Jason Kelly

Blogonomics: The Seeking Alpha Model

I've been a contributor to SA for a while now. It works for me because I make my money selling books and newsletters. Under that model, more exposure is better. As people meet me, see my work, follow it, and reach their own conclusions, I can pick up more paid readers who decide that I'm worth the price.

In a way, SA and other aggregation sites give me a free way to provide a trial period. Most newsletters include a cheap or free trial period, but there's work associated with providing it. The subscriber needs to be entered in a database, confirmed, and such. If potential subscribers can test me at SA and other places that show my free content, then by the time they decide to pay for my premium material I will have a greater chance of retaining them as customers. That's valuable.

Now, to the weakness of the SA model and how to reduce it.

Seeking Alpha is becoming Seeking Anybody.

As word gets out on SA and others, the original core group of high quality contributors gets diluted with a rush of less qualified contributors. If hundreds of new contributors are coming onboard SA every month, the signal/noise ratio is doomed to fall simply because:

> Most people in any profession are by definition average.

> Most people writing about finance online are not even in the profession, making them even less likely to be worth the reader's time.

> Material starts to duplicate, anyway, with now a hundred people making the same points about the latest Google headline instead of just the three people you know to be on top of the situation. How many me-too articles does it take to drown? Just the best presentation of each point is enough.

This deluge of contributions is serving to make SA less useful over time, not more. The quality filter needs to be tended to carefully or, as somebody else put it, SA will degenerate into another Yahoo Message board -- utterly useless.

To be clear, I'm not insulting the work of all part-time contributors nor claiming that all pros are great. I'm pointing out that any collection of material from thousands of people needs to be zealously filtered.

One idea is to create an algorithm that judges the accuracy of articles with a simple [buy, sell, hold] over the [short, medium, long] term categorization attached to the investment under discussion, and then track how that investment does against the S&P 500 or other appropriate benchmark during the time frame specified. Over time, a contributor would have an accuracy rating assigned to his or her history that could be viewed in a table like the one at CXO Advisory's Guru Grades:

www.cxoadvisory.com/gu.../

I have a feeling that the vast majority of contributors at SA would object strongly to such a system of accountability due to their accuracy hovering forever around 50%. Few want the world to know that their articles are no more useful than the nearest coin for a flip.

So, an alternative way to rate contributors would be to tally ratings assigned to their work by readers. SA is dipping a toe in this water with the "Did you find this article interesting?" question at the end of each article. That's not very sophisticated, though, and a system like Amazon's 5-star rating where you click on the number of stars that you think the article deserves would work better. I suggest ten stars for a finer scale.

As with DVDs on Netflix, a reader should be able to go back and change his or her rating of an article on SA later. That way, readers could judge whether the article was accurate or not in its forecast. If I write that Apple will hit $300 by Christmas, you don't know until Christmas whether I was right or not. You might rate the quality of my presentation 4 stars now, but want to move that up to 10 stars when Apple hits $300 at Christmas. (This is just an example.)

Similarly, SA should have a rating and categorization of commenters. They have zero accountability and, to be blunt, most commenters are morons. Offline, I've talked with other SA contributors who feel tarnished by the inane comments appended to the bottom of their articles.

If SA could separate the inane from the insightful and then prioritize them appropriately, that would improve the site's quality a lot.

With some further development by SA of rating systems, we can all help to put the best contributors, articles, commenters, and comments on top for easy research.

I hope these ideas help. All in all, I consider SA to be a fine effort and am happy to participate.