Content-delivery network Proximic, which has a unique contextual matching system, now has ads to sell that can help bloggers and others monetize their sites. The Munich Germany based start-up has signed deals to syndicate product listings from both eBay’s Shopping.com and Yahoo’s Shopping Network as contextual ads on other Websites. What other web sources claim the company is going to have more than 50 million product ad units in its data base coming in from both Yahoo! Shopping and eBay’s Shopping.com. Proximic estimates that Google, in contrast, has an inventory of about one million unique ads. Proximic’s ad network based on this massive inventory will launch at the end of January or early February 2008.
Web publishers are going to be offered with a way to place a widget on their sites, which Proximic is later going to use to serve ads on. Web site participating in the network are going to be later indexed and served up with contextually matching products as text ads along with contextually relevant content links. The ads and contextual links can also appear in a sidebar for anyone who has downloaded the Proximic Firefox add-on.
Proximic is neither matching context based on the keywords nor on the context itself. The company also says it doesn’t use semantic or statistical methodologies to understand the page’s meaning. “Semantic systems are not able to scale,” claims Proximic co-founder and CTO Thomas Nitsche. He also adds “If you hold more than one million documents, you run into a problem,”. Semantic search, he thinks, is too slow at this point for ad serving. Instead of keyword, semantic, or statistical approaches, Proximic uses proximity analysis to determine the page’s context. There is no much information publicly available as to how exactly it works, but from what we know and have read Proximic’s algorithm is translating each body of text into a pattern of characters that then becomes represented by a mathematical vector. Matches are done through traditional vector analysis. The company gives the following explanation:
We look at patterns of letters. We get a profile. The profile is a vector. We compare two vectors, and compute proximity by pattern distance. We can generate proximity between texts. The text can be one word, two words, 15 words, or a complete page.
We have read on other blogs claims of the sort Proximic is taking on Google AdSense, which has provoked us to give our 2 cents too and we think that such claims are, if anything, too boostful and not serious in any way and could be more harmful to the image of Proximic rather than brining anything like positive PR at the end of the day…
Ok, here we go with several potential problems, as we see them, Proximic is going to face and needs to deal with.
First off no site running Google AdSense is going to give up on its Google ads and earnings and replace them with an unknown start up that has little to no advertisers on its network. Why? Simply because Google does not allow your site or blog to run third party contextual ads (no matter what technology is used to match the context) on a page where their AdSense ad units run, which leaves little to no chance for Proximic’s contextual ads to stand off the ground any soon or at least not on sites that are currently Google AdSense publishers. There is clearly going to be a conflict of the two contextual ad units and Google is not going to be the one who will be dropped off by the web publishers.
If Proximic is indexing each page, as we read above, that becomes part of its network then they would also need 600,000 servers to get any closer to what Google is today (check the link for more info about the Google’s computation expenditures).
Revenue sharing with web publishers is not going to be very favorable for the web publishers who are going to participate in Proximic’s ad network after eBay, Yahoo! and Proximic itself all get their cut. We have read on Web that Proximic plans on giving participating websites 70 percent of any revenues after eBay and Yahoo! take their cut, which clearly leaves the publishers with a very small piece of the pie. On the other side, if they want to spread around Web, the way Google did, they have to pay web publishers serious money, lots of money, before even starting to think on competing with Google AdSense. Let’s put it that way: we see no way for Proximic to reach the payout Google achieved – $3.5B paid to web publishers in the first 3 quarters of 2007…
Proximic is not the first third party company to serve ad units from Yahoo! Shopping and eBay’s shopping.com. Even today you can sign up for Shopping.com or Yahoo! Shopping’s developer program and get listings up by next week. There are a number of other shopping engine syndication programs and most of them allow you to target to some extent. One of which is Shopzilla, among others, and Proximic is going to face fierce competition for the love of eBay and Yahoo!.
In tests, Nitsche says Proximic is seeing click-through rates as high as 1.5 percent, which is much greater than the 0.25 percent or less that is typical for an AdSense campaign. That’s simply not true. We have been Google AdSense publisher since 2004 and our average click-through ratio has always been way above 1.5%, so speaking for precise targeting we’ll have to wait and see what Proximic is capable of.
Proximic claims to be showing relevant results based on the content one is reading by gathering results from multiple sources, including Wikipedia but a weak point here is that they are not maintaining their own index massive, unlike Google. Just like with their third party sources of the information they deliver the same is with their product ads too, they are not theirs, which simply turns Proximic into an affiliate (middleman) company. Either way the company is vulnerable in case any of the third party information/ads providers leaves the game.
Proximic is a privately funded company based in Munich, Germany and Palo Alto, California. Investors include Wellington Partners and the Holtzbrinck Group, the publisher of numerous publications including Scientific American. The company is said to have 14 employees.
Other players on the contextual arena include Amazon, LinkedWords, Turn, Tumri, Shopzilla, Vibrant Media and Kontera and BlogRovr, among others.
LinkedWords is yet another, already fairly popular, company known to deal with the contextual aspect of Web and is known to be the pioneer of the in-text linked words approach, been around even before Amazon adopted this interesting approach for spreading its products among third party web sites’ context. It runs a massive contextual platform built upon tens of millions of English words and phrases, which web publishers are using to get contextually linked to each other through their platform by using in-text linked words, as the company’s name implies itself. (Disclosure: we are using LinkedWords)
Other ad companies that are known to have tried the same are Turn and Tumri, among others.