Check out the advertisement on the right side of the page:

Probably not the thing you want to see as you’re contemplating which refurb cell phone to purchase! They are using (as best as I can tell) the Newsfeed Maker service from IPD Group.

How would you solve this problem in general?  This is a tough one.   I don’t think you could rely on automated algorithms that rely on clickthrough volume, because truly awful news would (I think) get a relatively high CTR.   You could attempt to quantify each item based on some multidimensional matrix including such fuzzy criteria as “pleasantness” and “business-friendly”, but how to do that?  Keyword analysis?  If you have the word “cancer” in the title or blurb, you get -9 Pleasantness, and -14 Business Friendly?  How many keywords would you need to seed your keyword comparison database with?  How would you seed it?  Community information sharing portal?  Crawling UrbanDictionary.com for their keywords?

Tough problem.

Yesterday I linked up this blog with the Amazon.com Associates Program.  Registration was a snap, and I was very pleasantly surprised at how easy they make it to add product images and links.  For an example, see yesterday’s book review of the novel Hyperion, by Dan Simmons.

However, I don’t think I’ll be including Amazon links in most of the posts I do about books/music/products.  That extra couple minutes of building links/widgets just doesn’t pan out in a cost-benefit calculation, but more importantly, revenue generation is not the reason I blog in the first place.  Sure, I have dreams and goals and plans for building a successful software company, but Amazon.com 4% referral fees are not a key part of it.   If I think it’s useful for you, the reader, to see that extra information with a mouseover, I’ll include it, but probably not every time.

From Google UK.

[slideshare id=44638&doc=google-on-online-retailing-472&w=425]

Best Practice in Online Growth in Retailers

(h/t Fraser Mcculloch)

Here’s a short interview on the Practical E-Commerce website with Jay Steinfeld of blinds.com.

I loved the (one and only) comment at the end (emphasis mine):

This seems to me like a variant of the “Follow The Money” technique deployed in a rapidly evolving environment. It works quite well but it does require nimble thinking and a willingness to step into the unknown and experiment.

The merchandising side requires a very flexible ecommerce system, one that can be changed rapidly with minimal costs. On the marketing side it requires some careful risk management and superb data analysis. You try something and measure if things got better or worse, also called A-B Testing. Then you do it over and over and over, sometimes making incremental improvements while mostly abandoning the things that don’t work.

The most difficult aspects I see when employing this method is people have great difficulty dealing with the intensive data throughput, the rapidity of change, the impermanence of everything and an overload on their decision making apparatus. People look for “stable” solutions that don’t require such heavy demands on their mental processing. The constant bombardment of data and decisions frequently overloads the individual.

There are some very good ways to deal with a business at this level but all require a change in tradtional thinking. Not everyone can accomplish this.

I have a suspicion that Mr. Steinfeld is quite bright and has developed his own methods for handling and processing huge amouts of data and making an inordinate number of daily key decisions without overstressing his systems. It’d be very interesting to learn how he does it.

Posted by: Steve Strickland
Thursday, September 20, 2007