First Post!

Ok this was my very first post in February 2000. Talks about open source and ecommerce, although it's probably applicable to a wide variety of software types

The Importance of an Open Source Approach in an Ecommerce Deployment

This article is geared primarily to non-technical users,executives and management. It is not intended to be a technically rigorous paper but rather some general guidelines about electronic commerce and the importance of open source software in electronic commerce.

In order to understand the importance of open source software to your company's electronic commerce deployment, it is important to understand the structural changes and implications of the Internet and the new economy. The new Internet economy represents a strategic inflection point in computing and business. "Strategic Inflection Point" is a term used by Intel's CEO Andy Grove and refers to the periodic sea of changes the computer industry undergoes when the old rules no longer apply and a new paradigm begins to take hold. Intel has made it one of its missions to inculcate a culture that looks for these strategic inflection points so that they can avoid the fate of so many other companies in the computer industry clinging to the old paradigm. Minicomputer manufacturers are one example of companies missing the sea of change that personal computers brought on.

Clearly the Internet as well as related technologies represent a fundamental sea of change if there ever was one. In a number of industries we find web-enabled companies re-writing the rules of that industry. E-Commerce has become more than a buzz word, with CEOs launching ambitious projects to avoid being "Amazoned." These projects are important to the enterprise, often times the life of the company is at stake. The choices made in a company's approach and its software strategy often make the difference in the success or failure of the project.

When deploying these initiatives, rapid iterative approaches based on user feedback are a better way to maximize the functionality and profitability of the system. Too often, assumptions are made about the perceived or expected behaviors of users in regard to e-commerce systems. Too often, corporations take the same approach with a web site that they would with a brochure. Build it once and then leave it alone for a year. Expect to spend at least half of what you spent building the site on maintenance and changes in the first year.

With that said, it is important to design your site with flexibility and the ability to change in mind. Make sure that you build proper content management tools with the site and have a procedure for updating that content. If you produce a catalog, make sure that the content for the catalog can be re-purposed on the web easily.

Often the previous approach was to exhaustively select a vendor and then engage them in deploying the software. Given that most companies amortize this cost over a five-year period, selection of the vendor and technology is critical. In an industry that is only five years old, selecting the proper vendor and technology is very difficult.

Many companies make the choice, "Let's go with the safe solution." This often means the company's current vendor or the one perceived as the largest in the marketplace. While this appears to be the safe step, it most certainly is not the optimal solution for the company, especially since the previous safe option players may not be the best solution to deal with the new challenges presented by the inflection point. In fact, because these represent the previous "safe" vendors, they represent precisely the wrong assumptions in their approaches, methodologies and technologies.

The safe vendors, before the inflection point, are no longer the appropriate vendors after the inflection-point. Look for vendors that have developed around the inflection point.

Additionally, choosing the largest vendor may not make you safe. When selecting a vendor for e-commerce deployment, it is important to look beyond the life of the vendor. The CEO should assume that the vendor will not be with them in some way or another in five years. Software will change, the vendor may change focus, be bought out, or they may discontinue the product. Technology has proven time and time again, even safe vendors can go away in five years.

This means that in any e-commerce initiative, you embark on making sure you get the most important element - the source code with the proper documentation. This also might be a good time to re-evaluate your knowledge management procedures. If the vendor cannot supply the source code, look very carefully at them. What guarantees do you have with this vendor? Any vendor that cannot provide the source code to what will become the cthe core of your business initiative on the web shouldn't be taken seriously.

Remember that you are betting your business on this.
Think about your relationship with your current IT vendors. Is it everything you want it to be?

In assessing the impact of the Internet, it's important to keep in mind several things about the strategic shift it embodies.It's a networked economy. This means the company with the fastest, most secure network with the most network connections wins. Let's take a look at what this means in the context of IT deployments and e-commerce initiatives.

Fastest - This is fairly simple to understand. Any e-commerce system. needs to be fast. The server or servers shouldn't be underpowered. In fact, a good rule of thumb is to take your worst case estimates and multiple it by five. As far as speed of the network is concerned, it is a mistake to assume that your internal IT people can deploy enough bandwidth to satisfy your customers. The same can be said for regional ISPs and many of the national providers such as Earthlink or Mindspring. E-Commerce is the domain of national co-location and data center providers. The key in choosing one of these vendors is to select a vendor with a large set of peering arrangements. Peering arrangements can best be described as private connections with other ISPs such as America Online. This is a relatively small marketplace of providers, such as Exodus Communications, Frontier Global Center, Qwest, and Above.net. When choosing an e-commerce provider, make sure they have a data center in one of these facilities or have a relationship with one of these providers.

By speed we don't just mean the network and the servers.The time it takes to deploy to the site is critical. You need to get the site up yesterday. Make sure you can deploy in less than a quarter and continually make changes to the system based on user feedback.

Most network connections - For Amazon, one of their competitive edges has been their affiliate network. These connections to the Amazon site have a tangible benefit and represent one of the clear competitive advantages Amazon has. To date, it has generated 1,059,360 incoming links to the Amazon site. By contrast "competitor" Barnesandnoble.com has 63,254 incoming links to their site. Their sales figures represent a similar disparity (roughly 12 to 1). So when deploying an electronic commerce initiative, look for systems that maximize incoming links. One of the most cost-effective ways to do this is with an affiliate program such as Amazon's. Amazon developed their own software to track purchases, and currently they can boast a greater number of affiliate networks such as Clicktrade, BeFree, and Linkshare. These outside networks also increase deployment costs in addition to the software costs from the providers themselves. In the long term, it's better to use software with affiliate tracking built-in since this saves on integration costs and assures ownership of the affiliate relationship

Most secure network - Network security in the networked economy is of vital importance. As a recent break in at CD Universe showed, an online record store with three years of e-commerce experience, over 300,000 customer's credit card information combined with their personal information was lifted by Russian cracker. Undoubtedly it was the single greatest publicized e-commerce crack to date, and demonstrates the problem of not rigorously testing systems and performing regular security audits. It also highlights another problem - namely, closed source software. The piece of software which was cracked, was the closed source credit card authorization system. A recent study by Securityportal.com concluded that open source software was corrected much more quickly than closed source software from Microsoft and Sun Microsystems.

Having the source code allows you to have the source vetted, corrected and modified to fit your companies needs.

Without the source code, you just might find your business orphaned with no place to go for support or help. Not having the source code to your company's internal e-mail client isn't that important. Not having the source code to your e-commerce solution can prove devastating.

For example, in June 1999, ICat, a division of Intel and at that time a leading e-commerce solution, informed its entire customer base that ICat would no longer be supported. ICat also informed these customers that they would be bound by the terms of their license and no source code would be forthcoming.

For the IT managers and CEOs who took the "safe choice" with closed source, lost.

For this reason, access not only to source code, but the ability to change source code is paramount. Source code escrow schemes prevent you from proactively modifying that code for your site. For this reason, I recommend against source code escrow for most businesses.

For most businesses, this sounds like a much greater IT investment than they previously made and it probably is. The previous attitude of, "We'll just buy solution "X" won't knock the ball out of the park for e-commerce. It might get the site up and running, but don't expect to be on the cover of Fortune magazine any time soon for it. In the networked economy, the network itself becomes one of the most valuable assets

In the non-networked economy, if the company's e-mail system was down for a day, this represented an annoyance, but not a business-threatening situation. If this happened several times a year, it wouldn't represent a business threatening crisis, although it certainly it should be repaired. Keep in mind that being down three days a year still means a 99.17% uptime. Consider the reaction to this with an e-commerce web site such as Amazon or eBay - it significantly drove down their stock price!

The lesson here is clear, stability of the system and the ability to remain up under heavy loads is critical and probably 99.99 uptime should be the operating goal of an e-commerce initiative. Once again, I would recommend taking a look at software that is open source, has been peer reviewed and professionally developed. Ask about uptime as well as total cost of ownership. You are betting the future of your business on it.

If you are looking to purchase software, and want your internal IT to deploy it, ask them what platforms they feel most comfortable supporting with the stability, security and uptime requirements you need. More than likely, some of the Internet developed software and OS options represent the best choice since many of these grew up around natively in the strategic inflection point that is the Internet. More than likely, you will outsource this effort to a firm that specializes in e-commerce integration. If so, be sure to get source code for every component to be used in the site. It would hurt your business to be dependent on a closed source product without any recourse for support of that software. Additionally, ask about integration with existing ERP packages and business process. Usually, having the source code makes this task much easier.

Finally, be prepared to move at net speed. This means putting in place a system to handle customer inquiries in a timely fashion, and making sure fulfillment is rock solid. The net moves quickly and having the customer service systems in place to handle order volume and customer concerns is crucial. That is an area where traditional brick and mortar companies have an advantage over net companies. Customer support infrastructures are already in place. You need to make sure that your customer service center is properly integrated with your website and your existing infrastructure can support it.

Launching an ecommerce site can be a daunting task. By making the right decisions early, you can save yourself a lot of headache and be assured that it will positively impact your company's bottom line.

Classic Posts 2000 - Dot Com Graveyard

Ok here's another classic post from August 8, 2000- Talking about the coming dot com flameout. Of the companies mentioned in this post only Linkshare,(CJ& Befree have merged), Oracle, Intershop and Interact! are still in business - ouch

Fallout from the "Dot com" failout

While the graveyard grows with the graves of ill conceived ideas, funded by investors who should have know better. The problem of many the dot.coms wasn't that their ideas were bad, it was often that their structure for achieving these goals was completely top heavy. Would a local pet store hire 200 people, many of whom are top flight engineers?

That's roughly what petstore.com was doing while their sales were equal to a store in a busy metropolitian area. Is the inevitable failure of the site a surprise to anyone?

For web sites who ship physical goods, sustainable growth is the watch word, unless you were lucky enough to be Amazon who have parlayed their first mover advantage into a infrastructure advantage. One of Brave New Worlds first customers was and still is Interact! a great shop that sells new and used CD-Roms and DVDs. They have done so by selling product for more than they pay for it and providing sound customer service. Not a magic formula, but one that enabled their site to be profitable from the start. Maybe it's just me but isn't the point to make money? BTW they don't have an engineer on staff. Reselling software, like shipping dog food has a very low margin.

This column is about the other shoe about to drop on the e-commerce sector, namely the some of the companies that will be adversely affected by the correction that took place in the marketplace. This are largely businesses that have relied heavily on the dot coms. These includecompanies like Net Perceptions - an e-commerce personalization provider, Linkshare - which manages affiliate relationships for ecommerce companies. I haven't singled these companies out because they specifically are in trouble (although that may be the case, but rather the segments they represent are in trouble.)

Net Perceptions sells real time personalization software for e-commerce sites This is highly similar to Amazon recommendations feature and is really handy if you want to sell online. However the product carries an extremely high price tag (over six figures plus deployment and integration costs) and is facing increasing competition from the providers of e-commerce software (Intershop's Enfinity offers this ability as well as the open source and free All Commerceand Symphero. It's difficult to compete at that price point, especially when dot coms are dying left and right and open source software erodes your market. They face a more serious problem in that the bottom of the market is beginning to bundle these sorts of features and rightly so. Ecommerce products should provide a much higher level of reporting than they currently do as well as this sort of functionality.

Net Perceptions represents a first generation technology that is a great idea (personalized cross selling and personalization) that eventually gets wound into the core product. Auction software was the same way after Ebay's IPO. Everyone wanted to sell auction software at $55,000 - $125,000 per site. (Here's your $75,000 CD-Rom, want customization with that?) The problem was of course that the open source products killed the market for auction software. (That and there doesn't really need to be thousands of different auction sites. Network effects again.)

Outside of Oracle, are many companies going to be able to sell software at these price points? Not really since it effectively limits your customer bases to the Fortune 1,000 and Dot Com start-ups. So expect fallout among vendors who sell enterprise solutions without the marketshare of a Microsoft or Oracle, especially those selling software that is the Dot Com flavor of the month. (XML software providers come to mind ;-

Linkshare has a similar problem. They have provided an ASP model for managing affiliate networks. Sounds great until you look more closely. Linkshare, Commission Junction and BeFree all retain ownership of the affiliate, not the merchant, thus putting the merchant and the provider at odds. The pricing schemes vary as well. For example for Linkshare, there is a $5,000 set-up fee and 3% of the transaction, $1,000 monthly miniumum in fees and a three year contract term. This means at the end of your contract term, the merchant ends up without an affiliate network to call his own. All the marketing and promotion that the merchant adds value to the affiliate network which is owned by the ASP. Make sense? Not much, but out-sourcing the affiliate network made sense in the early days of electronic commerce when systems were patch work quilts hacked together with PERL and Unix pipes.

Brave New Worlds has always reccommended that the merchant retain ownership of the affiliate network, even with the added expense of administering the network. This sort of functionality is being bundled with open source e-commerce software (namely our own Symphero and commercial packages such as Intershop, so expect this business model to slowly fade. It will slowly fade due to merchant lockin. (Once you have traffic to your site coming from the affiliates, you are fairly rooted to that vendor. Uproot those affiliate roots and lose sales.) When you are a Dot com, speed to market and customer acquistion were the watchwords. Now it's a path to profitability and giving up 3% of your gross margin to an affiliate network makes no sense.

I have mentioned Intershop quite a few times. They too face this same competition from open source e-commerce packages such as Interchange, All Commerce and our package Symphero which provide similar functionality and the many other benefits that open source provides.

With the return to reality by the dot com deaths, we can expect more rational pricing by Dot com software and service providers as they readjust to the market realities - At the end of the day, you need to take in more money than you spent and you need to have a customer base which is profitable otherwise you are out of business.

Classic Posts from 2000

When I orginally started this blog four years ago I did everything by hand. As a result my blog has no posts for 2001 & 2002. Here's an original post from 2000 stuffed back in to Blogger.com. In this one I note how shutting down Napster was a strategic loss for the RIAA.

RIAA and Napster: A study in failed strategic thinking.

The recent injunction shutting down Napster has been hailed by the RIAA as a strategic victory for copyright holders everywhere and a victory of the major label system over the Internet. RIAA Senior Executive Vice President Cary Sherman
stated "the decision will pave the way for the future of online music." A truer statement has never been uttered, however, not in the way Mr. Sherman thinks. By forcing a federal injunction to close Napster, the RIAA has insured that music will remain freely downloadable and that the current definition of copyrights that sustains the current label system will eventually be eroded and replaced with something altogether different. By closing Napster, the RIAA has insured that their attempt to control the way music is distributed on the Internet will fail.

How is this possible? Given that they are on the eve of their greatest victory yet in asserting their control of copyright, it seems nonsensical. The RIAA could learn a lesson from the war on drugs and basic economics. Where a demand exists, a supply will arise to fill it. Clearly Napster tapped into the demand for digitally available music. Being the market leader initially, network effects were beginning to kick to insure Napster was going remain the leader for finding digital music online. (The much touted 78 million users by the end of the civil trial and the real reason for the injunction.)

Unfortunately for RIAA shutting down Napster is not going to change demand for the ability to digitally download music. Where
before that demand was focused on Napster due to the ease of finding the right music selection (Napster had the greatest number of users, thus the greatest selection of downloads which in turn reinforced their lead in users.
Ain't network effects grand?
Demand will now be forced into other outlets. Since the RIAA and the music industry as a whole has not agreed on a standard for secure download, nor have they made any replacement service for Napster available, this demand will undoubtedly flow into
GNUtella, which is a close cousin to Napster with some key differences. This creates some serious problems for the RIAA.

GNUtella is fair use.

Without a centralized server to facilitate search, Gnutella truly is one to one sharing of files. It's harder to find what you want but nonetheless, it still can be found. It really is sharing a copy with a friend. Never mind that it is a perfect digital copy of the original or so close that the listener cannot tell the difference. This sort of sharing falls under the traditional definition of "fair use" and I would hazard a guess that any court cases the RIAA brings to challenge that will fail.

GNUtella is anonymous.

No more centralized servers. No more registration forms. No customer
database and absolutely no centralized control. It would have been in the best interests of the RIAA for Napster to be the most successful music sharing service. With 78 million projected users, Napster would have had a near monopoly on music sharing. It would have been far easier to then develop a licensing system with those users, gathered in a central place. By shutting down Napster without a ready replacement, they insure that music sharing will be driven "underground" in the GNUtella network.

With music sharing now distributed with a true peer to peer client(which will undoubtedly improve over time as an open source product)the RIAA has lost the ability to influence the way music is delivered on the Internet. Since there is no central company in charge of GNUtella, lawsuits will be difficult, if not next to impossible to serve.

Why did the RIAA do this then? The file sharing peer to peer technologies represent a technological dislocation. As such they replace previous structures suddenly and out of nowhere. The structure they are replacing is the traditional music distribution network, which was governed largely by the distribution of CDs. Since the RIAA has a vested interest in the current distribution network (and indeed record labels made far more with CD sales than they ever did with vinyl sales),they are unable to cannibalize their current business model for the next - namely digital delivery. Napster as a venture capitalist backed start-up represented a "target" for them to sue. Unfortunately for them the elimination of Napster will significantly speed the growth anonymous decentralized file sharing systems.

What will be the final upshot of this transition? Since my crystal ball is in the shop, I won't hazard many guesses, but one thing is for certain. The RIAA and the labels they represent lost a lot more than they won. They lost any hope for control of digital downloads of music or any centralized body controlling digital downloads.

They won the battle but lost the war, clearly failing to understand the strategic effects of eliminating the largest of the centralized file sharing systems.

Third Moon around Earth.

According to the BBC we may have picked up a third natural satellite. It struck me as interesting that we don't know anything about the second one. From the article,

"Earth's second one is called Cruithne. It was discovered in 1986 and it takes a convoluted horseshoe path around our planet as it is tossed about by the Earth's and the Moon's gravity."

So naturally I wandered/wondered over to Google to find this article at Space.com. It's a great article describing the how asteroids temporarily become moons and after thousands of years wonder off again. From the article ,

"We found new dynamical channels through which free asteroids become temporarily moons of Earth and stay there from a few thousand years to several tens of thousands of years," said Fathi Namouni, one of the researchers, now at Princeton University.

"Eventually these same channels provide the moons with escape routes. So the main difference between the moon (we’ve always known) and ‘the new moons’ is that the latter are temporary -- they come and go, but they stay for a very long time before they leave."