Email Invite from Sears.com

When I got the above email from Sears inviting me into a new social shopping experience, I hoped that they’d found a way to combine MySears and Sears.com together more contextually and pervasively, letting me move easily between the “get advice before you buy” approach of MySears.com (with its action verbs being join, explore, and connect) and the shopping focused Sears.com.

They haven’t, but what they have done is introduce more social functionality into the shop. Visit sears.com and in the utility navigation right underneath the multi-brand bar (Sears, Kmart, Crafstman, Kenmore, Lands End, etc) you should see an option which toggles between “visit our social site” and “leave our social site.” Clicking on “visit our social site” and you’re greated with this splash screen explaining the new experience:

Sears' New Social Experience

It seems Sears has added a bit of social networking functionality as an overlay to the shopping experience, letting you “message” and “follow” other users, seeing their (on site) social activity. You can view profiles, see who other users are following and followed by, get badges, join groups, and do many of the other activities we’ve come to expect in the era of Facebook and Twitter.

The Bearded One - a Sears Associate and one of the public profiles on Sears.com Social Experience

What’s more unusual is the relationship this new social site has to the existing store and community. MySears.com continues to exist on it’s own tab, suggesting a separation between the researching activities (“get advice before you buy”) and the shopping activity of Sears.com. But the new social site is essentially a kind of layer over the regular shop.

Social Site Version of Product Detail Page

Non-Social version of Product Detail Page

Look at these two cropped product detail pages. Both are of the same product in the Sears.com product catalog – a Blu-Ray DVD player. One image is from within the social site (note the toggle says “leave our social site”) and the other is from outside the social site (“visit our social site” being the link). The language suggests that the social site is a separate place (one you visit and leave) but the experience suggests that social is more like a layer you turn on and off – like the virtual reality overlay in an iphone application which adds review information to your camera view of the street on which you’re walking.

In other words, there’s only a few rows of difference, mixing some social activity (Q&A) in place of specifications, adding a product wiki, and repeating the social network information (people in your network who like/want/own the item being viewed) again later in the page.

Unfortunately there’s a lot of empty social areas at this point – not many items that lots of people in the network have liked/disliked/wanted/owned. It also seems like asking a lot to require users to now find new folks to follow, rather than importing and matching their social graph from Facebook, Twitter, or an email address book search. You can invite people via email, but I didn’t see any easy way to build a new network.

My Network is not very robust - nor I suspect will most user's be at first

What do you think of what Sears is trying here?

I love the idea of activity streams (and the standard for syndication of them) in the context of shopping (see Swipely, Blippy, and Shwowp), but I’m not sure how Sears will get enough critical mass built up if users have to specifically choose to follow each other, and their activity is only visible inside Sears’ own walled garden.

Like Automattic, they’ve essentially avoided the Facebook “like” button in favor of their own – but will users go to the trouble of liking a product in multiple places, or will Sears just end up with a quiet network where the only active users are associates?

Hybrid (photo by Burning Image)

Late last week, Plaxo and Google unveiled an implementation – currently in limited testing mode – of OpenID and OAuth working together to create an improved user experience. In essence, the implementation affects Gmail users receiving invites to join Plaxo Pulse. They call this a “hybrid approach” and I think it will have a significant impact as it significantly simplifies the flow.

Plaxo created a custom landing page, based on knowing that the user received the invite at a gmail address, which means that the user has a google account, which means that the user also has an OpenID. (It wasn’t clear to me if the landing page is triggered by a query string parameter or wholly different url embedded in the invite itself, or by a referrer check or the like).

Given that knowledge, the landing page offers just two choices: one big button labeled “Sign up with my Google Account” and a non-graphic link which says “Or, use another address.”

If the user clicks “Sign up with my Google Account,” they get the optimized flow, and get a consent page served by Google which tells the user what they are being asked to consent to, including their gmail address and a request to allow Plaxo to access their Google contacts.

What’s great about it is that when the user accepts, they’ve used OpenID to authenticate to Plaxo based on their Google Account, and they’ve used OAuth to authorize Plaxo to access their Google contacts – but the process never mentions either standard. It’s two great things which are even better working together, and it creates a better user experience.

Technology, like design, is at its best when it disappears.

Peanut Butter Cup Heart (photo by Bob Fornal)

Of course, similar kinds of behavior can be accomplished through Facebook connect – but the difference in this case is that both Plaxo and Google are big supporters of the concept of the “open stack.” All the technologies involved are open, in the sense that they can be implemented by any party (and in fact have associated open source libraries in multiple languages to ease that implementation). To top it off, the whole implementation itself is being released as an open source project called step2.

This means that the same approach – requesting an OAuth token (access to some particularly scoped functionality, like Google contacts access in this example) as part of an OpenID authentication exchange – can be (and most certainly will be) used by Plaxo with other webmail providers, by Google with other social networks / membership sites, and in contexts where neither Google nor Plaxo have any involvement.

For more info:

• Plaxo Blog post – “Introducing Two-Click Signup“
• Google Data APIs blog post – “Bringing OpenID and OAuth Together“
• Watch Google and Plaxo developers discuss the integration on Episode 26 of Social Web TV (and then subscribe to watch the whole series – unfortunately not yet compatible with Miro so you have to go to the site to watch)
• Check out the step2 project on Google Code
• Google’s Federated Login API may be the simplest way to add OAuth and OpenID interaction with Google

Building on the momentum of all the (OpenSocial based) applications they added a few weeks back, LinkedIn is now rolling out events. In this video, Christine Wodtke demonstrates how the application leverages your social graph, showing who in your network is attending various events:

Its a great idea, and I’ve already found or created events for all my conferences coming up. (I’m tempted to create events in the past, as a way of adding conferences where I’ve presented to my LinkedIn profile. The “add an event” flow doesn’t seem to prohibit that, though I haven’t followed it all the way through yet).

I wish the recommendations (which events they suggest you might want to attend) were a bit more precise, but I guess that’s a result of relying on things like “industry” set in your profile (mine is set to “Internet” which must be hard to match on), or job title (“Next Generation Internet Strategist” is not on many event planners’ lists of target job titles), or even education (my educational background is pretty varied and not neatly tied to what I do now). I think it’d be great to allow me to configure the app to add some tags of interests – and maybe let me choose how recommended events get sorted (date, distance, relevancy, or some combination thereof).

It would also be good to have a simple way to get an event’s URL – for now I’ve been to the event’s “page” and clicking on the “Share” link, then pulling the short url out of that message. That results in a url looking like this: http://events.linkedin.com/pub/12514
Rather than one looking like this: http://www.linkedin.com/osview/canvas?_ch_page_id=1&_ch_panel_id=3&_ch_app_id=30&_applicationId=2000&appParams={%22from%22%3A%22my_events%22%2C%22go_to%22%3A%22events%2F12514%22}&_ownerId=2757022&completeUrlHash=gXn-

I assume the nasty url is a result of OpenSocial, in the sense that the hosting site needs to know which application to load and then pass info to the application – but since they are already creating url aliases, why not expose them more directly?

These suggestions aside, it’s a welcome addition which makes LinkedIn much more useful, especially to those not in job-seeking mode.

(If we’re not connected on LinkedIn and should be, here’s my profile).

However, based on the twitterstream (hashtag #gsp), sounds like lots of interesting presentations there:

• Charlene Li on the Future of Social Networks (CNet coverage)
• Charlene Li on the Future of Social Networks (the Slides on Slideshare)
• Beth Kanter’s notes (and her own slides on “Giving Good Poke: Using Social Apps and Media for Social Good)
• Chris Messina on Diso (slides)
• Slideshare tag gspwest2008 and tag gsp seem the best places to look for presentations to be uploaded
• Jeremiah Owyang’s notes on David Glazer’s presentation Open Social

I know I’m also missing lots of others blogging the conference – links in the comments appreciated.

It’s an important and powerful critique, though one with which I ultimately disagree. Greenfield essentially argues that:

1. Social networking applications must, necessarily, oversimplify human relationships: they couldn’t possibly represent the complex and dynamic nature of any graph connecting a pair of individuals, let alone the mesh of a whole community.
2. As a result, they inevitably create emotional distress, anguish, and pain for users (and sometimes even for non-users)
3. Therefore, we should not use them.

The problem, as Greenfield sees it, is that we’re allowing technical architectures to intrude upon the pre-technical, social space of human relationships. We’re allowing the web of human relationships as-modeled-by-software-systems to reduce, pollute, and corrupt the web of human relationship as modeled in the human psyche and history of culture.

Here’s one of the critical paragraphs of the piece, though you should read it (and the comments to it) in full:

What these commentators do not or cannot admit, though, is that the whole milieu in which these concerns of openness and portability are contained is broken – and not just a little broken, but badly so. All social-networking systems, as currently designed, demonstrably create social awkwardnesses that did not, and could not, exist before. All social-networking systems constrain, by design and intention, any expression of the full band of human relationship types to a very few crude options – and those static! A wiser response to them would be to recognize that, in the words of the old movie, “the only way to win is not to play.”

Greenfield takes apart the XFN standard, noting that it prohibits, by design, “negative relationships,” and goes on to assert that negative relations are critical to the social fabric. However, it is important to be able to keep some of those feelings (and their dynamic nature) to yourself:

social comfort and coherence require that by far the majority of actual feelings regarding the people in our lives not be made explicit. In my experience, any degree of smooth and compassionate human concourse absolutely requires plausible deniability, and a certain degree of dissembling regarding your actual, operative feelings for the people you’re engaged with, however much you love them.

Finally, Greenfield concludes:

I believe that technically-mediated social networking at any level beyond very simple, local applications is fundamentally, and probably persistently, a bad idea. From where I stand, the only sane response is to keep our conceptions of friendship and affinity from being polluted by technical metaphors and constraints to begin with.

It’s almost enough to make me shutter my Facebook account, but then it’s my move in Scrabulous.

My issue with Greenfield’s account, however, is that he assumes that simply not playing is a viable answer. Like Bartleby‘s “I would prefer not to,” Greenfield’s renunciation of all software-modeled relationships risks a slippery slope which ends in renouncing all online participation.

After all, doesn’t blogging software also create social discomfort and awkwardness which didn’t exist before? (Didn’t you read my blog post on X? I can’t believe the comment Y left on Z’s blog!)

It is vitally important to remember that there is (and will always be) a reduction inherent in transforming the complex and dynamic mesh that is human relationships down to a “social network” as understood by Facebook, LinkedIn, and the like – but I have to disagree that the only appropriate response to that reduction is to take my ball and go home.

Where social networking applications cause emotional pain we need greater education and contextualization. I don’t know about your teen years, but I was certainly familiar with artifical indicators of popularity and mechanisms of exclusion in mine.

This is not to say the mechanism of bullying, exclusion, and oneupmanship aren’t different in an online social networking world, but that we need to learn to understand, explain, and mediate those differences, not ignore the social networks and hope they go away.

Shindig first, direct from the OpenSocial API Blog “Let’s get this Shindig Started“:

Shindig is a new project in the Apache Software Foundation’s incubator (as per the formal proposal) that aims to provide an open source reference implementation of the entire OpenSocial stack — Shindig’s goal is to allow new sites to start hosting social apps in well under an hour’s worth of work.

This am was the initial commit to the Shindig svn repository. In other words, there’s already code, in the best open source fashion:

This commit represents initial versions of the first two components, the Gadget Container JavaScript and the Gadget Server — the latter written in Java. The Gadget Container JavaScript provides code to generate IFRAMES pointing to gmodules.com, offers some basic gadgets functionality (e.g. dynamic height), a layout manager, the edit dialog box, a cookie-based user preferences store, and an option to point IFRAMES at your Gadget Server instance instead of gmodules.com. The initial Gadget Server provides extensible scaffolding for processing gadgets: retrieving XML, parsing it, and processing it into a form that allows rendering of the gadget to a user or retrieval of its metadata.

I won’t likely have time today (or tomorrow for that matter) to dive into this, but it is great to have some actually code in advance of the holiday week.

The second project I’m excited about this morning is DiSo, which is Chris Messina, Will Norris, Steve Ivy and others working on a social networking platform “with its skin inside out,” starting with WordPress as a platform. It’s a chance to take the concept of using XFN, hCard, OpenID, OAuth, FOAF, and related microformats and open standards to create a truly distributed social network.

(See also GigaOm’s coverage from Tuesday which I just found through the news feed in my WordPress dashboard – The Next Social Network: WordPress)

Glad to see both of these projects kicking off in the transparency of the open source world – gives me good hope that we’ll actually make some significant progress on the social network portability front.