The pace of technological innovation is continually increasing. The internet of today bears only a superficial resemblance to the internet of 5 years ago. And itâ€™s very likely that the services still evolving today, now mainstream trends, will look very different even only two years from now. GigaOm took a brief, general look at the 20 years of evolution to todayâ€™s social networking services. While the past is a valuable guide, the ability to combine new technologies (mashups, the mobile web, peer-to-peer) with current ones will send them down paths impossible for the past to predict.
While there are innovations made by pioneers such as SixDegrees that still have yet to be integrated, we can no more predict the future of Facebook by looking at Friendster, than we could have predicted the evolution of the modern day cell phone from our land-lines by looking at the telegram. So Iâ€™m writing a series of posts contemplating the important changes weâ€™ll see in social networking services (SNSs) as early as 2010, and no later than 2015. The first of these isâ€¦
Recommendation & Review Integration
Many may not realize it, but Facebookâ€™s Beacon â€™serviceâ€™, in which purchases you made were displayed to your friends, was a poorly thought out, horribly misguided attemptâ€¦ at something we all want (even if we donâ€™t admit it): our friendsâ€™ opinions. But in addition to the privacy-infringing implications of Facebookâ€™s premature program, they failed to realize that a purchase does not equal praise. Sometimes we buy things because we have to, or for other people. Sometimes we buy things everyone else is buying and spend months thinking â€˜what a waste of money that wasâ€™.
Thatâ€™s why reviews are so valuable. Everything from experience-based reviews of the seller (eBay), to objective reviews of a product (Consumer Reports), to subjective reviews of entertainment (music/movies/books). While both of the former can have a subjective component to them, especially for the latter itâ€™s as important to have contextual information about the reviewer as it is about whatâ€™s being reviewed. Knowing someone personally is the very best context, and thatâ€™s the largest part of the value SNSs can bring by integrating reviews. As useful as it may be to know that â€˜500 people liked X, 100 people disliked Xâ€™, itâ€™s much more useful to also know â€˜1 of your friends liked X, 5 of them disliked Xâ€™. Of course, if X = Harry Potter, that could also be a sign you need new friends.
Next time Facebook or another SNS takes a stab at this, it wonâ€™t be by saying â€œLet us keep track of everything you buy and promote those things to your friendsâ€. Instead, the way to do it that offers the most value to you, your friends, and SNSs, is by saying â€œTell us which purchases youâ€™d like us to keep track of (via cookies, or importing data from Amazon/eBay/iTunes or Mint/Wesabe). Weâ€™ll keep them confidential until youâ€™d like to rate your purchases, with the option of keeping your rating anonymousâ€.
Benefits of Social Network Reviews
Any recommendation engine built on your data wonâ€™t consider a purchase as praise. Everything you bought because you had to, or didnâ€™t care about, or disliked (but not strongly enough to give a negative review), is simply ignored.
Your friends get reliable data on recommendations for and against from a context thatâ€™s meaningful to them. If enough of their network has reviewed a product/service, they might even have the option of segregating those reviews into more narrow contexts (coworkers at business A, friends from school G, contacts in area L).
You can establish yourself as an authority. Drink a lot of wine? The SNS should offer you the option of displaying your wine reviews prominently on your profile. An SNS tracking review trends across networks can recognize: â€œThese two people have the most dissimilar views on wine that are still agreed with by the most people in region Q. Letâ€™s display their reviews as a kind of point/counter-point to everyone in that regionâ€.
Friends can still use people whose recommendations they trust the most. â€œView all of Suzyâ€™s reviewsâ€ (or â€œ[category] reviewsâ€) would be useful. Of course, anything Suzy has reviewed anonymously isnâ€™t displayed here, unless itâ€™s Suzy herself viewing it. The next step is being able to ask, â€œBased on Suzyâ€™s reviews, would she like item X?â€ The SNS can use the same recommendation engine it uses for Suzy herself to give a percentage chance of approval by Suzy to whoeverâ€™s asking, probably someone whoâ€™s liked Suzyâ€™s other reviews.
Social Network Services themselves have much better monetization options. Rather than advertising everything, hoping desperately that it generates enough sales for a retailer to keep advertising with you, SNSs can become a more direct part of the sales process. Identify the top 10 most highly-reviewed products, or most highly-reviewed of the month, or most highly-reviewed in a category, or most-quickly gaining reviewed productsâ€¦ and sell rights to a â€œBuy this nowâ€ link from the review. Better yet, setup an auction between multiple retailers of the product, for whichever one offers the SNS the highest percentage of the sale at the lowest sale price for the users.
The Key Issue of Reviews
Reviews are not a sales platform for the long tail. Theyâ€™re classic Pareto Principle: 80% of sales will come from 20% of the reviews. SNSs can make the most money, and save their users the most money(a classic strategy for gaining more users, therefore making more money) by putting sales-conversion-rights for each of those 20% up for bid, and either ignoring the rest, or let Amazon/Walmart/Target bid on the other 80% as a whole. Not only will specialty retailers be able to compete against giant general retailers for items within the top 20% of reviews, they may be the only ones who supply popular-but-expensive items within the top 20%; their ability to bid lets SNSs profit early from trends quickly bridging the early-adopter/mainstream gap.
More on the future of social networks soonâ€¦ what do you think theyâ€™ll look like in 2010?