The rise of edge content continues – iNods, which launched today, is the most recent example.
iNods aggregates blog and other edge review content and displays it in a centralized way for users. Unlike similar services kritX and blogcritics, which aggregate general review information contained on blogs (and are excellent), iNods is completely focused on shopping.
Like the others, iNods has realized that there is a massive amount of review information out there in the blogosphere. And more is being created every day. They are harvesting that information, making it searchable, and driving highly targeted traffic right back to the blogs to read that content.
They also require blog registration to include content, although they do already have fairly deep content and so I assume they have added some sources without their registration. They say that they will support microformats in the future, but at this point it is not a requirement. This means bloggers do not need to format their data in any particular way to have it included on iNods.
I’d like to see these services expand their technology so as not to require blogger registration or microformats (kritX, blogcritics and iNods all require one or both of these). This is a harder problem to solve because they need to go through every blog post and figure out what is relevant to their service.
Structured blogging initiatives and/or blogger registration certainly make it easier for services like this to recognize and gather the specific edge content they are looking for, I do not believe it is necessary. Existing tagging and syndication platforms are all the tools these vertical search engines need.
There is no way centralized review sites like Yelp, Riffs, Judy’s Book and others can compete with the blogosphere over the long run. Those sites will also have to gather decentralized content, or become meaningless. More and more people are realizing this and writing about it – a good post by Rachel Cunliffe (one example) is here.
Mike, I agree there is value in the scraper model – grabbing content from other sites and packaging it in a way that adds value.
But your assertion that…
“There is no way centralized review sites like Yelp, Riffs, Judy’s Book and others can compete with the blogosphere over the long run…”
…assumes that the blogging format is going to be not just A winner in the user generated content space, but the ONLY winner.
I disagree wholeheartedly.
I would point to the success of centralized, destination Web sites such as MySpace, Craigslist, wikipedia, etc as evidence that people LIKE congregating online where other people are.
Centralized, destination Web sites provide a conversation, they provide an audience, they provide reputation systems and trust systems, and more.
The blogging format works very well for some – but not everyone has the time, passion, and dedication to manage the consistent commitment that a blog entails.
For the rest of us who choose not to blog, online communities such as Yelp and Judy’s Book offer a convenient, lightweight alternative to blogging.
For scraping to continue to be a viable model, at least some destination sites will need to survive, or else there will be nothing left to scrape – unless you really think that blogs are destined to be the only user generated content format left standing.
Lawrence,
I distinguish betwen a “destination site” and “centralized content”. A destination site could be a place where decentralized content is aggregated, mixed and mashed.
I think you raise a point about centralized content that I agree with, though. The collaboration (wikipedia) is not easily handled with decentralized content. That’s why wiki’s are so great. Is Yelp really a wiki? Maybe…
Conversation, however, can happen in a decentralized way. Memeorandum shows us that.
Also, Myspace is really just a bunch of blogs. And craigslist…well, let’s see what happens over the next few years.
I’ve got to agree with Mike on the aggregation of edge content – radical decentralization will win out. On Yelp, I said:
“It’s a nice site, but I feel like they’ve got the same problem as Riffs – there are already hundreds of reviews out there in the blogosphere, and yet Yelp is asking people to create content specifically for their service. This is where structured blogging and microformats come in. Back in November 2005 I wrote about how Riffs could aggregate reviews from the blogosphere, and in mid-December I spoke about the opportunities created by structured blogging. In the long term, I think closed services like Yelp and Riffs are going to lose out to services which aggregate and filter reviews from the edges of the network.”
More here:
http://mashable...siness-reviews/
Yeah, after a while the definitions (blog, wiki, forum) tend to overlap and blur a bit.
I’d say that MySpace is more a collection of profiles, than it is of blogs. But let’s say that you could subscribe to any Yelp reviewer via RSS? I know that you can do that on Riffs, for example as well as on a few other review sites. Would’t that make those sites “blogs” too, and therefore not guilty of the centralized content label?
We may just be disagreeing on terminology. I can’t imagine that Yelp isn’t going to enable RSS feeds of its reviewers. I guess where I’m losing you is this: Yelp aggregates all these individual publishers under the umbrella of a centralized review site. Blogger aggregates (with walls between blogs) a bunch of individual publishers under the umbrella of a blog site. Assuming both offer RSS feeds, how is Yelp any different than Blogger, besides the fact that they enable collaboration between reviewers?
Lawrence – Yes, I agree with you. If Yelp does this, essentially it becomes a collection of specialized review blogs.
Now what happens when/if Yelp trys to stop other services from aggregating this content, as craigslist did with oodle???
For structured blogging to work, you need the help of blogging software vendors i.e MSN spaces, blogger, typepad, wordpress etc. Trick is to make every party in the transaction of a blogging content or for that matter any content, provide and realize value ($$$) for all parties. That is not that easy.
Mike: Thanks very much for reveiwing iNods.
I wanted to quickly confirm on a couple of items. We do not require reviewers (whether the reviews are on their sites or blogs) to register with iNods for their content to be included in the iNods index. We are assimilating our reviews index constantly, and currently use a combination of methodologies to do this. We are working towards refining the automated addition of review pages to our index, but I would not say we have solved the problem fully. This would definitely be a big part of our IP.
Having said that, reviewers can benefit by registering their sites and blogs with iNods as we roll out newer features for them.
Also, we are heavier on the shopping side but are also covering the broader review space and we do expect to evolve it to include all review content in the coming months.
Lawrence, you raise good points but there are different dynamics in play, I believe, when we look at centralized review sites and sites like MySpace, Craigslist etc. MySpace is driven my its community. Craigslist is a web1.0 company and built the momentum over the years – do you think one could start a new Craigslist today?
Also for reviews, fundamentally, consumers are used to searching on the web for reviews/advice and there is just a lot of fragmented advice information on the web today (see the millions of affiliate sites) – and it is all good information. Though time will tell, it might be difficult for sites like Yelp, Judys Book and others to beat the depth of review content that is already there…and growing fast with blogs.
If Yelp opens its content via RSS feeds, it will be awesome, and we’ll be looking forward to adding it to iNods index and deliver traffic back to Yelp.
Mike, there is always that chance of sites blocking us from crawling them, like Craigslist did with Oodle but I believe review content is much much more fragmented and will continue to be fragmented. So even if a particular site stops us from crawling (which I hope does not happen, since we are delivering traffic back to them) we might not be as badly affected as Oodle. Also, I think, local search may not be best suited for this decentralized aggregation model due to lack of a large number of edges with local content.
I can’t really agree with that last statement of yours Mike. There is a much larger perecentage of general Internet users that do not maintain blogs than there are that do. In fact, even going forward many years I’m willing to bet serious bloggers will not become a majority. How, exactly, will the blogosphere completely overshadow sites that don’t rely on it?
Furthermore, I’m having trouble grasphing how structured microformats provided by a certain site that plans to aggregate is any more powerful than just going to the site in the first place. A blog is still a destination on the Web, it’s not as if were fighting physical boundaries that make writing on your blog magnitudes simpler than writing elsewhere. Making aggregated subjective content truly meaningful will require thoughtful structure – in my opinion, a site focused on providing that structure is effective, whether the author writes the review on his blog or on the site itself.
Ashish (#8):
Let me clarify my opinions. WRT your first paragraph, I think the definition of “blogging” will expand greatly. What if there was a site you controlled that you could simply aggregate your reviews? What if Amazon posted that for you based on all of your reviews? That, to me, is a blog, and I think that anyone who takes the time to write content online will, eventually, be wring on the edge. Legacy centralized plays like ebay can survive for a time. New startups have no chance. This is a point that is certainly up for debate, but it is my very strong opinion and I think that new decentralized services to launch in the future will prove this out.
WRT paragraph 2, I think microformats are nice, but I think technology can easily find a way to deal with unstructured content. There is simply too much value sitting out there, and you cannot force people to structure what they write, especially if they have no incentive to do so. Structure is far less important than the content itself, and new services will make that content “less invisible” by solving a hard problem.
I agree with Mike – structure is far less important than the content itself, and new services will make that content “less invisible” by solving a hard problem.
Edge content aggregators really do not have to dictate that the content is in any particular structure or format. This seems especially true in case of review content. Reviewers can write on blogs, simple web pages or other sites and it should be crawled by aggregators like iNods.
I tested a few queries on iNods and it seems to return pretty good results. Mike, I guess they are crawling and generating their own index of review content.
Two issues:
1) Type “canon sd550″ into iNods and “canon sd550 review” into Google. Not a whole lot of difference. Why leave Google to come to iNods?
2) If and when iNods needs to start bringing in some revenue, their current (noble!) mission to “encourage review content writers to create and own their own content and deliver targeted traffic to their sites enabling advertising and other potential income streams” might come in conflict.
Hi Rick:
1. If you found iNods results close to Google results, we would consider good as we are just getting started
On a more serious note, it will be likely that some queries yeild equally good results on Google (Google is definitely a competition to all of us who are building search-centric services), but you will see that the Google results quality dimishes after first few results AND the results may not be as good for all categories of products and services.
2. We do expect to add advertising and affiliate-based revenue streams in the future, but that will not really compromise our mission of delivering traffic to sites/blogs of review content writers. The monetization of that traffic will, ofcourse, depend on the content owners. Also, we do plan to introduce a number of features in the coming months, that will enable review content owners to further improve their revenue streams.
To combat the Google-comparison, you will also see a lot more features for consumers that we will be rolling out soon that ease the buying research process.
I hope this addresses your points to some extent. I do appreciate the feedback – thanks much, and stay tuned to iNods feature announcements on our blog.
Mike, just to answer your question about what happens when Yelp starts blocking iNods from scraping their content, I guess I’d say that I understand where Craig is coming from.
Anyone who has ever had anything to do with an online community knows that it’s all about creating an environment where “sheep sh$t grass.” Where it’s a give and take – members of the community contributing, and those same members benefiting from those contributions. The more freeloaders, the more the community becomes at risk.
So with craigslist, they had built this incredible community where site visitors act like owners. This sense of community has allowed craiglist to become an web institution and to rock the newspapers industry to its foundations.
Along comes Oodle, a (venture funded, for profit) business whose goal it is to make it easier for visitors to TAKE from craiglist. Yes they send traffic, but in the process they essentially transactionalize Craigslist – stripping away the community aspect, and condensing craigslist to a quick way to buy your bike, ticket, and move on to the next thing.
I know exactly where the craigslist guys were coming from when they told Oodle to shove it. By attempting to transactionalize / commoditize the craigslist service, Oodle is endangering that which makes craigslist unique – its “sheep that sh$t grass,” its community give & take.
I love openness, but it seems that those that are the most enthusiastic about it are those who would be takers, and ignore or marginalize the giving part.
I don’t know if Yelp will ever block iNods from scraping its content and making money off that content, but whatever they decide to, I would imagine that they will take a careful look at what iNods brings to the Yelp community, as it’s clear what they would be taking.
iNods (search product reviews, advice) seems very similar to become.com (search product reviews, research). Might be interesting to do a side-by-side comparison of the two.
“radical decentralization will win out”
Care to elaborate?
I think everyone here can see there is potential for structuring/aggregation. Heck, I explored that side in an earlier post on hReview (hope a self-link is okay) hReview review
Now that I’m looking into building a user-review mgt system, it becomes clear that destinations have a ton of options, limited only by their creativity.
First, there is the issue of identity / repuation management. For example, how will aggregators know Joe Bob Sixpack is “Joe_Bob” on Amazon and “Joe_Sixpack” on Epinions? The implications are non-trivial, and the workarounds are non-ideal.
Next, there is the double whammy of getting enough structure & being able to leverage that structure to the fullest, without overstepping the boundaries of attribution. I think “radical presentation” will be just as important as radical aggregation, and the edge, at least initially, will go to the destinations.
+ + +
re: iNods, many of the results are not reviews -about- the products/keywords, or even the category in general, just pages that happen to mention the keywords once or twice? That would seem to be the first level of structure that needs to be addressed, before we can start thinking about the fancy stuff.
It seems that the reviews iNod collects are reviews of actual products (iPod, etc.) and not reviews of the online/offline retailers where one might buy these products. What about BizRate + the surveys customers are prompted to purchase after they’ve been referred by a shopping comparison engine? Will iNods also aggregate these retailer reviews? I searched for the name of one of my clients — who I know has thousands of customer reviews floating around the Internet — but their brand didn’t pull up any results.
One more business model type comment:
I think there is an opportunity / need for sites like iNods to come up with something creative about their business relationships with the sites that they are scraping.
While Google can pull off the “let our bots hammer your site, and we’ll send you traffic” contract with Webmasters – because they control 70% of search traffic, I don’t think this gives a carte blanche to every other scraper out there.
For example, it’s just not appealing to me as someone who runs an online community to let every spider in to drive our bandwidth up and “borrow” our community’s content. It makes it even tougher to swallow that these scrapers/aggregators then look to skim advertising revenue off that content, before (maybe) sending (the least profitable) traffic back… as another poster mentioned, seems like a bit of a conflict to me, and an area where there exists the opportunity for some creative thinking around how to structure a relationship in which both the scraper and the scraped community benefit.
Hi Leslie:
You are right, we are not covering retailer reviews as much right now, but we will be broadening our index to include pretty much all types of reviews including retailer reviews. It will take some time though as we are ensuring our index remains clean (centered on review content) to ensure more relevant results.
Thanks Vaibhav — I’ll have to keep an eye on the iNods blog to make sure I know when that happens.
From the POV of an internet retailer, aggregating a company’s reviews would also provide a great opportunity to respond to and potentially rectify bad reviews without trawling for them all across the web.
For example, I’ve investigated many of the poor reviews my client has received, and though whatever the customer was angry or disappointed about when they wrote the review has been resolved, they are extremely unlikely to return to the internet to pen a positive review or edit their existing poor review.
It’s frustrating because with a BBB complaint there is an official resolution process — not so on the web. As a merchant you’re basically just playing the law of averages, hoping the good outweighs the bad, since both are (ironically) pretty much set in stone once written.
Anyway, since iNods is concered about keeping reviews “clean,” I hope it’s something you guys might consider.
Leslie:
This is an interesting point – something we had looked at but not understood as well. Your example sheds light on how we can look at addressing this issue in someway. I will keep you posted.