jump to navigation

Going social: Recommendations engines need to factor in consumer reviews November 28, 2011

Posted by David Card in Uncategorized.
Tags: , , , , , , , , , , , , , ,
add a comment

Last week eBay announced it was buying Hunch, reportedly for $80 million. Even though it has its own recommendations system based on searches and popular items, eBay grabbed Hunch’s recommendations engine. The acquisition illustrates the trend of adding social media to the data-powered recommendations credited with increasing Amazon’s sales by 25 percent. But get ready, because the next stage of commerce recommendations will further mix in human-powered consumer reviews.

Pioneered by Amazon, the first wave of e-commerce recommendations engines were based primarily on collaborative filtering. That is, they compared a lot of data: “Customers who bought product x also liked product y.” As with most big data analysis, recommendations are not only about pattern matching but also about which patterns best predict future purchases. And like all pattern matching, the company with the most data usually wins. Social media, via open APIs from Facebook and Twitter, bring new data for analysis and lower the cost of entry for recommendations engines.

Game-changing social data from APIs

Hunch creates a “taste graph” of a user’s interests based on analyzing available social media data from Facebook, Twitter and others that is fine-tuned by user responses to a Q&A process on the Hunch site. That’s why it could deliver useful results without suffering from a “cold start,” or lack of enough data to be predictive.

Effective e-commerce and media recommendations will mix in big-data analysis of different information sources including product factors, observed purchase behavior, and social and interest graph data from APIs. They will create algorithm-driven engines from that data and present it alongside customer reviews that themselves can be filtered and analyzed by adding structure through categories, comparisons and ratings schemes the way Amazon and Best Buy do.

Another recommendations pioneer, Pandora creates personalized radio stations based on factor-based analysis somewhat like Amazon. Pandora programs based on mapping and matching song and artist characteristics, along with a user’s likes or dislikes. Its effectiveness makes it the leader in online radio, with 40 million active users and $75 million in third-quarter revenues. Lately, Pandora has been scrambling to build out social features to enhance its website and apps and get more inputs.

A startup with an MIT Media Lab background, The Echo Nest has begun licensing its music recommendations engine to online music programmers like KCRW and Clear Channel’s IHeartRadio. The Echo Nest automates its song-characteristic analysis by running each track through audio analysis. It also scans online blogs and music pubs for information that informs its audio analysis. This makes its approach more scalable — and cheaper — than Pandora’s engine.

A blended approach in the future

The established engine technologies still work, and social media data lets new players build them effectively. Last year, Amazon started experimenting with available social data. When connected via Facebook, Amazon remembers a user’s friends’ birthdays, suggests items popular among friends and makes recommendations based on favorites the user has listed explicitly on his Facebook profile. But so far, all of those results are on a separate, “beta” page of recommendations that Amazon links to from its personal recommendations page. They’re not mingled with Amazon’s traditional recommendations and user reviews on product pages — yet. Likewise, as it learns what social data is most predictive, I expect Amazon will incorporate the social signals it gathers from Facebook connections into its recommendations-ranking algorithms, the way Microsoft uses Twitter and Facebook in search.

Successful recommendations will take this approach of blending social and data. At GigaOM’s RoadMap conference, Wal-Mart sketched how its social commerce strategy, driven by its Kosmix acquisition, would focus on search, recommendations and local, in-store context rather than stores on Facebook. Wal-Mart is one retailer with serious data smarts. This is just one example of how, for commerce, it’s wisest to think of social media as data sources rather than shopping hubs.

Question of the week

Who do you think will build the best recommendations engine?
Advertisements

Predicting Twitter’s Best Business Opportunities April 5, 2011

Posted by David Card in Uncategorized.
Tags: , , , , , , , , ,
add a comment

Last week, Twitter’s original CEO, Jack Dorsey, confirmed he was re-joining the company to head up product development. Dorsey returns to Twitter to correct some mistakes and address backlash from vocal digerati and, more importantly, from members of the Twitter ecosystem. Blogger complaints killed Twitter’s QuickBar, an iPhone app feature with a badly executed advertisement. Soothing the companies trying to build businesses around Twitter APIs will be more difficult. Twitter partners and competitors alike want to see how Dorsey will align Twitter products with its best business opportunities.

Dorsey laid out some of his early thinking at a Columbia University appearance. He conceded Twitter needs to be clear about its platform and product direction, and advised third-party Twitter developers to stay away from mainstream client apps. Rather, they should focus on integrating technologies like geolocation, recommendations, filters and mobile sensors. Actually, Dorsey acknowledged that Twitter client maker TweetDeck was great for a minority of high-value Twitter power users. Twitter itself, he said, should focus on attracting and serving more mainstream users — the ones that are consumers of Twitter content rather than creators.

Serving Mass-Market Consumers

Developing for the masses will help Twitter continue its evolution from an incestuous microblogging tool for techies, journalists and social media professionals into something a lot like a broadcast medium. ComScore tracks about 20 million U.S. monthly users of the Twitter site (undercounting mobile and client access, perhaps by 20 percent). One API watcher says the vast majority of Twitter accounts follow fewer than 10 others. Twitter must fix that if it’s going to bring value to mainstream content consumers.

Twitter’s history leads it to focus too much on connecting users to other users, rather than users to topics. Its first-screen promotions to “see who’s here” and view “Top Tweets” link to people or brands, or to individual tweets. Popular “Trends” displayed through a local filter on a user’s personal page is more topical, and more in line with mainstream online media approaches, where current headlines, “most popular” and local news/weather/events lead. Mass sites tell me “most popular” is far more effective in generating clicks than “related items.” Dorsey should prioritize collaborative filtering over complicated content management taxonomies.

But Twitter should also collect channels of topics to help unsophisticated users follow more relevant feeds. Twitter already partners with Sulia to deliver curated topic channels to other media companies based on Sulia’s editorial and algorithmic analysis of expert content. It should use those topic and time-driven channels itself. Twitter could promote recommendations with a smarter version of Twitscoop’s real-time topic cloud.

What About Advertising?

Though its ad platform is a product, Dorsey didn’t say much about revenue generation at Columbia. He admitted it was a challenge for marketers to tie together Twitter’s three current ad formats: Promoted Trends, Promoted Accounts and Promoted Tweets. Lately, Twitter has been telling advertisers to concentrate on Promoted Accounts and Promoted Trends at the expense of Promoted Tweets that run in a user’s feed.

In theory, the site takeover approach of a Trend could mimic timely, mass-reach advertising used by portals like Yahoo and AOL to great success for movie studios and holiday-themed sales. But a Promoted Trend now is a barely highlighted little text unit. Twitter’s attempt to feature it on the QuickBar attracted derision from digerati, who complained of its lack of relevance (and who probably use TweetDeck on their desktops, anyway). A smarter play would be a flashier ad unit on the Twitter.com site, where mainstream users congregate.

Better contextual targeting could alleviate some of the complaints about relevance. (Promoted Tweets show up as a result of Twitter searches.) If Twitter doesn’t want to manage a targeted ad marketplace, it could draw on the expertise of OneRiot, a company that’s trying to build a real-time ad network for other Twitter clients.

Question of the week

How should Twitter prioritize product development?