The only attraction on most robots, from my experience with store owners, is if there is a forum or blog associated with the ecommerce site. This is usually used to increase rankings for SEO. The main thing that Ecommerce faces is Fraud orders, where a person will flood a site with bogus orders. neither captcha or authentication will do anything to keep this from happening. IP addresses and countries are usually enough to spot such bogus orders. Also note that a lot of People don't want any information outside their address to be delivered to, so they will put in a bogus email address. The email verification will save a bounced emails when the notification emails are sent. On the other hand once a bounce email is returned . that my be, depending on the bounce, a way to verify the email address.
David E Jones sent the following on 9/29/2008 10:01 PM: > > On Sep 28, 2008, at 11:17 PM, Bruno Busco wrote: > >> 2) Since we have no captcha function I am worried about robots that could >> generate thousands of false accounts filling up the database. Having >> neither >> captcha and e-mail verification could expose the ecommerce too much, >> don't >> you think so? > > What is the worst case scenario for these? Extra registrations? > > There are many forms of denial-of-service attacks, I must admit I've > never heard of one that specifically targets automated account creation > on ecommerce sites. Usually automated account creation is used for sites > where public postings can be made, like free email and public forums and > such. > > Whatever the case, it's really not my decision unless I'm running an > ecommerce company myself. That's something I'd generally leave up to a > client to decide on. If someone did ask my opinion, I'd say the > inconvenience to a customer (and corresponding abandoned carts) may not > be worth it. Of course, it should also be considered that adding captcha > and/or email verification may not have any effect on customer conversion > rates and what what. If a company really wanted to know either way the > best approach would be to do random testing of using and not using each > across a large customer base. > > -David > > > >
