What is the exit strategy for low-traffic Internet Search Engines? This is a question I’ve been secretly wondering about for the past few months as I study the growing number of companies in this popular category. No Lack of Contenders…

BTW, your methodology is correct but your performance assumptions were way off. We’re in good shape

Thanks for acknowledging that my general assumptions were right (Arrington did not understand the rationale) eventhough your contention is that the cpms will be higher. I have not seen any ads yet, so I don’t think you’ve begun to sell the inventory. It is odd that you contend that advertising is a small piece of your revenue and subscriptions fees for data will generate significant revenue. You’ll find that data fees will not yield much and advertising will become your major source of revenue. I realize that you are not from the advertising/media industry and you do not have anyone on staff that’s focused on sales. When you get an ad pro, you’ll change your tune.

1) Our CTR is much, much higher than your estimate, higher than industry norms for links including Google AdSense.

2) Advertsiing is only a small piece of our model. We also receive subscription fees for data services and related article links we do for a number of partners. We’re doing fine revenue-wise and we have been super careful in how we spend money.

3) CPM’s are higher than $3 because the majority of our impressions take place in high quality branded site publisher sites that charge premium based cpm’s. Given that the user actually clicks on something and there is only one ad in the plug-in that they see (vs. several on a typical page that also competes with all the content for attention).

I find the dialog here helpful but also a bit critical in absence of all the data. Bottom line is that we’re trying to do something innovative, something hard – if it were easy, everyone including big players would be doing it. It doesn’t always work the way we want but overall, it’s producing high quality results and our partners are very supportive and happy.

As a result, we generated over 200 million links to blogs last month from a mainstream internet user base and hopefully that plays a small role in the blogosphere reaching a much broader audience.

My email is conrad@sphere.com if you have any inputs. Thanks.

relax its friday, Richard B

1) 1 billion LINKS
2) .005 ctr to get to the content widgets = 5M pages
4) Divide that number by 1000 (cost per thousand)
3) A generous $3 cpm
4) Equals $15K
6) 60% rev share equals $9K

Slow enough for you?

FYI : Your link goes to http://www.crunchnotes.com/ and it is down. I just got a page not found error

I thought it should be http://www.techcrunch.com

Cheers

Granted….my content ain’t what you would call “mainstream”, but still…..

but anyway, a billion page views at a $3 CPM is $3 million. If they are charging a CPM, then click rates are irrelevant to revenue production. And if you did apply a ctr, it would be to impressions, not revenue. you need to rethink all of your logic and math.

They state that they delivered 1.1 billion links last month. They include a 125×125 ad unit that can probably yield an average of a $3 cpm. Consider that a realistic ctr is .005. This produces a gross yield of $15K. With a rev share of 60%, the partner is getting $9K (and I was being generous with the averages). For Nytimes.com or WSJ.com, this is bupkis.

As far as the quality goes, I’ve found the results are usually pretty good and I like seeing what’s being said about an article/post, especially for the articles that don’t have comments (like The WSJ).

In principle, I’m delighted to see SphereIt and BlogRovr, although I uninstalled BlogRovr because of two reasons: The related content wasn’t particularly interesting and I didn’t like their interface (too intrusive).

BTW, good results with a lousy interface may be a bit trying on one’s nerves (think Factiva, no less), but lousy results with an iPhone-quality interface is useless.

Good UXD (user experience design) means both good UI design AND good app results (e.g., good search results, although latency is certainly an issue with search results as well).

BlogRovr failed, IMHO, on both results and their UI. But I’ll give them another try in six months or so. And I’ll give SphereIt a try in a few months, too. Both have the right idea, but need to focus on execution.