December 13, 2008

Rent-sharing Matching Mechanism Visualization

Filed under: Portfolio — admin @ 12:56 pm

If in an effort to avoid arguments and hurt-feelings, you and your roommates want to find the most efficient way to divide rent and decide who gets which room… fret none. “Room assignment-rent division: A market approach” by Abdulkadiroglu, Sönmez, and Ünver details a simple, clever and elegant mechanism for resolving this dilemma. I’ve put together a bit of a visualization for those (like me) whom need a little aesthetic assistance in understanding how the mechanism works.

See it in action!

A Matching Mechanism for Allocating Attention to Tasks in Open Source Projects

Filed under: Research — admin @ 12:04 pm

In this paper we discuss using the often celebrated top-trading cycles mechanism to increase participation in open source projects. Open source projects is typically characterized as voluntary activity where a small set of participant perform the majority of tasks. In addition, open source projects can substantially reduce the cost for business adopting the software. If we can increase participation in open source projects, we can potentially increase quality, time-to-market, reduce operational costs for businesses, and overall increase welfare for society.

Download and read!

July 9, 2008

The glass is half full (or not)… maybe

Filed under: Left-field,Observations — admin @ 8:26 pm

A while back I stumbled on this podcast via Phil McKinney of Killer Innovations. Basically, it’s a very fun exercise in thinking creatively and generating different perspectives. We all know the classic saying.

An optimist says:

Awesome!!!11… The glass is half full.

A pessimist says:

Dang. The glass is half empty.

Phil suggests trying to think of different perspectives on this classic saying. My take—look at it from a Bayesian perspective.

The Bayesian optimist says:

Well I expected a full glass, but only got a half full glass :(

The Bayesian pessimist says:

I didn’t really expect anything at all, but was really surprised to get a glass half full :)

ooooh… that’s awkward.

btw. My first interpretation was this.

The Economist says:

Obviously, the water is scarce, and we need make the best use of it by determining who we should allocate it to. Of course, I’m economically rational and propose that I should get the glass of water.

May 25, 2008

What’s Strategically Interesting About Free

Filed under: Incentives,Observations,Strategy — admin @ 7:01 pm

Chris Anderson has started a lot of discusstion about free in “Free! Why $0.00 Is the Future of Business“.

Although, I didn’t see him mention it (and maybe it will be in his upcoming book “Free”), there’s something strategically special about free. Simply, it’s as if producer incentives are as they would be in perfect competition.

What do I mean? Let’s take, for example, Google Analytics (a free service offered by Google). We certainly can assume that Google offers the service for free because it indirectly increases the quality of advertising on search results. But offering the service for free has costs (labor, technical support, etc..). Let’s imagine that it wouldn’t take much cost to turn the service into a losing venture for Google.

Now, Google could charge a price for the service. They’ll lose market share, but it would more certainly make the service sustainable. BUT because Google has some market power, we could speculate that the incentives of the developers would be to lower the quality of the product in order to make more money. For instance, if the service charged per minute of use, we might expect the developers will create interfaces that are a bit more complicated and take more time to use. This would normally be tempered in perfect competition (but, hey, we don’t have an infinite set of web analytics producers). Google could pay users to use the service. Obviously, we can expect that would distort incentives such that users would only use the service to get a bit of money.

What’ left? Exactly, free. When offered for free, Google increases market share as much as possible (and increases the utility derived indirectly from search). Consumers consume as much as they want. More importantly, the developers of the service have a huge incentive to decrease the cost of production as much as possible. We might speculate that decreasing costs correlates with increasing quality. For instance, less demand for technical support (a cost) means that the service is of higher quality.

Simply, Google Analytics is competing against themselves! And competing against one’s self looks a whole lot like perfect competition.

May 23, 2008

Masters Thesis: A Framework of Attention as a Medium of Exchange

Filed under: Attention,Research — admin @ 8:12 pm


Research in the economics of attention generally approaches attention as a scarce resource and the design of systems to efficiently allocate that resource. This paper approaches the topic with a slight twist. Instead of designing systems to allocate attention, we look to designing systems where attention is a medium of exchange. This does not advocate that attention isn’t scarce, but, much like money, we can use attention to exchange for and allocate the goods that we really want (i.e. information goods). From this basic premise, this paper presents a conceptual framework and some insights into the nature of attention as a medium of exchange. The utility of the conceptual framework is derived from broad applicability, and this is accomplished in several ways. First, it adds coherence to an array of seemingly disparate research papers on the economics of attention. Second, existing systems currently using attention as medium of exchange can be described using the framework. More interestingly, though, the framework provides a common language to visualize new systems or augment existing systems. Lastly, the framework allows for decomposing systems in a way that is amenable to economic analysis. With that, we can compare the efficiency of differing systems or simply understand the implications and dynamics of a given system.

Download and Read!

And the presentation I gave using Google Docs

September 5, 2007

Another Recruiting Issue

Filed under: Insights,Signaling — admin @ 6:27 pm

Seth Godin discusses short-sighted recruiting techniques here. One observant commenter recognizes that the typical cover letter, interview, and attention to detail format might actually reveal desirable characteristics in a candidate. Even if those activities were just completely wasteful, this technique might be useful in separating low-ability individuals from high-ability individuals (i.e. screening).

On the other hand (and what I think Seth is driving at), those screening techniques are not very useful anymore. It may have worked back in the day, but now individuals can effortlessly create business networks (LinkedIn), resume profiles (Monster), and, in some cases, simply pay to have a good cover letter/resume written. Those old signals are just simply weak, and we need to look for more appropriate ones.

August 28, 2007

Government Correction

Filed under: Entrepreneurship,Markets in Everything — admin @ 6:49 pm

I love to read the various and amazing ideas entrepreneurs devise, but I can’t help being especially entertained when they create solutions that circumvent government action.

Here are two examples. I plan on collecting more.

Getting caught insurance pointed from Marginal Revolution
A private entrepreneur offers ticket insurance for public transportation as an alternative to actually buying a ticket. In the unlikely circumstance that you are caught, the insurer pays the fine.

A market for public-parking spaces
Innovator SpotScout creates a mobile service that allows individuals to purchase public-parking spaces from each other.

August 27, 2007

How to Not Hire Someone Via Craigslist

Filed under: Adverse Selection,Observations — admin @ 6:28 pm

Guy Kawasaki offers some good and humorous advice for poorly thought-out recruiting advertisements here How to Change the World: How to Not Hire Someone Via Craigslist.

Guy is a bit optimistic on outcomes. These kinds of advertisements end up doing much worse in that they create an adverse selection (I’ve actually seen it happen). Experts willing to slum it might apply. Not likely. And certainly, applicants that are honest about not having the specified qualifications will be overshadowed by those who are very skilled at being deceitful. So, in the end, you end up with individuals that aren’t as skilled as you thought, and you don’t even know it.

There is a bright side. Those type of applicants might actually be well matched to those organizations. In other words, deceitfulness, misrepresentation, and overall fronting might actually be valued characteristics (whether the organization realizes it or not).

July 17, 2007

Improv Presentation Application

Filed under: Portfolio — admin @ 5:18 pm

Presentation Mode Library Slideshow Splashscreen

This is an application I carried out from concept, design to development. I did have some help in the early stages from my good friend Dave Bowman. The software is intended to be used by Sales Technicians conducting proof of concepts at prospective customer sites. A proof of concept in enterprise software involves actually installing the software and taking screenshots which will be used in a future presentation.

Presentation! Why not use Powerpoint? The main finding that Sales Technicians discovered was that clients were more receptive (and likely to buy) when the presentation did not use Powerpoint. Possibly, not using Powerpoint is perceived as a high level of effort, and, thus signals a high quality of product. In addition, Powerpoint has a lot of features, most of which were not appropriate for this scenario. So, this is designed to be as simple as possible, removing any unnecessary features.

The application was created in Flash and uses Zinc to enable operating system level commands (e.g. “create a folder”, “rename directory”). Just like Powerpoint, a Sales Technician can use the same application to edit and present. As you can see, the edit mode is clearly denoted by the construction motif. The navigation on the left side allows for pages, directories, sub-directories, and so on… We conducted contextual research (i.e.actually watch people give presentations) and noticed that presenters and the audience wanted to go back to previous material or skip ahead. In other words, the preferred a non-linear format. That is not conducive to linear presentations, and hence the name Improv.

The application includes a library where all assets are managed. While typically screenshots are imported, any file type is allowed (e.g. Documents, Videos, Scripts, and, yes, Powerpoint files). A slideshow can be assembled from the image assets. Think of this as creating a playlist in iTunes. Then a slideshow (or other document) can be linked to a navigation item.

This is really the beginning of a vision that I have where off-line presentations, online ads, and online user-driven presentations are connected together, and lots of metrics are captured. I call it “Integrated Mechanisms for Presentations using Real-time Online Vehicles”. Okay, I admit it, that took me a bit to coerce that name into Improv. But it’s not terribly far off.

July 16, 2007

Recruiting Advertisements

Filed under: Portfolio,Signaling — admin @ 5:59 pm

screen_mensa_know-it.jpg screen_mensa_airplane.jpg

“Know-it-all” Recruiting Ad PDF
“Airplane” Recruiting Ad PDF

These are some recruiting advertisements that I designed. They were printed in Mensa magazine as well as in the Detroit News classified section. They were quite successful and have been reprinted many times.

This is probably one of my first attempts to think analytically using economics when designing. They are based on two concepts: signaling and screening. The objective was to induce innovative and creative IT-focused individuals to apply for employment.

You might think that simply saying “Are you innovative? Apply with us.” would suffice as an advertisement. Unfortunately, the recipient cannot tell for certain our commitment to innovation (i.e. an asymmetric information problem). Any company can say they are looking for innovative people (and most probably do). It used to be that simply creating an advertisement was a strong signal. It was costly and is a non-redeployable asset, and a rational organization would not incur such costs if they were not serious. Information technology is changing that landscape, though, and what used to be strong signals are now weak. In other words, anyone can make an advertisement. Signaling now exists in the content of the ad.

To accomplish this we delved into the concerns of innovative individuals. Specifically, they could be at odds with the social system they belong to. They think outside the box and what appears to be unproductive behavior is actually genius at work. They experienced that at school, and realistically could experience that in work as well. By going through this effort to understand what makes innovative individuals, we can credible say that we understand and working for us will be conducive to you.

The second issue is screening. An ad that lacks credibility might create an adverse selection in that truly innovative individuals are not convinced and simply apply to opportunities that appear better. This ad is fair but not exceptional on this issue. I was inspired by one of Google’s ads. It featured a vending machine of items with a puzzle. Simply solve the puzzle and your application is bumped up to the front of the queue. Clever. Those who don’t do well on puzzles will more than likely not apply (which is what Google wants). It’s just quite amazing to think that an ad can be designed such that only the individuals you to want apply. In addition, this lessens the burden of sorting by the Recruiting department.

Next Page »