task

Guy Wins $1,000,000 For Pitching A Perfect Game…In A Videogame

Source: http://gizmodo.com/5532088/guy-wins-1000000-for-pitching-a-perfect-gamein-a-videogame

Guy Wins src=The folks at 2K Sports offered $1 million to the first person to pitch a perfect game in Major League Baseball 2K10—a supposedly difficult task. 24 hours after the game was released, they had to write a check.

Using Braves pitcher Kenshin Kawakami as his avatar, 24-year-old Alabama resident Wade McGilberry was able to complete his million dollar game in less than 90 minutes after returning home from work.

Great news for Wade because he recorded his attempts according to 2K Sports’ rules, but not so great news for them because as “insurance companies couldn’t possibly come up with the odds of throwing a perfect game, 2K Sports didn’t take out insurance and now will pay McGilberry a lump sum of $1 million out of its own pocket.” Oops. [CNBC via Sporting NewsThanks, Ezra Tenenbaum!]

Tags: , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

Wednesday, May 5th, 2010 news No Comments

How Google Crunches All That Data

Source: http://gizmodo.com/5495097/how-google-crunches-all-that-data

If data centers are the brains of an information company, then Google is one of the brainiest there is. Though always evolving, it is, fundamentally, in the business of knowing everything. Here are some of the ways it stays sharp.

For tackling massive amounts of data, the main weapon in Google’s arsenal is MapReduce, a system developed by the company itself. Whereas other frameworks require a thoroughly tagged and rigorously organized database, MapReduce breaks the process down into simple steps, allowing it to deal with any type of data, which it distributes across a legion of machines.

Looking at MapReduce in 2008, Wired imagined the task of determining word frequency in Google Books. As its name would suggest, the MapReduce magic comes from two main steps: mapping and reducing.

The first of these, the mapping, is where MapReduce is unique. A master computer evaluates the request and then divvies it up into smaller, more manageable “sub-problems,” which are assigned to other computers. These sub-problems, in turn, may be divided up even further, depending on the complexity of the data set. In our example, the entirety of Google Books would be split, say, by author (but more likely by the order in which they were scanned, or something like that) and distributed to the worker computers.

Then the data is saved. To maximize efficiency, it remains on the worker computers’ local hard drives, as opposed to being sent, the whole petabyte-scale mess of it, back to some central location. Then comes the second central step: reduction. Other worker machines are assigned specifically to the task of grabbing the data from the computers that crunched it and paring it down to a format suitable for solving the problem at hand. In the Google Books example, this second set of machines would reduce and compile the processed data into lists of individual words and the frequency with which they appeared across Google’s digital library.

The finished product of the MapReduce system is, as Wired says, a “data set about your data,” one that has been crafted specifically to answer the initial question. In this case, the new data set would let you query any word and see how often it appeared in Google Books.

MapReduce is one way in which Google manipulates its massive amounts of data, sorting and resorting it into different sets that reveal new meanings and have unique uses. But another Herculean task Google faces is dealing with data that’s not already on its machines. It’s one of the most daunting data sets of all: the internet.

Last month, Wired got a rare look at the “algorithm that rules the web,” and the gist of it is that there is no single, set algorithm. Rather, Google rules the internet by constantly refining its search technologies, charting new territories like social media and refining the ones in which users tread most often with personalized searches.

But of course it’s not just about matching the terms people search for to the web sites that contain them. Amit Singhal, a Google Search guru, explains, “you are not matching words; you are actually trying to match meaning.”

Words are a finite data set. And you don’t need an entire data center to store them—a dictionary does just fine. But meaning is perhaps the most profound data set humanity has ever produced, and it’s one we’re charged with managing every day. Our own mental MapReduce probes for intent and scans for context, informing how we respond to the world around us.

In a sense, Google’s memory may be better than any one individual’s, and complex frameworks like MapReduce ensure that it will only continue to outpace us in that respect. But in terms of the capacity to process meaning, in all of its nuance, any one person could outperform all the machines in the Googleplex. For now, anyway. [Wired, Wikipedia, and Wired]

Image credit CNET

Memory [Forever] is our week-long consideration of what it really means when our memories, encoded in bits, flow in a million directions, and might truly live forever.

Tags: , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

Wednesday, March 17th, 2010 news No Comments

the economics of advertising sucks, but it will suck a lot more soon

it’s a simple matter of supply and demand. Let’s do a thought exercise.

1.  eMarketer forecasts that retail e-commerce will grow roughly 10% per year for the next few years. This means that the total “pie” of people spending online will only grow by an average of 10% per year. Note that sales is (or should be) the goal of advertising. So that’s why we are looking at e-commerce sales and comparing it to online advertising because both are completed in the same medium and we can eliminate cross-media uncertainties and breakdown of tracking.

e-commerce

2. online advertising is still exploding with trillions of pageviews per month, thanks to social networks which throw off ungodly numbers of pageviews when people socialize with others. The Compete chart below shows the top social networks which rely on banner advertising (impression-based advertising) to make revenues. Notice that just Facebook and Myspace alone generate 115 BILLION pageviews a month. And if you consider that Facebook shows 3 ads per page, that would be 250+ BILLION impressions per month served by Facebook alone. Furthermore, the rate at which pageviews grow is 250% – 1,000% per year, depending on the site in question.

pageviews

3. In the online medium, we have end-to-end tracking from the advertising (banner impression) through to the sale (e-commerce). The banner is served (impressions); a percent of users click on it to go to a site (click through rate – CTR); a percent of those make their way through the site and end up completing a purchase online (conversion rate). Those users who are looking for something and who are considering buying something will be online searching and researching. Those are the ones who are likely to click on banner ads, compared to others who are online to do something else, like write email, socialize with friends, etc.  And if the purchase is their ultimate end-goal (to make a purchase) we have a farily reliable indicator of the growth in not only such interest but also the completion of the task — namely, e-commerce, which grows at 10%.

4. Now, if the number of people who will click grows that 10%, but the number of advertising impressions grows at a slow 250%, the ratio of clicks to impressions drops dramatically because the denominator is growing 25X faster than the numerator. Serving more ads simply will not get the amount of e-commerce to grow significantly faster. The point of diminishing returns has been reached and passed, so incremental ad impressions are ignored and useless. The number of people who will end up buying will not increase significantly faster. And given the tough economic climate the amount of sales may actually decline before it goes up again.

5. If we generalize this back to all retail commerce, it grows at an EVEN slower pace than ecommerce. When you compare this to the dramatic increase in ad impressions and the shift from traditional channels (TV, print, radio – whose impressions and audience sizes are dwindling) to online channels (portals, news sites, social networks – whose impressions and audience sizes are skyrocketing) again the ratio of sales to available advertising drops dramatically. This is a measure of the effectiveness of advertising (sales  divided by advertising spend). It was already small — it sucked — and it will get dramatically smaller soon — it’ll suck more soon.

A way to mitigate this “sucking” is to peg advertising expenditures on a success metric which is an indicator of user intent — cost per click — versus a traditional indicator of reach and frequency — ad impressions served — which from the above is NOT an indicator of consumers’ intent to purchase.  This way, advertisers only pay when someone clicks. Those “someones” click when they are looking for something and are more likely to complete a purchase than those who don’t click.

“CPC banner advertising” anyone?

Tags: , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

Sunday, March 15th, 2009 digital No Comments

Dr. Augustine Fou is Digital Consigliere to marketing executives, advising them on digital strategy and Unified Marketing(tm). Dr Fou has over 17 years of in-the-trenches, hands-on experience, which enables him to provide objective, in-depth assessments of their current marketing programs and recommendations for improving business impact and ROI using digital insights.

Augustine Fou portrait
http://twitter.com/acfou
Send Tips: tips@go-digital.net
Digital Strategy Consulting
Dr. Augustine Fou LinkedIn Bio
Digital Marketing Slideshares
The Grand Unified Theory of Marketing