Supernatural Agency: Online business tales » e-commerce

The beginning of the end for affiliate marketing?

martin — Fri, 30 May 2008 20:35:31 +0000

We’ve had three major brands for whom we are affiliates end their relationships with us because we are located in NY State. Never mind the fact that our servers are somewhere on the West Coast and that these retailers have no brick and mortar presence in NY. Our totally dysfunctional and short-sighted state government decided that out of state e-commerce companies selling on the web in NY should pay our insane (8.5%) sales tax even though the law for mail order companies states that they don’t have to charge the tax as long as they have no physical presence in the state.

Apparently a website that does not even sell the products and does not physicaly exist here is now a cause for ordering merchants to pay sales tax. So, what about the millions of paper mail order catalogs sent into our mailboxes from out of state companies- aren’t they a physical presence? And since when are online media that advertise services on a commission basis (very common in print, radio and broadcast) a cause for claiming a physical presence?

I do not think this will hold up in the courts as it is a poorly though-out law filled with loopholes. In the meantime don’t plan on building an affiliate marketing business in NY.

Update: Two more merchants have sent notices specifically detailing that the NY law was the reason they were removing us, hopefully temporarily (as they said). Interestingly, Linkshare followed with a generic notice saying we had been removed for unstated reasons- I think they’re worried about losing a lot of other merchants if they get wind of this.

So, if Google has PayPerAction (PPA) programs in which they get paid when a conversion takes place should they be paying taxes? They have a big physical presence in NY.

This is the lid coming off a big can of worms…

If you want to make money with a blog pick your subject matter carefully

martin — Tue, 15 Apr 2008 15:27:53 +0000

Even though we put some Adsense units on this site we never had any expectations for them beyond a few clicks here and there. That’s because readers don’t come to this site with a purchase or product research in mind. The subject matter (Internet marketing and business) isn’t closely associated with buying.

Gawker Media, the big blog network, announced that it was selling off a few assets including Wonkette, a very popular inside Washington political blog that had over 5 million pageviews last month. With that kind of traffic you’d think they’d be crazy to dump an otherwise successful site. My guess is, that with a downturn economy, a site like Wonkette which never had a strong advertising model- readers go there for opinions and gossip, not shopping, was probably seeing a big drop in ad revenues. I’m guessing their CPMs were in the toilet despite the high traffic.

To make money with a site you have to choose your subject matter carefully. Is the information you’re providing something people seek during a buying process? Are the types of purchases associated with the site Internet e-commerce friendly? Things like books, jewelry, gadgets, courses and other high value, easily shipped, high margin products work best. Things that are heavy, cheap and have low margins like pet supplies, large appliances and groceries don’t cut it.

Is the subject matter information-intensive? If the buying decision is easy you can’t add enough value via a web site or blog. Look at the ads on general news, opinion and lifestyle blogs/sites- they are often big brand ads for cars, insurance and other generic subjects that are neutral in relevance to the content. These kinds of ads generate almost nothing in CPMs (measured in cents per thousand impressions) and contribute to banner blindness.

Getting a formula that delivers higher revenues requires something a lot of Internet media entrepreneurs don’t want to do: hard work. I’m in the process of inserting relevant affiliate text links into the hundreds of articles on one of our kitchen design sites. We have display affiliate ads but they draw very poorly. The text links are carefully chosen for very high relevance but it’s a lot of work:

Determine what the reader of the article is most likely to be interested in at that moment. I have an article that recommends getting material samples when planning a new kitchen. The reader wants to know where they can get these.
I go through the various affiliate programs and find a vendor that not only offers samples but pays out for sample requests, in this case $15 for a sample order (because they know that if you have a flooring sample you like you’re likely to order that specific flooring- resulting in a big sale). This takes a lot of time to research.
I build a text link with the affiliate code that includes the appropriate anchor text. Free Flooring Samples from XYZ, for example
I choose the appropriate place the article and insert the link text inline with the content. I don’t hide that it is a link (different color) nor do I hide the brand because I want it clear to the search police that I’m offering legitimate relevant content even if it is commercial. And I want the reader to know that I considered their interest, time and needs when selecting this revenue model.

This is a lot of picky busy work but when I’m done I’ll have an information site that is optimized for revenue-generation in a way that the consumer appreciates. And it is sustainable because once the work is done the site is self-sufficient. On to the next one…

The Deep Web: Google now crawls all possible results of entering data in forms

martin — Sat, 12 Apr 2008 15:12:35 +0000

This is pretty crazy.

If you place a form on your web site Google can now enter any possible combination of information, radio button choices, checkboxes, etc., and index the results that come back for the form. They call this accessing the ‘Deep Web’, that part of the web that, until now, could only be accessed by a human entering information. It is deep in the sense that they now index all the possible underlying data in an online database that is not protected by a security layer (at least that’s how I read it).

Implications? Optimize those forms and your results pages and make sure they return relevant results or you might get purged.

And, BTW, dump your Microsoft and Yahoo stock- they are not even on the same planet with these guys.

Parsing the Hecta Media domains purchase

martin — Wed, 19 Mar 2008 20:47:09 +0000

I’ve been following Hecta since they went public in London last year and I spoke with Clark Landry regarding their plans a few months ago. Now, they’ve made a few big purchases of domain portfolios, notably a 60,000 name group bought for $1.42 million US. or about $24 per domain. They project a gross annual revenue of about $750,000. from these domains or an average of $12.50 annually per domain.

Clark told me their business plan was to look for domains that were underdeveloped due to lack of cash or bandwidth to do so. This purchase, even without any idea what domains it includes, looks like a no-brainer. I’m guessing the owner was looking to cash out but selling for 2x revenue doesn’t look that great to me. As has been pointed out by others, domain registrations are going to eat about one third of the revenue each year. I’d guess Hecta will cherry pick the good domains and offload the rest, especially if they are going the development route. It will be interesting to learn more about their approach.

A sale like this should be an increasingly rare event in the domain world going forward. Until recently domainers have looked like land speculators, grabbing swathes of swampy land, marginally improving it and selling off parcels to the next wannabe magnates. As the domain world wakes up to the fact that being a developer is where the big money is they’re going to be keeping the choice pieces for themselves and building luxury condo towers on them. Instead of a nice 300% return (like the Hecta deal), you’re going to be looking at much bigger returns on much smaller portfolios, IMHO.

If this is the case we could see a flattening of the resale market for marginal domain names as big portfolio holders dump the crap. While we don’t sell much, I do watch the auctions, Sedo and Afternic, etc., and it looks like this is already taking place. Personally I’d rather have one strong dot com domain than 1000 oddball ones. The definition of a ’strong’ domain for me is one with an obvious development roadmap and a high potential for monetization.