Check out TheFind Beta. Congrats to Siva and his team!
I’m excited about the concept of TheFind as comprehensiveness has been a pet peeve of mine since I first talked to Sergei Burkov at Dulance. One problem with Shopzilla, Shopping.com, NexTag, etc. is that these guys work off of a CPC model. To play, the merchant has to pay. Unfortunately, this automatically cuts down on the comprehensiveness of these engines as merchants are forced to remove low margin/low cost items from their feeds.
Shopwiki, Windows Product Search Live, and TheFind don’t charge for inclusion and crawl the web for products as opposed to working predominantly through feeds.
However, playing devils advocate for a moment (as I’m sure my friends at Shopzilla, Shopping.com, and PriceGrabber will have plenty of comments for me today) comprehensiveness isn’t everything. Relevance is another BIG piece of the puzzle.
Testing out TheFind…
A search for Polo Jeans didn’t exactly bring up what I expected. And the first result for red sweater, a $525 mens designer sweater from Bluefly which is ‘not currently available’, didn’t really delight me. Oh, and a search for nintendo wii returned a Super Mario Galaxy Badge before the Wii AND TheFind returned a price of $249 from Pandora’s Cube when Pandora’s Cube indicated a price of $599. 0 for 3. If Jose Reyes of the Mets had played like that last night, I wouldn’t have been able to function this morning.
So take a look…the problems you find will be addressed. TheFind is well funded and has a smart team on board…and it doesn’t hurt that the site is searching over 500,000 stores and finding over 150,000,000 products.
Hi Brian, are you saying that comprehensiveness and relevence are in opposition? If so, could you please explain why you think the CPC model is conducive to providing users relevent search results? Or am I misunderstanding your point? Thanks.
Well Brian you might call a crawler “second generation” except of course that Shopping.com had a crawler initially (You should remember something like $135 million in VC went into shopping.com). What they found was that many merchants unable to generate an XML feed were also lower quality merchants. In this regard, they are using the CPC model as quality indicator. The idea that you provide more stores and this is a “better” option isn’t necessarily true.
TheFind.com
A new comparison shopping site TheFind.com launched today as an all-encompassing shopping search site. The Company has garnered a bit of buzz, with an interesting write-up at GigaOm and Comparison Engines. Although I like the principle, it’s r…
Totally irresponsible / uninformed journalism.
First of all, its simply impossible that there are 500,000 retail stores that offer prices in US $s on the internet. The upper bound is something 50 stores.
Second of all, their “grid” is the anti-definition of consumer friendly. Part of the reason to conduct a comprehensive search is to understand pricing / vendor trade-offs in a logical fashion. Their grid doesnt sort by price, and somehow incorporates no-name merchants such as “smalldog” in its top set of results (even though they are relatively expensive). Its totally unclear how the information is ordered / presented.
Finally (and only because Im not interested enough to debunk the rest of this junk coverage) – because I could rant and rave all day long about how searching by things like color is the most rediculous “feature” ever:
“So take a look…the problems you find will be addressed.”
Ha, yeah right. Extraction at scale and automated normalization are two of the hardest problems being addressed on the internet. They wont just be addressed – it doesnt work that way in real life.
well, in response to EarlyMiser’s comments about Shopping.com, I just want to clarify a few things about Shopping.com;
1. That company never had as its core business practice crawling websites. It would crawl stores that had no feeds, but it never encouraged it.
2. It never raised $135 million. That I can assure you.
I am sorry that you can “assure me” that shopping.com didn’t raise $135 million over various rounds of venture capital. I can assure you they did and they lost buckets of money on marketing.
In 2001 alone they spent something 70 million in marketing. I assume you currently work at shopping.com or at least you imply it. Remember Shopping.com started as Dealtime. Their third round of funding in 2000 was $50 Million. This implies of course two previous smaller rounds of funding.
Their second round of funding was $20 Million in 1999.
Second Round
http://www.findarticles.com/p/articles/mi_m0CGN/is_3748/ai_55770999
Third Round
http://news.com.com/Short+Take+DealTime+raise+50+million/2110-1017_3-236968.html
So just by trolling two of their rounds totaling $70 million. I am still looking for their initial round and I know they had a fourth round.
I like shopping.com I have been a partner with them for a while now. When I find the other rounds I will post them.
Here’s a company profile from 7 years ago.
http://www.vault.com/jobs-company/Shopping.com.html
Founded by two Israeli high-tech entrepreneurs in December 1997, Shopping.com, formerly DealTime.com, went live with its online shopping comparison service in June 1999, backed by venture capital funding from firms like Israel Seed Partners and Odeon Capital Partners, as well as by strategic and financial partners including AOL, Time Warner and Bank of America. The site uses an automated shopping “bot” to comb hundreds of web sites — including online merchants, auctions, classifieds and group buying sites — to find the lowest prices on thousands of items.
As you can see the shopping bot was an integral part of the company.
After much searching I suspect I got the figure for their funding wrong. It’s certainly between 80 – 110 million but the 135 million dollar figure is also the amount of venture capital that boo.com got. Part of the problem is that the first round was by a set of Israeli VCs so the figures are harder to come by. I will keep plugging away though
Please note – I am not including the $44 million in VC that eOpinions recieved in the total. Let’s call their first round of funding $5 million – that puts us at a total of $75 million in known funding. Add in eOpinions and now we are at $119 Million so the $135 Million dollar figure doesn’t seem so implausible now. Still I suspect I conflated the two and my apologies for the confusion.
Maze designer I meant to reply to you earlier. When I was talking about CPC models as a quality indicator, I was speaking generally. It’s not that comprehensiveness is the issue – it’s the fact that there are not 150,000,000 different products that can be effectively comparison shopped on. Rare products (such as collectibles, antiques etc) by definition cannot be comparison shopped up like a new dvd player. Comparison shopping works effectively on goods that are already commodities. People point out that the long tale will benefit smaller suppliers in the market when it will have the opposite effect. Only the larger suppliers will be able to carry the breadth of inventory to supply the long tale. The long tale far benefits suppliers able to achieve economies of scale.
My point about comprehensiveness is this – for goods you will want to comparison shop on – vendors comfortable with the CPC model are more than likely going to be high quality vendors. The idea that out there somewhere is a retailer you have NEVER heard of who is increasingly unlikely.
TheFind.com is not a price comparisson site, but a shopping search engine. Taking datafeeds and charging CPCs from larger merchants is fine for electronics, computers, etc. But for hard to find items, apparel, and home goods, there are many specialty retailers that do not list on shopping.com et al. By including these stores, thefind, shopwiki, and the other crawler engines create a more comprehensive shopping experience. Don’t bucket them in with the “big” guys – they are a different animal.
[...] TheFind Beta Launches (Comparison Engines) October 19, 2006 [...]