• a product or service that can be sold over the Internet
• the ability to handle remote transactions
• service provider support
• a strategic plan that incorporates, but is not dominated by, e-commerce

The last ingredient is the most important. Though for several years investors have been falling all over themselves to throw money at Internet-based businesses, our recent experience has shown us that such businesses fail as often as any other type, and they fail faster because of the speed of the Internet. We can't count on investment, or customer response, just because the word "Internet" appears in our business plans.

We make several mistakes in overestimating the value of our Internet presence:

• having online effects that add no value to our products or services, because they're "cool"
• having data handling capabilities that aren't justified by our level of business, such as full database backing for a small number of queries
• excessive e-mails sent to previous customers, trying to entice them back more often than their buying habits justify -- and sometimes being accused of "spamming" for it

The fact is that the Internet hasn't made the marketplace more competitive for small businesses unless the strategy taken by such businesses is suited to their markets. Too much money or effort applied to e-commerce can weigh a business down and keep it from doing what it does best. The failure of banner ads as a business model is a lesson history teaches us:

• ineffective if untargeted and/or untracked
• hits and clickthroughs are both less reliable as measures of Web effectiveness than registered users or viewers of multiple pages
• banner ad effectiveness can be enhanced (though with no guarantee of real success) by user interaction with sponsors and/or awarding prizes -- both of which cost money and resources

Here are some strategies for keeping e-commerce costs down:

• remotely-hosted scripts for surveys, password protection, site searches, etc. -- these can be found at (for instance)
  • beseen.com
  • hostedscripts.com
• third-party handling of message boards, etc. such as at
  • websidestory.com
  • greenspun.com
• third-party handling of mailing lists, such as can be found at listbot.com
• advertising via link trades instead of paid banner ads, such as the plan offered by linkexchange.com
• using graphics only as they add value to your product or service
• conducting sales transactions through e-mail (even with phone verification of credit cards) instead of through secure online connections

How transactions are handled online

It is possible to handle transactions completely online, but it's sufficiently complex that some small businesses don't go there. You need, as a minimum,

• to handle credit cards, or online credit, or both;
• to provide "shopping carts," or other similar means (usually written in some combination of JavaScript and CGI) for site visitors to order a product via an HTML form
• to have product information online -- enough to enable a customer to decide to order

This may force you to create a product database and have your Web site able to access and query the database. Many businesses with large product lines and substantial capital investment in e-commerce, such as amazon.com, have taken this approach.

You may also find it necessary to subscribe to a secure verification service, such as Cybercash, to overcome customer fears of stolen personal or credit card information.

How service providers are chosen

Besides cost, there are three fundamental questions that can be asked regarding a service provider hosting an e-commerce site:

1. What are my data handling capabilities?

   This is the cheapest part of e-commerce, and many users find it's the area in which most providers are strongest. They can all provide you with up-to-date devices and connectivity, and have various hosting plans that depend on

   • connection speed
   • reliability
   • redundancy
   • data security

2. Will my customers find my site convenient?

   This is the area of e-commerce that many of us struggle with. We sometimes can't know what customers will have problems with until we let them access our sites, any more than we can predict consumer product failures. As is the case with consumer products, it's a good idea to get as many points of view as possible before releasing an e-commerce site to the public.

   Some technical people have problems in this area because we have a tendency to see customer difficulties as personal criticisms of our work. Successful e-commerce takes a thick skin. Typical areas in which customers have problems are

   • availability (we've grown used to 24/7)
   • bandwidth, or data transfer rate (we want the best we can get for our connection to the Internet)
   • uninterrupted access (we don't want to be disconnected during a transaction)

3. Will my staff find site administration and business transactions convenient?

   When we take on a service provider, the concerns are not much different for business than for personal use:

   • our favorite operating systems
   • e-mail accounts/aliases/forwarding/autoresponders
   • available and reliable customer service
   • technical expertise
   • domain name hosting and seamless transfer
   • site management and/or Web development tools

References

Spector, R. amazon.com: Get Big Fast. NYC: Harper-Collins, 2000. ISBN 0-066-62041-4
interliant.com provides some interesting tutorials, particularly regarding how providers are chosen. (Their interest is obviously in selling their own service, so you must keep that in mind.)
The List contains information about nearly every service provider in the world. (But there are millions -- or so it seems.)

1. Make a realistic assessment of your needs before large capital investments.
2. Don't do business online just because "everyone else is." Have a strategic plan, and see e-commerce instead as one aspect of that plan.
3. If you push forward with e-commerce, be careful not to "cut your losses" too soon. Most companies selling online are still losing money based on direct online transactions!
4. Associate programs are among the best known and loved (if there is such a thing) means to drive traffic to a site. An associate program will encourage independent owners of Web sites to link to a major commercial site with products of interest to their visitors. The commercial site increases sales; the independent site increases its usefulness to its visitors (and the owner makes a small profit).
5. Banner Ads are ineffective if untargeted and/or untracked. Hits and clickthroughs are both less reliable as measures of Web effectiveness than (for instance) registered users or viewers of multiple pages. If you must depend on banner ads, their effectiveness can be enhanced by
   • interaction w/sponsors
   • prizes!