Friday, August 22, 2008

Universities with hidden talent

Cross post from CoNotes blog:

In the recruiting world, there is so much subjective opinion that it is very difficult to decide if Candidate A is better than Candidate B. That is why brand name universities and companies carry so much weight on resumes.

What ends up happening, though, is that every startup looks to hire some undergrad engineers from Stanford. They all send an email out over various Stanford listservs, hoping for a hit. Sometimes you’ll get a hit, but a lot of times you are just another fish in the sea of startups recruiting at Stanford. And while students at Stanford are great, there are tiers of students. Some startups are inherently going to grab the top tier students, and some the 2nd tier…but the question is whether it’s worth going for the lower tier students or whether you should really start spending time searching for hidden talent at other schools?

So I wanted to see if there was some way to prove there was talent at lesser known schools. The tough part is finding a common, easily-available metric for “talent” at different universities. I couldn’t come up with anything good, so I settled on what data I could find easily—SAT scores at different universities. Before any of you jump on me for using SAT scores as a proxy for qualified candidates for a startup, I ask that you help find me a better metric that can serve as a proxy for such a purpose.

Given that caveat, I really just want to show that there are some non-brand name universities with just as much talent as brand name universities. I also wanted to narrow down the scope a little since a large number of entry positions at startups are engineering related. So I used US News’ list of top undergraduate engineering schools. The names most people would expect are on there (Stanford, MIT, Berkeley), as well as some other schools less well known for their engineering programs (Michigan, Cornell).
  1. Massachusetts Institute of Technology
  2. Stanford University
  3. University of California—Berkeley
  4. California Institute of Technology
  5. Georgia Institute of Technology
  6. University of Illinois—Urbana-Champaign
  7. University of Michigan—Ann Arbor
  8. Cornell University
  9. Carnegie Mellon University
  10. Purdue University—West Lafayette
More surprising were the schools on a second list for universities without doctorate programs.
  1. Rose-Hulman Institute of Technology
  2. Harvey Mudd College
  3. Cooper Union
  4. Cal Poly—San Luis Obispo
  5. United States Military Academy
I personally know of these 5 schools because of friends that attended each school, but I doubt a lot of people know that these schools have some great undergraduate engineering programs.

After gathering SAT data on these schools, I then ranked them by percentage of the student body with SAT scores above 1500.
  1. California Institute of Technology
  2. Massachusetts Institute of Technology
  3. Stanford University
  4. Harvey Mudd College
  5. Cornell University
  6. Carnegie Mellon University
  7. Cooper Union
  8. University of California—Berkeley
  9. University of Michigan—Ann Arbor
  10. University of Illinois—Urbana-Champaign
  11. Georgia Institute of Technology
  12. United States Military Academy
  13. Rose-Hulman Institute of Technology
  14. Purdue University—West Lafayette
  15. Cal Poly—San Luis Obispo
(Scribd doc with more data at end of the post or link to doc here.)

What stood out to me were 3 specific schools: Harvey Mudd, Cornell and Cooper Union. Harvey Mudd at 32% of their school with SAT scores above 1500 is pretty much neck-and-neck with Stanford. And all 3 have higher percentages than Berkeley and Illinois, two schools with much better known engineering schools.

Now if you want to then look at universities in terms of absolute #s of students with SAT scores above 1500, you get this ranking:
  1. University of California—Berkeley
  2. University of Illinois—Urbana-Champaign
  3. Cornell University
  4. University of Michigan—Ann Arbor
  5. Stanford University
  6. Massachusetts Institute of Technology
  7. Carnegie Mellon University
  8. Georgia Institute of Technology
  9. Purdue University—West Lafayette
  10. California Institute of Technology
  11. Cal Poly—San Luis Obispo
  12. United States Military Academy
  13. Harvey Mudd College
  14. Cooper Union
  15. Rose-Hulman Institute of Technology
This ranking only reinforces Cornell’s standing as a great place for startups to recruit. With just as many undergrads scoring 1500 or above as Illinois, and more than Stanford or MIT, it seems logical to at least make a small effort to reach out to Cornell engineering students.

At the same time, these are the #s that really hurt schools like Harvey Mudd and Cooper Union because companies feel like there aren’t enough qualified applicants in absolute terms at the schools—despite the high percentage of very smart students. But both these schools are in very convenient metropolitan areas (Harvey Mudd is in LA and Cooper Union is in NYC), so the incremental cost is not that high.

While I want to emphasize again that SAT scores are not a great proxy for talented people, it does indicate that there are some schools with hidden talent pools that startups really should be targeting. And moreover, they might want to spend more time at these other schools than some of their current target schools.

Read this document on Scribd: Universities with hidden engineering talent


Anonymous said...

So your method of recruiting has nothing to do with the school whatsoever, but rather the school's SAT selectivity? That is kind of stupid, no offence intended, in that you have not considered the quality of education that each school offers.

Andrew said...

I would definitely consider "quality of education" if there were a quantifiable, unbiased metric for it that was readily available for a large number of colleges. However, given the lack of metrics for a lot of more relevant attributes of universities, I went with SAT selectivity. It's crude and inaccurate, but the main point I wanted to make was that there are extremely high-caliber individuals in schools that are not considered "name brand" schools.

bc said...

If you admit that it is a crude and inaccurate metric to begin with, maybe there isn't much of a conclusion, either for or against, regarding your original hypothesis/opinion.