In this post, we review the past decade of e-commerce and the key trends. Advances in recommendations technology, together with the emergence of social media and mobile commerce, have combined to change the way e-commerce is transacted. In a follow-up post, we look at current statistics for online retail.

Recommendations Technology Advances

Over the past decade the online retail industry has seen great strides in the use of recommendations technology. Amazon has consistently led the field in this, with its sophisticated blend of personalized, social and item recommendations.

Many of the retail recommendations in use today rely on implicit user data. These systems typically track user data, which is then analyzed with a set of usually proprietary algorithms. The end result: recommendations for users. Earlier this year we looked into Baynote's recommendation system:

"Baynote observes real-time user behavior on a site and looks for implicit, emergent patterns. It uses collective intelligence and an affinity engine to analyze the data. Common behaviors which it tracks include page refers, queries, mouse movement, time spent on a page, peer behavior."

Other similar recommendation technologies we've profiled include MyBuys, ATG and richrelevance.

Social Media Takes Retail to Blogs, Social Networks

As with nearly every other industry, shopping sites have increasingly used social media to promote their wares.

According to Shop.org's recent eHoliday Study, 47.1% of retailers surveyed will be increasing their use of social media this holiday season. Specifically, more than half of retailers have "added or improved their Facebook page (60.3%) and Twitter pages (58.7%)" this year. Nearly two-thirds (65.6%) have "added or enhanced blogs and RSS feeds" over the same time period.

One result of this has been a big increase in implicit social recommendations data across social networks and blogs.

Another trend with ecommerce sites is distributed sales. Anyone can embed an Amazon store into their blog or social network these days. As Kurt Collins of social commerce vendor Cartfly told us in December, this won't replace "end destination e-commerce" - but it will "augment sales tremendously" at the edge of the network.

Mobile Commerce Arrives, Albeit Slowly...

The growth of mobile phones has been a big trend this decade. However, as Sarah Perez wrote in September, mobile commerce in the U.S. market has struggled for momentum.

According to data from eMarketer, more than 70 million U.S. mobile phone users will access the internet from their devices this year. Despite this, the m-commerce market remains immature. In an April 2009 survey by RIS News, privacy and security concerns are still at the forefront of both shoppers' and retailers' minds.

There is some promise that mobile commerce will finally gain traction in the coming decade. Mobile payments firm Billing Revolution found that on-the-go consumers are happy to purchase small ticket items like pizza and movie tickets, for example.

One market that has shown strong signs of mobile commerce growth is Japan, according to Morgan Stanley.

See also our analysis of mobile payments.

Conclusion

New recommendations technologies make it easier every year for consumers to find what they want, social media has driven a lot of retail activity to small websites and social networks, and mobile commerce has slowly but surely gained a foothold in e-commerce.

These are just some of the trends in e-commerce over the past 10 years. While Amazon.com and eBay continue to be the giants of online retail, the Social Web and advances in web technology have both had a big impact this decade.

See also:

• Online Retail Thriving: 8% Growth Expected This Holiday Season
• Top Internet Trends of 2000-2009: Online Music
• Top Internet Trends of 2000-2009: Democratization of News Media

Does this mean Shazam is about to give Pandora and the like a run for their money?

The free Shazam application is best known for its nifty tune identification trick. Mobile users can hold their iPhones up next to a speaker or other source of music and the application "listens" to what's being played in order to identify the song and artist. It also lets you read track and album reviews, read artist biographies and tag songs to share with friends over Facebook and Twitter.

The new application, Shazam Encore, adds even more functionality including improved speed performance, trend lists that highlight what's popular among other Shazam users, a search function that taps into a database of 8 million+ songs, music recommendations and a "drive-and-tag" feature that lets the app recognize when it's in an in-car dock so it can identify what's playing on the radio while you're driving.

But How are Those Recommendations?

Out of all the new features, however, it's the music recommendations option which is the most interesting. Recommendations are the killer feature which can either make or break a mobile application these days. With services like Last.fm and Pandora already providing mobile users with playlists based on a user's likes or dislikes, Shazam needs to be able to do recommendations well - really well - in order to compete with these already popular applications.

In addition, the up-and-comer streaming music service from Spotify also partnered with The Echo Nest's music intelligence platform earlier this year to help improve on Spotify's playlist and music discovery functions. The end results of that partnership have been touted as being like the iTunes' "Genius" feature, only better. Although not yet available in the U.S., Spotify's mobile application is one of the most highly anticipated applications as it provides a new way to enjoy music - through playlist creations that can be listened to both online and off. It, too, will be heavy competition for any application entering into the music recommendations game, including, of course, Shazam.

So where does that leave Shazam Encore? At the moment, its recommendations offering provides you with a list of other songs you might like based on the one track you have pulled up. While this might help you discover new music, you aren't able to create a playlist based on those songs. Instead, Shazam's focus remains more on the sharing of music via tagging and posting to Twitter and Facebook.

As far as how good Shazam's recommendations are, we would need to do a lot more testing before giving a solid opinion - the app is just too new. In fact, it's so new that it wasn't even showing up in an iTunes Store search at the time of writing. The provided screenshot in the App Store doesn't look all that encouraging, though. (Really, a fan of indie band My Sad Captains wants to listen to Katy Perry singing about "kissing a girl?" I don't think so...)

But whether or not the recommendations are up to speed, it remains to be seen whether iPhone app shoppers will be willing to fork over the $4.99 US (£2.99/ €3.99) to have access to them, especially when there's no playlist option included.

Those interested in trying the new Encore application can find it now in the App Store by clicking here.

When you also consider that the wisdom of the crowds is often dominated by small, powerful groups, then the validity of user ratings is further called into question. So why not just get rid of explicit user ratings and use implicit recommendations instead?

YouTube wants to know if a thumbs up/thumbs down system would be be more effective (two options), or even just favoriting (one explicit action to say you like an item).

However possibly a better option is to remove explicit ratings altogether. Does YouTube even need to ask its users for ratings, given the wealth of user interaction data it has?

Earlier this year, ReadWriteWeb profiled some sophisticated recommendation technologies which rely on implicit user data. Many of these systems track user data and, with a set of (usually proprietary) algorithms, come up with recommendations for users. This type of system could well replace ratings altogether in YouTube. While YouTube probably already makes use of the ratings data in its recommendations, as noted above such data is typically unreliable and not very valuable.

As an example of how this could work on YouTube, here is our description of Baynote's recommendation system:

"Baynote observes real-time user behavior on a site and looks for implicit, emergent patterns. It uses collective intelligence and an affinity engine to analyze the data. Common behaviors which it tracks include page refers, queries, mouse movement, time spent on a page, peer behavior (see note about communities below)."

Other similar recommendation technologies we've profiled include MyBuys, ATG and richrelevance.

Are explicit user ratings still valid in consumer apps such as YouTube and Amazon? While we're arguing that implicit recommendations data could enable YouTube to scrap user ratings altogether, on the other hand products like RateItAll are still built around the star system. Let us know your thoughts in the comments.

The Good Stuff Machine

While my6sense is a capable feed reader in its own right, it's what my6sense's founder Barak Hachamov likes to call the company's "good stuff machine" that makes all the difference. While traditional feed readers just organize items chronologically, my6sense actually watches what you do while you read your feeds. The app, for example, looks at messages you read, links you click on, items you share and the position of these items in your stream, as well as items you skip. You can also actively mark a message as interesting by clicking the 'I like' button.

After you have spent only a few sessions with the app, my6sense will have already gotten a good sense of the items that are probably most relevant to you. Of course, the more you use it (we used the web app regularly since the beta launch last December), the better the recommendations get.

In our experience, my6sense's algorithms do a great job at figuring out a user's interests. If you are a real news junkie, you will probably still sometimes want to switch to the regular timeline mode that organizes items chronologically. After all, the items you don't usually think you would be interested in can sometimes really grab your attention (which is, to be honest, a problem that all recommendation systems have to grapple with).

Import and Share

It is worth noting that the app can also import your streams from Facebook, Twitter, FriendFeed, LinkedIn, and Flickr. From within the app, you can also share the most interesting items you find on Facebook, Twitter, and FriendFeed.

My6sense can import your feeds from Google Reader, iGoogle, MyYahoo, Newsgator, and NetVibes. Thanks to this, it's quite easy to get started. Sadly, though, the app doesn't sync with any of these services, so items you read on my6sense won't show up as read in your Google Reader subscriptions. My6sense also offers a curated lists of feeds that new users can subscribe to.

A Few Things to Improve

There are a few nagging omissions in the app, though, that keep it from being really great. Most importantly, you can't tell the app to only display items that were posted in the last 24 or 48 hours. In a way, this makes sense - after all, the app is trying to give you the most relevant items, including those that you might have overlooked. But often, we just want to see what the most interesting items posted today are, and as of now, my6sense can't do that.

Another problem with the app is that once you import your social network feeds from Twitter, Facebook, or FriendFeed, your timeline often gets overwhelmed by these items. At least for us, my6sense performed far better when we disabled these feeds.

Currently, my6sense also doesn't offer an offline mode, so you can't use it to read feeds while on a plane or far from the nearest cell tower.

Overall, we recommend you give my6sense a try. After all, the app is free, and importing your feeds from your current feed reader is easy enough - just give it a day or two to see if the recommendation system works for you.

Two year old wine social network Snooth announced today that it is now powering wine recommendations for the 25,000 editor tested recipes on Conde Nast's food site Epicurious. Snooth says this is just the first of a number of big sites that its custom algorithm will power recommendations on. That cassoulet? Snooth suggests you serve a Montevina Terra d'Oro Syrah 2002 ($15) with it. Nice.

Wine with food has got to be one of the most familiar kinds of recommendations offline, but the online recommendation technology industry is a fast growing one. The belief is that quality recommendations will serve as searches you never knew you wanted to perform - helping users navigate from one logical option to another, possibly making more purchases as a result and hopefully being better served by the websites they visit.

A food site with good wine recommendations sounds pretty tasty to me. Snooth says its recommendations are based on ingredients, cuisine and cooking method.

Here's how it works. First, the keywords are parsed out of a recipe, then they are run through an extensive food dictionary and a long decision tree is then followed. Is it a soup, is it a salad, what is the primary taste? Beef and nuts tastes mostly like beef; beef and liver tastes mostly like liver. How the ingredients are to be prepared is determined by their proximity to preparation words in the recipe. Recipes with expensive ingredients will see more expensive wine recommendations, inexpensive ingredients (lobster vs. shrimp, for example) will yield less expensive wine suggestions. Goodbye old one-liners about "if you're eating chicken!"

Nibbledish, Cookstr, Chow? All cool recipe sites but no wine recommendations, much less very sophisticated ones. It's easy to see how recommendations can provide a competitive advantage in a niche like this.

If that team withstands the month long period of scrutiny that begins now, it will not only mean fame and (some) fortune for them and a big boost for Netflix - it could signal a key turning point for recommendation technology on the web.

In January of this year, we took an in-depth at the Netflix Prize, asking if 2009 could be the year that the goal would be met. In that post we discussed a New York Times profile of the contest as well, where we learned that the company's existing recommendation engine called Cinematch is credited with driving 60% of Netflix's rentals. That system is especially good at predicting "long tail" movies, older more obscure titles that are less well known but make up 70% of what Netflix customers pick. Improvements in Cinematch's effectiveness plateaued in 2006 and the move to offer a big cash prize for outside innovators has captured the imagination of thousands of engineers and their fans.

How Does it Work?

How do you judge improvements on recommendations? Netflix provides contest participants with huge piles of anonymous data about what movies certain customers rated highly, then the teams built algorithms to predict which movies other customer profiles would rate highly based on past patterns. BellKor's Pragmatic Chaos says it can now guess what people will like with a 10% improvement over Cinematch's success rate.

That gets difficult when movies like Napoleon Dynamite, which some people loved and other people hated, get thrown into the mix. It's nearly impossible to predict whether a person will like films like that.

Most of the predictive recommendation systems entered in the Netflix Prize are reported to be quite similar - so we asked in January whether it was going to take a radical breakthrough to top 10% instead of just continued iteration.

That breakthrough may have come when the four teams put their heads together, or it may have been an iterative victory. Time and science will tell.

Some people believe that recommendation as a technology has the potential to be even bigger than search. In our favorite article on the subject, written eight-teen months ago now, Dr. Rick Hangartner, Chief Scientist at recommendation engine Strands, puts it like this:

In the near term, search engines will increasingly incorporate simple recommender technologies to handle approximate queries (e.g. "You asked for this, and based on similar queries/behavior by others, you might be looking for this."). But in the long term, the recommender industry will be larger, and recommender technologies will be more pervasive than the search industry and search technology as we know it. [Because there will be recommendation going on all over the web.]

The application is also available in white-label OEM form and LouderVoice thinks it might be ideal pre-install for many networks. According to the company, "if you don't have an Android phone, trust us, you're going to want one, it's a fantastic system, particularly if you use GMail, GCal, GTalk etc. This will be the biggest mobile platform ever released".

If you want to see this new application "in the wild", check out this video review we've uncovered and keep an eye out for LouderVoice's own screencast coming soon on its blog.

The feature O'Neil wrote about appears to be nothing more than the latest FriendFeed rip-off: truncating repetitive activities. (Ex-Googler Paul Buchheit's FriendFeed is like a Facebook R&D lab without stock options.) Whether Facebook is doing more than that publicly or not, you know they have to be working on recommendation behind closed doors.

O'Neill wondered whether this feature might give the user an option to view other items from the same or other users that Facebook deemed similar to the original post. I say I'm drinking coffee and Facebook shows me a movie, a picture and another message about coffee from one of my friends this morning. That would be pretty awesome for us as users and it would increase ad impressions for Facebook.

More likely is the explanation offered by AllFacebook readers, that this new feature is just a way to scrunch up items that are basically the same so users can't spam their friends' newsfeeds and so that newsfeeds are more pleasing to scan down. In other words, in the image above, the two-headed person on top probably posted about "besplatno-ing" like four times in a row. Facebook decided to show just one of those messages and add a link to view the rest.

That's how FriendFeed does it and it works really well. This seems like a plausible explanation of this screenshot, but it's also a real lost opportunity. Facebook's corny "your friend is a fan of this advertiser's stuff" may be more creepy than compelling - but automated recommendations of all types of items could be great.

We're big fans of recommendation technologies here at ReadWriteWeb, from relatively simple "people who like X also like Y" to more complicated algorithms. The systems are fun to learn about, but the fact of the matter is that recommendation doesn't have to be hard. The hard part is amassing enough data and interested people to be able to make recommendations. Facebook has plenty of data and people, though its labyrinth of privacy restrictions might complicate things a little.

So if this isn't it, and we suspect it is not, we sure do hope that Facebook will soon surface the recommendation technology we assume they are working on behind the scenes.

In the following post we take a look at three of the most exciting startups entering this emerging market. The movement is a logical development now that millions of people are comfortable posting information online. The web's next step is to leverage machine learning. These are three companies to watch who are doing just that - combining user input with technology that improves its performance by gathering and processing data. In this case they are doing it in order to help people make better decisions, but these are just some of the first consumer technologies that will enter the cyborg-like space that combines people and machines in order to better serve people.

Aardvark

Premise: Ask any question by IM and your question will be routed to a tagged "expert" on the topic, among your friends and their networks.

Logic: There appears to be some semantic analysis of the tags given users by their friends and themselves, cross referenced with semantic analysis of the questions asked in order to find the right fit. We presume there is or will be some logic judging the history of successful answers from users so as to rank relative expertise.

History of one query.

User experience: High coolness factor when a real person quickly answers your question. How reliable that person is regarding the topic of the question is not readily apparent. Interesting IM interface facilitates relatively sophisticated interactions based on short commands. Fun to browse through open questions; smart deference to email when people aren't available by IM. Can be irritating to be interrupted by other people's questions by IM, but not such a big deal. Web interface is quite nice but I've hardly ever seen it -- just asking and answering questions through IM.

How It Differs From the Others: IM interface offers almost zero barrier to entry and a powerful hook to return to the service over time. Machine learning focuses on identifying human experts, and search is rich with human interaction, thinly mediated by a smart system. You could call this a friend-network-based, semantically powered expert discovery and conversation system.

Stage: Closed beta; new users get 50 invites. Has been in the works for years and is relatively well baked.

Backing: Made up largely of ex-Googlers. The parent company is called The Mechanical Zoo and has raised $6 million from very hip VC firm August Capital and Ron Conway's Baseline Ventures.

For more info, see this review on VentureBeat.

Hunch

Premise: You may like the same advice for common questions that people with similar tastes like.

Logic: A series of decision topics have been populated with questions concerning factors to consider for each decision. Users go through and answer those questions and are then presented with a series of answers that other people who answered the questions the same way and who have similar tastes have said they are happy with. It's hard to explain but really easy to use. Users can add "factors to consider" questions to any question. There's a really interesting social networking component to it as well.

Home page: random questions; taste-profile-building question about you, users.

User experience: Using Hunch is an odd experience, but it's a whole lot of fun once you get it figured out a little bit. Much of the User Experience design is a model that you'll wish every website followed. It's quite game-like. That said, the site can be overwhelming and make your brain hurt. The service tells me that most people who said they think clowns are funny (as I did), and who don't do video editing on their computers, also liked the answer "no, you probably don't need to upgrade your Mac's RAM." I don't really know what to make of that. You'll probably want to go back, though, and you'll probably want to clap your hands and smile each time you do.

How It Differs From the Others: By far the most "involved" for users of these three services. The user experience is very structured but it's also a lot of fun. You could call this a profile-driven, crowd-built recommendation system.

Stage: Closed beta; new users get a very limited number of invites. One co-founder says it's still quite rough around the edges, but if that's the case we sure can't see it.

Backing: The company has raised $2 million in VC funding and has an executive team of successful startup founders who've sold other companies, most prominently Caterina Fake, one of the co-founders of Flickr, who is now Chief Product Officer at Hunch.

To read more about Hunch, see the company's official FAQ.

Swingly

Premise: Answers to any question you have can be found out around the web. Swingly finds those answers hidden in plain text articles, databases and other Q&A sites. Then it makes them structured for easy sorting in response to queries.

Logic: This un-launched company uses Seti@Home-style distributed computing to perform Natural Language Processing on pages all around the web, hunting for information that can be turned into Questions and Answers to serve up to Swingly users. The company believes that "next-gen search should [include] 'micro-retrieval,' rather than return pages, and return only the content (word/sentence/paragraph) you need."

A screen shot from earlier this week.

User experience: We've not been able to test Swingly yet, but it looks relatively straightforward so far. There will be any number of additional services built out as well, including a widget for bloggers to offer Q&A functionality on their sites. When you talk about billions of pieces of structured data that you can query with common questions, almost anything is possible. That said, Q&A is a field that several other companies have done a good job nailing already, from Yahoo Answers to ChaCha to Mahalo.

How It Differs From the Others: Swingly is the most mysterious of the three services and the most likely to become "a platform." It's also the most likely to suffer from the Powerset dilemma: hype, hyper-nerdy ambitions, big expectations, lackluster launch, $100m payday from Microsoft, getting turned into a term of derision among some in the industry and maybe buying a yacht.

Stage: Closed alpha right now. Starting to make the first public rumblings with screen shots, Twitter presence, initial PR outreach. "Alpha coming in late March and a public beta in mid-May. The alpha version will use an index of about 850 million question-answer pairs (more than all the Q&A sites put together) and will only be searchable. The beta release will consist of about 5 billion question-answer pairs and will include full questions and answers plus semantic search capabilities." - CEO Andy Hickl, last month

Backing: Dallas, Texas-based Swingly CEO and founder Andy Hickl is also CEO and President of the very related-looking Language Computer Corporation. CNN calls that company "closely held."

One thing's for sure - we're going to hear a lot more about Swingly. The company is working with Porter Novelli's Josh Dilworth, one of the smartest and most effective PR agents in the tech industry. Dilworth has a history of working with uber-nerd companies and getting them huge media coverage. His recent clients include database super-search engine Wolfram|Alpha (our review) and the most-discussed consumer semantic web company to date, Twine (our most recent coverage).

To follow the unfolding of Swingly, check out Hickl's personal blog.

Those are three companies we'll be watching closely as they break new ground in the combination of social and machine learning online. Which would you be most likely to go to first with a question? We'd love to hear from readers who have thought about this field, who are doing work in it as well, or who have initial impressions about these services that they would like to share. We expect to see a whole lot more like this in the near future.

Title photo Cyborg 2.0 by Y0si CC on Flickr

Stumble in an Iframe

The DiggBar, of course, sparked a lot of controversy though StumbleUpon's new toolbar does not include a URL shortener. So, unlike the original DiggBar, there is probably little reason to assume that the new StumbleUpon toolbar will steal too much search engine 'juice,' even though it uses an iframe to show the original page. As users aren't likely to share the long StumbleUpon links or use them to link to a site from their own blog, this shouldn't be too much of an issue. But it should be noted that, as far as we can see, StumbleUpon does not return a canonical link which would tell search engines like Google to ignore the StumbleUpon link and index the original page instead.

We talked to StumbleUpon about this earlier today, and the team there didn't seem too worried about this, but instead emphasized that the bar was easy enough to close. It should be noted, though, that whenever a StumbleUpon user shares a story by email, the recipient will see the toolbar by default.

Until just a few weeks ago, StumbleUpon was part of eBay, but now, StumbleUpon is once again an independent company after its founders and investors agreed to buy the company back. It's good to see that the company continues to roll out new products on its roadmap while it is going through yet another transition, though the company is obviously going see a lot of competition from other content discovery services and social networks in the near future.

Even more significant, few sites today are constructed solely from internal site resources. Hosted metrics, recommendations, news, store locators, stock tickers, friend followers, and so on and so on are rapidly deconstructing the whole notion of "place" through the active participation of the "web-fabric" layer of the web community.

From this perspective, most recommendation services are still stuck in the feudal worldview: the black box recommender knows you (whether "you" are a visitor or place-owner) better than you know yourself and determines, in its infinite wisdom and authority, what content should be presented to you. The place-owner may have some input into presentation and even, though less so, content, but only in a very limited way.

While this situation is useful in certain cases because of the total passivity afforded the place-owner and visitor, it severely limits the potential contribution of recommendation technology.

Personal, Real-Time Conversation

There is a broader view of recommenders, though. The business value of recommendations is that they bring the place-owner into a one-on-one, real-time, conversation with the visitor. As such, a recommender must be able to accommodate the active participation of both the place-owner and visitor. Recommenders play the role of the salesperson, the agent in the company who has one-on-one contact with each shopper. This is in contrast to the site designer, who is more akin to the display designer in a bricks-and-mortar store and who can only target segments of the population who are expects to pass the display, not individual shoppers. Recommendations are also narrower in concept than personalization tools, which are analogous to store greeters: they may personally greet you when you arrive, but they generally don't follow you through the store as you shop or interact with you in real time.

Okay, but why a conversation? Consider the typical interaction between a sales agent and shopper in a bricks-and-mortar store. The shopper enters the store and starts looking around. At some point, the sales agent asks, "Can I help you?" "No thanks, I'm just browsing," By this point, the sales agent has probably already observed the shopper and made some inferences about the shopper's intentions and receptivity and about associated sales opportunities. The shopper, in turn, has been assessing the store's inventory and pricing.

Like these sales agent, place-owners have a tremendous amount of knowledge about shoppers, sales tactics (like cross-selling, upselling), and their own business objectives, both short- and long-term. Much of this knowledge is unavailable to automated recommendation engines, no matter how much data they gather (and the ultimate prize for optimizing discounted infinite-horizon shopper value is computationally intractable even if we had the data). So, the recommender is better tasked to take advantage of the wisdom of the place-owner "in the moment." Of course, an uninformed recommender is just a degenerate case and may still be useful.

One advantage of the web is that transaction costs are low. Most place-owners can't afford to have human representatives in sessions. Most explicit communication by the place-owner must be in the form of policy or strategy, rather than actual real-time communication. (Notwithstanding this, interaction with a live sales agent may well be an appropriate option for a recommender to trigger in certain situations.)

Situation/Response

One way to think about this is like "situation/response". The situation description might cover visitor location, web page visited, catalog, date (e.g. if it is a holiday), place-owner internal item information (e.g. from a supplier catalog or internal access and sales statistics), visitor community information (e.g. sales ranking, review ranking), or even external information (e.g. Google search ranking, Amazon ranking). The response should be a specification over recommender behavior, as well as resulting recommendation content (e.g. show a pair of Nike's under $50), and presentation, both style and modality (e.g. use an animated GIF showing all available colors). Perhaps, as mentioned above, modalities even extend to bringing a live sales agent into the real-time conversation.

While limited work has been done on place-owner participation in recommendation-system content and presentation, the situation is far more dismal for the visitor. A broad array of modalities are available for visitor interaction, but few if any are available in most recommendation systems. A simple "No, that's not what I'm looking for" (e.g. a thumbs-up or thumbs-down icon on a recommendation thumbnail) might go a long way to making the shopper feel noticed and appreciated. I can say to a human store clerk, "I'm looking for a pair of Nike's under $50" -- why can't I tell the average recommendation system the same thing? Notice that this starts to overlap with the expressivity needed on the place-owner side. The main difference is that the visitor is always in the moment, so there is (usually) no need to specify context.

The above sketch is intended to crack open the door on the enormous range of possible capabilities, modes, and time-scales of participation by place-owner and visitor. Once we've opened this door, there is no reason not to open it to the visitor community and the web-fabric community as well. There are three primary points:

1. A place is no longer a feudal domain; all stakeholders now demand a voice.
2. A recommendation engine is the locus where understanding of content and understanding of the visitor-in-the-moment come together.
3. As a result, recommendations are the logical ground for crucial real-time conversations between place-owner and visitor.

Given our initial definition of place, we might also ask about the role of and opportunity for participation among other stakeholders. For example, can the interaction between the site designer and the visitor or web-fabric community also be viewed as an ongoing conversation, rather than an episodic, one-way information flow at the time of site design? The answer is yes, but that is a topic for another time.

Conclusion

Recommenders need to open up to allow increased place-owner, visitor, and community participation in both content and presentation. This is best done with the assumption that a recommender is meant to facilitate situated, in-the-moment conversation between the place-owner and visitor.

This was a guest post by Bruce D'Ambrosio, VP and Chief Architect, OnDemand Personalization at ATG, Inc. He was the founder of CleverSet, which was acquired by ATG. He is also a former Oregon State University computer science professor.

Why We Need This

Sites like Yelp, Amazon, the iTunes store, and others have been coming under fire for not having trustworthy reviews. Thanks to anonymous user IDs on some sites, reviewers can be anyone with any agenda. Often they are. On Lunch, however, those drive-by reviews contributed by someone associated with the company or product being reviewed (or with an axe to grind) will not be prominently featured. The reason? Lunch.com's Similarity Network.

The Similarity Network

The Similarity Network is probably the most important feature of this new community - without it, Lunch would just be just another Yelp. After signing up, you kick start the matching engine by playing "ExhilaRATE." Although that name is somewhat unintuitive, clicking the link takes you to a section of the site where you can - guess what? - rate things like movies, books, food, sports, politics, animals...whatever. The experience of rating items here is a lot like that of Amazon's recommendation engine. If you've ever killed a few minutes on Amazon training it to get to know you better, you'll find Lunch.com's engine fairly similar.

The difference is that Lunch.com's engine groups things to rate into categories with titles that sound a lot like Facebook Apps (Top Movies of 2009, What's your Favorite Wine?). The Facebook flavor to these "games" makes sense because in the future, Lunch.com will launch a Facebook connected-experience, perhaps even a standalone app. In the meantime, however, you must go to the site to rate items.

The more you rate on Lunch, the better your matches become. You can see your matches and the percentage of compatibility between you and those like you. There are also tag cloud displays that show what items you both like and which ones you don't.

With Lunch, You Can Rate Anything

If you're still wondering why you would migrate away from more mainstream sites to something like Lunch.com, there's another reason this particular community holds appeal: it allows you to make anything ratable. Again unlike Yelp, ratings don't have to focus on products, services, places, etc. They could also be opinion pieces - like what you thought of Michelle Obama's new outfit for example. That opens the door for a much wider range of recommendations and - since you're matched with those like you - those recommendations will be relevant to your interests.

Lunch.com is in private beta, but you can try it now with the invite code "ReadWriteWeb." To use it, just click the link on the right-hand side of the screen that says "Have an invite code?"

]]> Discuss]]>

Loomia is similar to Sphere and another app we reviewed recently, Apture. These services all aim to serve up more clickable content options on media websites - which means more user engagement and time spent on site for publishers.

We spoke to Loomia CEO David Marks and asked him how Loomia compares to Sphere, which at first glance appears to have much in common with Loomia. Marks said that Sphere is trying to do "semantic classification", i.e. analyzing the content of an article and recommending further content based on the findings. However Loomia focuses more on the user and so it does behavioral type recommendations. This can result in a more diverse set of topics, because users typically have a range of content preferences. It depends on the article though, said Marks.

Loomia currently has 2 types of deployment:

• Content (e.g. WSJ)
• Video (e.g. Brightcove)

Marks told ReadWriteWeb that video advertising is currently selling well for big media publishers. Accordingly these publishers typically now want to drive users to their videos - and Loomia has a widget to do that.

Marks told us that a lot of their publishers are "dollar focused" this year, therefore recommendations have become more than just an interesting feature on a website - they can drive more advertising dollars. As an example, Marks told us that a media website's Finance section may sell out with ads, but its Politics section may not (fairly common in big media websites). But the Politics section tends to get bigger page views, so to address the imbalance Loomia's recommendations widgets can drive users from Politics to Finance.

We've been looking at how recommendations are being used in the retail sector a lot, and Loomia is a neat example of how the same technology can have real value for the media segment. Let us know in the comments what other recommendation technologies have caught your eye in publishing.

MyBuys' Technology

The services approach means that MyBuys deploys a variety of algorithms and doesn't favor one approach - unlike, for example, ATG, which uses a method it calls "Statistical Relational Learning" (SRL). This is really the crux of the difference between MyBuys and the other companies we've profiled so far. The likes of richrelevance, ATG and Baynote all have a defining technology (usually patented) which for each is the foundation of its recommendations approach.

MyBuys has no specific algorithmic approach. Rather it appears to license technologies from companies such as Blue Martini, BroadVision, MarketLive and Microsoft. However MyBuys does still have a patent on the technology which brings all these disparate algorithms together - it calls it a "patented portfolio of algorithms".

MyBuys' recommendations are a javascript include for their clients' websites - i.e. the heavy lifting is done on MyBuys' servers. Their clients can see their stats in a MyBuys portal, and also summary stats are emailed to the clients.

Understanding Consumers

MyBuys has a team of people that focuses on site performance for its retail clients. This team - which works across all of MyBuys' client base - focuses on driving performance using a variety of tools and processes. They also do experiments for clients to find out what works best. Rosenblum noted that MyBuys is almost always paid on performance.

On its website, MyBuys says that it "creates deep consumer profiles based on both explicit information we collect from you and from shoppers when they sign up for alerts and implicit information we collect as shoppers interact with your site." Rosenblum claims that MyBuys "understands consumers at a deep level", whereas he said that its competitors don't necessarily do. He told us that the Baynote approach is "strange" because they don't focus on the individual, but rather the 'wisdom of crowds' (which he said is a 'lowest common denominator' approach). Further, Rosenblum claimed that many of MyBuys' competitors don't understand the product catalog, that they suffer from the "cold start" problem - i.e. with a new product there is no place to start, unless you know about consumer retail behavior.

Side note: we're sure that MyBuys' competitors would disagree with some of the above assertions, so we welcome feedback from them in the comments below. One thing we've found in this series is that each company in this space is very willing to talk down their competitors! A sign of a very competitive market.

Examples

On to examples of MyBuys' approach. One is World Market, a retailer of furniture and other goods from around the world. It has a 'May we recommend' section on its homepage, which Rosenblum told us is based on MyBuys' algorithms and what other people have done on the site. After the user hits the homepage, MyBuys tracks that user - they know where they came from, they pay attention to what the user clicks on next, and so on. On product pages, there are a variety of different recommendations on the right of the page under the heading 'More great finds'. The categories under this heading can differ (e.g. for some products there may be no 'featured' recommendation).

Another example is Golf Galaxy, a web retailer of golf gear. This has recommendations such as "Other great ideas" and "People also bought". It also serves up recommendations in the shopping cart: "You may also like".

MyBuys doesn't just do website recommendations, it uses email a lot. If they know the email address of the customer, they will send follow-up emails (e.g. if a user abandons the shopping cart). Rosenblum told us that this works very well, however he assured us that emails are 100% opt-in. He said that for every dollar MyBuys drives through the site, another dollar comes through the email channel.

Conclusion

So how effective are MyBuys' recommendations? According to the company, when recommendations engage consumers (i.e. a user clicks on a recommendation), they're 5 times more likely to convert than when there are no recommendations. Rosenblum told us that its clients see an increase of overall site revenue between 5-20%, which is a similar figure to that which other recommendations vendors have given us. The addition of email usually results in even higher conversions, the company claims.

As to how MyBuys compares to its competitors, as we've noted in previous reviews it's very difficult to make a judgment on that. However we're interested to note that a recommendations vendor can compete well in this market without having its own unique patented algorithm. MyBuys pushes the 'services' approach much more than the other vendors. We're sure that some of MyBuys' claims about the competition would be challenged, nevertheless it appears to be a successful business in the retail recommendations market.

There's a certain seductive attraction to the idea that collaborative filtering and artificial intelligence could hand us our heart's desire before our hearts even think of it. Think of the time and effort I could save if I didn't have to make decisions about what to eat, buy, wear, listen to, watch or read. When it comes right down to it, free will is a genuine time suck, and seriously cuts into my blogging schedule.

More Noise to Signal