QuitRank: Google Algorithm Predicts When Employees Will Quit

Oh my goodness. According to this WSJ article (here), Google has developed an algorithm to determine which of its 20,000 employees are most likely to quit:

The Internet search giant recently began crunching data from employee reviews and promotion and pay histories in a mathematical formula Google says can identify which of its 20,000 employees are most likely to quit.

Apparently, Google’s newest algorithm is using data from employee reviews, promotion histories and pay scale histories in the mathematical analysis. But much like its PageRank search algorithm, don’t expect anyone at Google to disclose any details on this algorithm:

Google officials are reluctant to share details of the formula, which is still being tested. The inputs include information from surveys and peer reviews, and Google says the algorithm already has identified employees who felt underused, a key complaint among those who contemplate leaving.

I wonder what they are going to name this algorithm. I’m suggesting PeopleRank or QuitRank or FireThemBeforeTheyQuitRank.

It’s not very surprising that Google is worried about people leaving. They have lost a lot of big names in the past few months (Tim Armstrong, David Rosenblatt, Santosh Jayaram). Google has grown so much over the past several years. 20,000 employees already? I’m thinking that Google is not as attractive as it once was.

When attractive start-ups grow too big, the romantic appeal fades. And quickly. The ladder gets taller, and it can become tougher to get things done. And it can feel more like you’re part of a machine rather than a person who can affect the machine’s design and success. Furthermore, some former employees speak of little to no training and an impersonal HR department. Check out some of the reviews that were leaked here. [Note: After reading those reviews, I realized that no matter where you work, you will complain about it. It’s just human nature.]

On one hand, I hope Google succeeds with this algorithm. They are treating people as data, and maybe humans are predictable. Maybe not. On the other hand, if this helps Google address employee concerns, more power to them. I guess. I’m conflicted on this one.

Google has found a way into most of our internet lives in some way or another. I really hope they don’t lose any more of those brilliant minds. But I’m selfish. I want more/better products from Google in the future.

The Google Search for Google

Today I thought I’d do a really random search, so I searched for google on google.com. The question: Will Google rank Google.com #1 for a search for google? Or if they do rank themselves #1, why would they do that? If someone is already on google.com and they search for google, would they even click on the search result for google.com? It would make no sense. Confused yet? It only gets worse from here. And here are the results:

search results for google on google.com
search results for google on google.com

Well, there you go. Wait a minute! Google.org?! WTF? What’s that? Oh, it’s the ‘philanthropic arm’ of Google. Why would they do that? According to a linkdomain: query on Yahoo.com, google.com has about a jillion links (actually ~1.19B links) and google.org only has about 270k links. From what I’ve heard about Google’s ranking algorithm, the key to ranking is the quantity and quality of backlinks and anchor text. However, google.org does have a lot more content. Maybe that is the key element driving the top ranking. But seriously, what’s going on here? How does google.org pull a #1 ranking? That’s like me showing up to SkyBar with Megan Fox (btw, she just separated from that dude from 90210). I mean, I’m a nice guy and all, and I’m kinda philanthropic. Plus I once rescued orphaned nuns from a burning cathedral, but that’s Megan Fox. She’s like the #1 ranking these days. [And what about the fact that Google.com is advertising Google.com as the only result in the paid search ads? Is that logical? I guess it’s about as logical as searching for the term google on Google.com.]

Obviously, something is going on here. Most likely it has to do with the fact that if you’re on google.com and you’re searching for Google, you probably don’t want to see google.com as the #1 result. Google recognizes this. They know you’re probably looking for something else besides google.com. I think its for this reason that Google has purposefully placed google.org in the top spot. After all, according to Google’s Keyword Tool, google.org only got ~40,500 searches in January. Google.com got ~13.6M searches, and the term google got ~83.1M searches. It stands to reason that google.org needs some help with exposure, and Google probably wants to get more exposure for the philanthropic arm of their company. Aw shucks. How nice?

It appears Google likes to help out the tired, the hungry and the otherwise unable-to-rank-for-themselves type of sites. But is that philanthropy extended outside the family? Let’s find out. How about a search for the term search engine:

google search results for search engine
google search results for search engine

Okay. Now I’m going to throw up. Dogpile.com? Seriously? That’s a little too obvious because dogpile.com is a site that let’s you search for anything, and then it piles (get it?) all the results from G, Y, MSN/Live and Ask into one results page. [Once again, it’s very odd to go to google.com and then search for the term search engine, but it doesn’t change the fact that these results are not necessarily up to par with Google’s reputation for providing the most relevant search results for a given query.] I guess Dogpile is a reasonable #1 result, but the strange thing is that this site is not the most well-known search engine. Google is. The word google is used as a verb in our vernacular these days for heaven’s sake! Furthermore, according to Alexa, google.com is the #1 trafficed site in the world. Dogpile is #1,825. Typically, Google gets this kind of stuff right.

According to the Google Keyword Tool, the term search engine actually gets some searches:

Google Keyword Tool: search engine
Google Keyword Tool: search engine

See that? Google reports that the term search engine had over 2.4 million searches in January 2009. If I had an awesome anchor text analyzer, I’d totally compare the anchor text in links to google.com vs. the anchor text in links to dogpile.com. I bet that 99% of links to google.com have Google as the anchor text. Dogpile.com, on the other hand, might actually have more links with the term search engine as the anchor text. I can’t imagine that would be the case, but it’s certainly a possibility. It also occurs to me that Google shouldn’t even rank itself for this search. Why would a user click on the google.com result when they are already on Google. It really doesn’t make any sense. Kind of like this post. It’s filler. And I’ve dragged it out as long as I possibly can. Now I’m kind of hoping no one even sees this one.

Update: I just did another search for google, and after refreshing the search a couple of times, I saw some shifting in the organic results. Google.org, google.com, Google Reader, etc… – they were all changing places. WTF? I’m still confused. It’s pretty clear that this search result does not play by the typical algorithm rules. Me thinks Google has its hands in the search for google.