No offers. Bummer.

4. Whereas, a whopping eight offers were made to applicants who had found their way to us all by themselves with no time and effort on our part: four who’d done web searches; and four who’d run into or followed Joel’s work over the years.

So, career fairs turned out to be a great recruiting strategy, right?

Not so fast, buster. Sure, career fairs turned out to be a great way to attract strong candidates. But most of the candidates we gave offers to who had found us through career fairs – shock! horror! — rejected our offers.

Chart #5 Results of In-Person Interviews


One big thing to note:

  1. Curses upon Google, Facebook, and Microsoft in the ongoing battle for top developer talent.

In conclusion…

It was exciting to get so many more applications this year as we tried career fairs and tracked applicant data for the first time. But we knew all along that more applications wouldn’t necessarily mean more superb applications or, ultimately, more interns. After all, the goal of recruiting is not to attract as many applicants as possible. It takes a lot of time and work to process a mountain of résumés, so they’d better be high-quality résumés. The goal for recruiters is to zero in on applicants who are “the right fit” while avoiding as many candidates as possible who aren’t “the right fit.”

Did career fairs get us closer to candidates who are “the right fit” for Fog Creek Software? Not exactly. Career fairs definitely gained us access to some of the top coding talent in the country. But we’re looking for the few top coders out there who will be inclined to choose us over Google and Facebook. We know that’s a tall order. We’ve discovered that, for the most part, our target candidates aren’t tossing their résumés on our table at career fairs. They seem to be students who are already part of a self-selected group. They’ve been plugged into the Joel Factor since they were teenagers. Or they’re the kind of students who’d rather skip the on-campus dog-and-pony show and just go online to learn for themselves about our culture and our products.

Will we repeat career fairs this year? The jury’s still out. Career fairs seem like a painfully arcane way to connect top coders to software companies. But, possibly because universities have money at stake in bringing students and companies together on campus, career fairs remain a staple of recruiting strategy across the country. They DID bring us two members of our current intern class and we wouldn’t want to write them off entirely. This was just our first time trying them and, at the very least, we might find that our repeated presence on campus will help build brand.

In other words, attending career fairs for the first time was a valuable experience for Fog Creek. And we’ll be sharing even more of what we learned in tomorrow’s blog post, “Do Career Fairs Have to Suck?” (The answer’s not what you think.) But, more than anything, career fairs taught us something we already knew: Fog Creek’s best recruiting strategy remains, quite simply, making great software and being a great place to work.