A tiny 14-man startup in Munich called Proximic wants to give Google a run for its money. But it is not going after search. It is coming from behind with an attack on AdSense. Proximic has signed deals to syndicate product listings from both eBay’s Shopping.com and Yahoo’s Shopping Network as contextual ads on other Websites.
This is the first time that either Yahoo or eBay has syndicated these listings. In one fell swoop, Proximic will gain an ad inventory of 50 million ads (20 million from each eBay and Yahoo, and 10 million from other sources). 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.
Prospective Web publishers will be able to place an ad widget on their sites. Proximic will index the sites and serve up contextually matching products as text ads along with contextually relevant content links (see demo screen shots below). The ads and contextual Web links will also appear in a sidebar for anyone who has downloaded the Proximic Firefox add-on.
So if you are reading about Mideast peace efforts, books about the Mideast might appear in the ad widget. If you are reading about the Mozilla Foundation, you might get an ad for Firefox track jackets from the Mozilla store, along with a link to Twitter message about Mozilla hiring the guys from Humanized. (This is what actually came up in the Firefox sidebar when I tried it. Proximic also populates its widget results with the content links from 900,000 RSS feeds it has indexed and the top 500 or so Websites.)
What makes Proximic different is that it does not come up with contextual matches based on keyword search like Google or Yahoo would. It doesn’t use semantic or statistical methods either to figure out what a page is about. “Semantic systems are not able to scale,” sniffs Proximic co-founder and CTO Thomas Nitsche, a former computer chess champion. “If you hold more than one million documents, you run into a problem,” he concludes. Semantic search, he thinks, is too slow at this point for ad serving.
Instead of keyword, semantic, or statistical approaches, Proximic uses proximity analysis. Nitsche is vague about exactly how it works, but it boils down Proximic’s algorithm 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. Or, as Nitsche explains:
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.
Using this method, Proximic can also create matches between product listings and Web pages, thus opening up what is now an inventory of product search results to the world of contextual advertising. 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. Of course, Proximic has to split any ad revenues it makes with Yahoo and eBay on one side and the Web publishers on the other. Proximic plans on giving participating Websites 70 percent of any revenues after eBay and Yahoo take their cut, leaving it with a very small piece of the pie. The only way to make up for it is by generating much higher click-through rates (by improving the relevance of the ads).
It is an ambitious undertaking for a German startup which has raised only 3 million Euros from Wellington Partners in 2006. But what Proximic is trying to do is combine contextual search advertising with affiliate marketing. By analyzing the patterns of characters on a page, it is creating a machine intelligence of sorts. The nice thing is that Proximic does not need to leave cookies on anyone’s browser or track you across the Web. It makes its own judgements about what is contextually relevant based on what you are reading.
If Proximic is successful in matching relevant products, imagine what it could do as a general-purpose search engine? But Nitsche knows better than to try to take Google head-on in search. He is happy, for now, focusing on the advertising end.
“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.”
And Nitsche for 1 million dollars, what is the N ORDER of that?
Nitsche: excuse me?
“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.”
For the market that I know about, all the pattern recognition analysis software for financial markets have been utter garbage in terms of consistent results, though the fin market works under different set of conditions, one of which is trying to predict the future.
Beyond that, I hope these guys succeed.
There is a definite demand for useful competition to Adsense and 70 percent payout to publishers sounds competitive.
Where you say ‘Prospective Advertisers’ I think you mean ‘Publishers’. (3rd para).
That had me confused.
By “prospective advertisers” I understand you mean “prospective publishers”?
Their website (from a publisher point of view) has to be one of the most confusing around.
For one, the examples page (screenshots) on how the widget works make no sense whatsoever, at least to me.
This is an insanely good idea.
Hopefully it does better than Yahoo’s publisher network.
So, the contextual method they’re using is the same vector space model[1] that’s been used for information retrieval for about 40 years?
Sweet.
This isn’t to say that the VSM isn’t useful or valid for this purpose (it produces fine results in many cases), but hyping technology based on it seems sort of bogus, although possibly it’s TC’s take on it making it come across that way.
The company may do fine regardless, anyway. There are plenty of advertising firms doing fine without Google’s expertise at IR concerns.
1: http://en.wikipedia.org/wiki/Vector_space_model
This sounds like auctionads + real time information retrieval.
I tried to signup for their widget but they appear to want me to provide them with free advertising?
The matching quality absolutely rocks!
The publisher program seems to start soon - that’s what they say on their web site. I hope they really mean “soon”…
I think his words is just an excuse for what they use. as well, what if those product listing providers boot them tomorrow to cut a middleman?
Being an eBay, yahoo shopping or shopping.com partner boils down to signing a direct contract or via Commission Junction with exactly the same inventory.
I don’t understand why Erick writes “first time” or “sign deal”. On another note, CPA is a source of financial loss for companies such as chitika, widgetbuck, auctionads etc…because the CPA must be transformed into CPC to pay the publishers with disproportionate ratios, to the disadvantage of these companies.
Its bullshit, and hype for nothing. 10 companies already offer the same service in CPA and in contextual mode.
Honestly the product has the same idea as Sphere it is just tackling a different market. However will readers download the Firefox Add-on? Which will be a huge turn off.
Because of the fact that you are downloading a widget that will offer ads..
No revenue model for publishers, examples on site do not show the product in action (don’t appear to show product at all!), completely confusing messaging and vague technology.
Vaporware- I expect a little more due diligence from Techcrunch.
They don’t deserve coverage until it is possible to actually monetize the service, otherwise what’s the point?
What Pascal said. Taking a syndicated Shopping.com feed is the starting point for many, many widget and AdSense-alternative firms that’re already out there.
Chitika sounded very similar back in 2005, and the hype others created around it (in terms of CTR mainly) led to lots of AdSense publishers trying it out, only to be severely disappointed with down-to-earth CTR’s and much lower effective CPC’s than they thought they’d have.
WIFM?
I don’t see anywhere to sign up so I get paid…
How come no one mentioned Amazon Omakaze ad units yet?
This is exactly the same idea - and it’s a good idea in theory - but it’s also fairly well established that it’s not a killer contextual ad unit by any means - in short it’s not worth running in lieu of Adsense.
Nothing to see here . . .
The overwhelming leader in the contextual space is Google. The #2 player is actually AOL/Quigo, not Yahoo. Can anyone name a major web publisher that exclusively uses Yahoo other than Yahoo.com? Anywho — this idea is being tried by Turn and Tumri. I’m very skeptical and believe Erick wrote about this because it’s something to write about and not based on merit. Sorry.
Erick,
Thank you for the coverage. Just to to clarify with regards to the deal with Shopping.com.
(1) Our relationship is with Shopping.com, not eBay, the parent company.
(2) As for the syndication with Shopping.com, the infrastructure we collaborate on is not a unique service to us. The main difference, however, is that we bring contextual relevancy to Shopping.com’s catalog data. When including Shopping.com in our mentioning to be syndicating its catalog for the first time, technically that was incorrect.
Best
Philipp
Philipp Pieper
- CEO Proximic -
If Proximic is indexing each page, as I 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 above for more info about the Google’s computation expenditures).
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: I see no way for Proximic to reach the payout Google achieved - $3.5B paid to web publishers in the first 3 quarters of 2007…
So, taking on AdSense claims and titles are, if anything, too boostful and not serious in my view, which is more harmful to the image of the company rather than positive PR…
My 2 cents
Erik - You really need to correct your post. As the CEO of Proximic points out, there are plenty of other people syndicating Shopping.com and Yahoo Shopping listings. This line: “This is the first time that either Yahoo or eBay has syndicated these listings.” is false. You can sign up for Shopping.com or Yahoo! Shopping’s developer program today and get listings up by next week.
As for the targeting, we’ll have to wait and see. Most of the shopping engine syndication programs allow you to target to some extent. Look at this search on Mahalo. I searched for iphone and got product ads from Shopzilla which seem pretty targeted to me: http://www.calacanis.com/2007/...../#comments
@Robert: It seems that there is more heart blood to it than a straightforward implementation of the vector model - which would work on “terms”, i.e. keywords, which by the explanation is not what they use.
They retrieve a “pattern of letters” which is *probably* compiled into a BIT “vector”…. More on the speculatory part:
*proximity* search within patterns + the guy is a former *computer chess* hero - which leaves me with a hunch that they have a genious algorithm inspired by pattern matching of moves on a chess board…
In any case people who spice up the business as ususal are just admirable!
Rock on