What he was doing was phone screening two people a day. Actual interviews were probably 1/20th that - so, approximately 400 Resumes = 100 Screens = 20 in person Interviews = 4 Hires. Which is a reasonable average over a year.
[edit]
My strategy has always been for posted positions that I'm actually acting as the recruiter for, is to do the initial resume screen myself (I can triage a resume in 1 minute for a position that I know about), and then hand over the screening to a reasonably competent (but unemployed) person in that area, ideally with a set of standard questions that the two of us have come up with together.
I.E. Hire a Unix Sysadmin to screen systems administrators, and give them a list of questions like, "what is an inode, what is the difference between a process and thread, how would you sum up the total size of files in a directory, what's your favor unix based operating system and why, tell me about the unix systems you've worked on?" - This usually takes 30 minutes/person.
The interviews with around 5 people takes about 45 minutes per interviewer (30 minutes to 1 hour) - so, about 4 hours per candidates.
Total investment per hire is about 136.5/4 = 35 hours. Only 7 hours/hire needing to be done by the hiring manager. The screening is the easiest stuff to be outsourced.
Of course, you can dodge 90% of this if you just get great internal references (removing the need for screening, filtering) - which is why internal references are so highly valued by companies. (That, and they don't have to pay $20K-$30K to an external recruiter)
Depends on your pool. Small companies need to be absolutely brutal about only hiring the best people, with the best skills, and the best fit for the company. I went to a local university career fair looking for two different junior level folks. I took back over 40 resumes, made 3 follow-up phone screens, and 0 hires. 500 real interviews for 4 hires is a bit much, but in general I pass on people all the time.
I've done the same (50 resumes for two positions, phone interviews with 15-20, follow-up in-person interviews with 4). That's 10X interviews. But 100X? Assuming anyone he hires goes through multiple follow-up interviews, I don't see what information he's getting from these interviews that couldn't be handled by a subordinate or by email. Evaluating someone for "fit" is impossible in a short interview like that.
That's not uncommon, assuming that an initial phone screen counts as an interview.
I have had stretches where it took me talking to ~80 candidates between initial phone screens, random chats, etc over two months before finding someone to hire.
And to be clear, I am not claiming that I found the 'best' person out of 80 people. 'Best' is a very custom definition depending on the environment, role, and fit.
Each candidate probably had more than one interview. 100 interviews for each hire is not all that high, especially if you don't have a lot of intern conversion hires (that's why larger companies can hire more efficiently).
Depending on how you define interview. In my experience a rough funnel of 100 people (random people through a job board, not referrals) -> 20 phone screens -> 5-10 actual interviews. Generally at large organizations in the Valley a rate of 10 on campus interviews to one offer is considered inefficient (shoot for ~5 from an efficiency perspective). Too many in person interviews generally suggests a weak filter in an earlier step (you can learn quite a bit from a phone screen, for instance).
HireHand is trying to help solve this common problem. We screen your candidates using screeners with skills matching the position you are hiring for. Then we present you with the set of candidates that have the skills required, leaving you to interview for "fit". Check us out: http://www.hirehand.com
Regardless, it never ceases to amaze me how controversial pricing discrimination is. Especially since most people have taken Econ 101: pricing discrimination eliminates dead-weight loss, which should be a good thing right? Nope, instant controversy.
Er. Um. No, most people have not taken Econ 101. Most graduates have not taken Econ 101. Most graduates in quantitative disciplines have not taken Econ 101.
Most people with economics degrees have taken Econ 101. A modest number of other people have. Some more have learned the basic concepts by means other than taking Econ 101. These are probably not the people complaining about price discrimination.
In any case, being surprised when people are upset because they think they're getting inferior treatment and saying "oh, but they should think of the gain in overall economic efficiency" seems like a sign of, well, not having taken Psych 101.
Price discrimination doesn't eliminate dead-weight loss (it really has nothing to do with dead-weight loss), it just converts consumer surplus into producer surplus.
Did you read the article you just linked to? Here's a direct quote "From a social welfare perspective though, first degree [perfect] price discrimination is not necessarily undesirable. That is, the market is entirely efficient and there is no deadweight loss to society."
Without price discrimination, unless you're operating tax-free and subsidy-free there is always a dead-weight loss. Perfect price discrimination eliminates it. It is literally right there in the article you just linked to.
Actually, it does both - capturing consumer surplus for buyers who pay more than the average and eliminating deadweight loss for buyers who pay less than the average.
[Technical note: By "average" I really mean the optimal price if the producer is only allowed to set a single price point. But let's not quibble.]
"buyers who pay less than the average" (which I'm assuming means buyers whose reservation price is below the market equilibrium) do not qualify as a deadweight loss. Deadweight loss refers specifically to:
a) buyers who cannot buy something which causes an economic benefit (they gain more than it costs to produce). Examples: shortages, monopoly pricing.
b) buyers who buy something that does not cause a net economic benefit (it costs more to produce than they gain). Examples: gov't subsidies artificially reduce price, negative externalities not factored into price.
Okay, let's take the monopoly example from wikipedia. Firm prices at monopoly price (60 cents), deadweight loss because all transactions for buyers with marginal benefit between 10 and 60 cents do not occur. Assume some identifiable group of people exists with marginal benefit < 60 cents (e.g., students). Firm introduces student discount => deadweight loss is smaller. Therefore, price discrimination can reduce deadweight losses.
Close, but your example is missing a crucial piece. Specifically, the marginal cost of producing the product.
Let's say the MC is 20 cents at the scale produced (Q1). The second group (students) get a MB of let's say 30 cents. They will not buy at a price of 60 cents, therefore there is a deadweight loss (MC < MB to students, but they do not have the good). If the monopolist can expand their output to Q1 + Q2 (where Q2 is the student sales) while keeping their MC under 30 cents, they can reduce the DWL.
Umm... so you agree? Price discrimination can reduce DWL?
By the way, there's no need to assume MC is a function of Q since that just complicates the example for no benefit (in the wiki article notice that MC is just a constant 10 cents). And yes, though I didn't stress it, of course the discount price must still be greater than the MC (and so in the range of 10 - 60 cents).
I didn't understand your argument at first; after you clarified I believe you are correct (with the corrections mentioned). My comment was intended to fix some flaws in your argument that rendered it incomplete, not incorrect.
I believe the constant MC assumption is dangerous though. MC is rarely fixed in microeconomics. It is almost always assumed to be an approximately linear function of quantity. Thus, increasing production will always increase MC. You'll need to hire more workers, pay more overtime, buy more expensive supplies, whatever (in the short run; in the long run firms expand production and the supply curve shifts). The assumption that MC won't change is almost always incorrect, and the MC changes affects whether the argument is valid or not.
Sorry if I'm sounding pedantic, just pointing out what my econ prof would've said to me about this argument.
EDIT: and I'm not trying to imply that you don't know any of this, just that it may benefit readers who have less economics exposure.
Where did the dead-weight loss come from? It's just consumer and producer surplus. There is no dead-weight loss in this situation as we've been discussing it.
From the trades that don't happen because the producer can't sell to certain segments at a lower price (for fear of reducing his profit on other segments). I suppose it's worth noting that this isn't an issue in perfectly competitive markets (since there price = marginal cost). But, outside econ-101, there aren't really many perfectly competitive markets.
By the way, as I mentioned in my first post, I definitely agree this is about capturing consumer surplus too.
It seems obvious why two people getting the exact same product / service but paying a different amount is a controversial matter. Especially, since the discriminating factor is the fact that you are willing to pay more!
If a vendor wants to price discriminate, there are ways of doing it, such as loyalty programs or rewards, which do not 'offend' the consumer and still manage to capture (most of) the consumer surplus.
In this case though, this is not what Orbitz is doing, they keep the prices constant but show different default options.
Isn't that the right way to do it? If Mac users choose options (a, b, c) the most often then these are the ones that you should default to. And the same for non-Mac users but with frequent options (d, e, f).
And in this case it happens to be that, on average, (a, b, c) have higher prices than (d, e, f).
Let's say Best Buy could magically learn your income and checking account balance when you walked in the door. And just before you walked in, employees rushed and hid all of the lower-priced TVs and computers. How would you feel about that?
Pricing discrimination is controversial because it shifts most of the surplus away from consumers. A lot of people come out of econ 101 thinking any more efficient solution is inherently better, but efficiency is only one factor to consider.
Price discrimination is revenue optimization for the vendor, it's not eliminating some efficiency -- it's steering me to part with more of my economic resources.
I expect the kind of behavior from a salesperson with whom I am negotiating -- that's a customary way to do business in that model.
In the United States, we learn about fixed-cost in the grocery store -- an apple cost $1 whether I have $0.50 or $5,000 in my wallet. When I'm buying a product or service on what appears to be a fixed-cost basis, I perceive your "pricing discrimination" as taking advantage of me.
Let me clarify my point. Our economy is based on the notion that things are worth what people are willing to pay for them. What if some people are willing to pay more than others? Then businesses have to choose: either they practice price discrimination, i.e. try to sell each and every unit for exactly what that person was willing to pay, or they set one price that will attempt to maximize their profit.
You can find many examples of each. Anyone who auctions off their goods practices price discrimination. i.e. 3 identical baseball cards auctioned separately on Ebay fetch 3 separate prices. Same thing with a seller in a street market who gives different prices to different buyers (tourists vs. locals) and then haggles with each and every buyer.
My point is, people seem fine with this kind of price discrimination in a lot of settings, but not all settings. For some reason, there are certain settings and triggers that change their view from "hey this is how things are supposed to work, this thing is worth what I'm willing to pay" to "hey I'm getting ripped off!" Why is that? And what are the triggers?
Businesses put a lot of work into finding pricing discrimination "tricks", including discounts and premium upgrades, and that on the whole is a fascinating topic to me.
The 1990's were fifteen years ago. Since then Carmack has spent 60 hours per week writing code for every conceivable console, PC, and mobile platform. Most people in his position would have moved into management by this point, he chose not to. As a result he has more expertise on more topics than just about anyone else in the gaming industry. And more than most in computer science generally. If he has an opinion on something, its because he knows what he's talking about.
I'm not saying he doesn't know what he's talking about, he just never liked consoles, he was always a PC guy and if he were truly "excited" about the nextgen that's like hearing Armstrong will quit riding bikes and be a F1 pilot.
And I'm saying, your opinion is stuck in the 1990's. John Carmack was very excited about both the xbox 360 and the Nintendo Wii. He's just not excited about the next generation because, in comparison to the most recent generation, he believes we're at diminishing returns for having more and more powerful 3d graphics. Which is exactly why he's taking VR headset development into his own hands.
Profiling is absurdly counterproductive, but not for the reasons the paper describes. We don't need detailed analysis for this one, simple math will do. Just to be clear, we're talking about heightened security procedures for passengers of Arab ethnicity (as a proxy for people of Muslim faith.)
The problem? The majority of muslims aren't Arab! There are large numbers of:
- asian muslims, Indonesia = world's largest muslim country
- black muslims, across various African countries, i.e. The Underwear Bomber
- white muslims, in the balkans and Chechnya for example
- And of course there's large numbers of muslims in India, Pakistan, and Iran who are not of Arab ethnicity, although they might or might not look similar enough (depending on dress most likely) to get thrown into the same profiling bucket.
The paper goes into the details of how a profiling system would be optimally set up, but the entire issue is moot. The premise that we have a good understanding of what groups should be receiving heightened security screening is itself wrong.
(Note: I couldn't find the statistic with a few minutes of Google searching, so I'll keep looking, but I know I have seen it reported that >50% of muslims in the world are of ethnicities other than Arab.)
The paper describes profiling as being statistically optimal, not counterproductive.
Also, no one proposes using Arab ethnicity as a proxy for being Muslim - who cares if someone is Muslim? Arab ethnicity (or being a Muslim) is used probabilistic evidence (not a proxy) for a person being a terrorist.
Uh-oh, lookout, it's a Highlight killer! Just like how facebook killed foursquare (fb check-ins), quora (fb questions), Craigslist (fb marketplace), and whatever it is they were trying to kill with fb credits, now they're coming for Highlight.
Might as well give up now, Highlight. You're doomed.
I don't think it's possible to celebrate Turing's life without mentioning how his life ended. Turing was without a doubt a war hero. His work allowed the Allies to crack the Nazi's enigma code, without which the war could have gone very differently. Nonetheless, a few years after the war the UK government, upon discovering that Turing was gay, prosecuted him for homosexual activities and treated him horrifically. His security clearance was revoked, and it's not hard to imagine that this had something to do with his suicide at age 42. Just something to keep in mind on this gay pride weekend.
It seems the barbarians of medievalism and quackery are always there just outside the gates, ready to drag us back into the dark ages if those who value science over superstition don't proactively keep the light burning.
Anyone know of any Turing events going on in the Bay Area?
There was the ACM Turing 100 conferences last week, but you had to register a while ago. It was a really great celebration, with interventions by a lot of Turing prices (Don Knuth, Vint Cerf, etc.). If you like theoretical computer science, you should definitely take the time to watch the video at [2].
It's easy to pretend that it's was the Nazi's and that the Nazi = "Speak German and worship Hitler", and so long as you don't do that, then you're OK and nothing you do or say can be linked to the end of Turing's life.
There is homophobia in many many cultures, all over the place. Even Great Britain.
Exactly, and my point is that these homophobes, racists, etc. have much more in common with the Nazis than their passports make them believe. You could also argue that fighting against Hitler laid bare the hypocrisy of racial injustice in the United States of America, sparking the Civil Rights movement.
haha, the whole time I was clicking around, a non-zero part of my brain was sitting there thinking, "ok, but where the hell did that daddy thing come from?"
I wasn't saying facebook just bought it for the domain name. I was saying face.com, as a startup, at some point bought that the domain name "face.com" despite having little to no revenue.
Wait a second, he was interviewing 100X more people than he was hiring??? Does that seem a little extreme to anyone, or is it just me?