I Know It When I See It

by Andrew McAfee on January 27, 2009

More and more often these days I get asked “Does [offering X] from [vendor Y] qualify as an Enterprise 2.0 product?” Established vendors of collaboration software are modifying their offerings and repositioning them as social software platforms that have all the features and functions necessary to support the new modes of interacting and getting work done. Smaller companies and startups often say that the established vendors “just don’t get it” and that the new features they’ve incorporated —  blogs, wikis, discussion forums, tags, etc. —  are just windowdressing on products that are still essentially geared for Collaboration 1.0.

So who’s right?  Whose products at present come closest to enabling Web 2.0-style collaboration and interaction in enterprise environments? Honestly, I have no idea. It would be a huge amount of work just to learn about all the vendors and their offerings, let alone to evaluate them. And evaluating them “fresh out of the box” wouldn’t be that relevant anyway; collaboration software usually gets highly configured and tailored by companies before it’s turned on, with only some features activated and only some of those highlighted. One company, for example, might want its employees to be able to blog, while another wants no such thing.

Because of these facts I usually dodge questions about specific vendors and their offerings, and instead answer how I’d look at any particular deployment of collaboration software to see if it met my definition of Enterprise 2.0.

I find this pretty easy to do. I check to see if the environment meets three criteria: Is it freeform?  How frictionless is contribution? And is it emergent? To put some flesh on each of these terms…

Freeform means that the technology does not in any meaningful way impose, hardwire, or make and enforce assumptions about
– Workflows
– Roles
– Privileges
– Content
– Decision right allocations
Instead, people come together as equals within the environment created by technology, and do pretty much whatever they want.

I want to stress one more time that technologies that are not freeform are not bad or shortsighted or somehow deficient. If I were in charge of a retailer, for example, I’d want an ERP-ish system that my customer service reps used to take orders from customers, then send them to the warehouse for fulfillment and also to accounts payable. I’d want that system to have a standard order-to-ship workflow that was very hard to deviate from. I would NOT want this system to consider all employees to be equal, to assign them all to the same roles, or to give them equal privileges or decision rights. Finally, I’d want it to never accept a ZIP code that contained letters.

In addition to this system, though, I’d also certainly want to have one or more completely freeform digital platforms in which employees and other constituencies could come together as equals to decide what topics were important for the company, and how to attack them. My experience is that over time people place themselves into roles within these platforms (As USC’s Ann Majchrzak and her colleagues found in a study of corporate wiki users), but the important point is that they’re not assigned into roles up front, or by any external party.

Frictionless means that users perceive it to be easy to participate in the platform, and can do so with very little time or effort. One measure of friction is the total time required between having an idea for a contribution (while sitting in front of the computer, carrying the iPhone, etc.) and the appearance of that contribution on the platform.

Sign-ins, navigation through many web pages, and clunky user interfaces are all perceived as hurdles by a platform’s potential users, and increase friction. So does the need to massage a contribution like a blog post to look like it wasn’t put together by a complete hack. I can already tell, for example, that I’m going to post to this version of my blog more often than I did when it was hosted under the hbs.edu domain name and used a different and clunkier interface for posting. I felt like I had to tweak each entry for a long time to make it look OK, and it was a disincentive to post.

Tweetdeck, on the other hand, makes contribution to Twitter pretty frictionless. It sits on my desktop as a separate client, and I zip over to it whenever I have an idea. It’s quick and painless to send a standard tweet, a reply, a direct message, or a retweet, and to shorten and include a URL. With Tweetdeck I can convince myself to take a timeout from my deep academic thinking (coughcough) more often because each timeout is so short —  literally just a matter of seconds.

Emergent is both most intuitive of these three terms and the hardest to pin down. It really does bring to mind Justice Potter Stewart‘s famous yet unhelpful definition of obscenity “I know it when I see it.” My best-effort definition of the phenomenon is the appearance over time within a system of higher-level patterns or structure arising from large numbers of unplanned and undirected low-level interactions.

I wrote about the mechanisms of online emergence here and here; they include linking, tagging, friending (as on Facebook and LinkedIn), and following (as on Twitter). These are all activities that help patterns and structure appear, and that let the cream of the content rise to the top for all platform members, no matter how they define what the cream is.

Without these mechanisms, online content becomes less useful —  less easy to navigate, consume, and analyze — as it accumulates. With these mechanisms in place, just the opposite happens; the platform exhibits increasing returns to scale, and becomes more valuable as it grows.

The Web as a whole, and especially the Web 2.0 portions of it, is wondrously freeform, frictionless, and emergent. It stands as our clearest example of the kinds of energy and benefit that can be unleashed by the new technologies of interaction, and the communities that form on top of them. Some specific sub-segments of the Web like Facebook and especially Twitter are almost perfectly freeform and frictionless, but are less able to foster all forms of emergence. As I wrote here and here, it’s hard for me to use them to separate signal from noise and let the ‘best’ content rise to the top.

Too many corporate collaboration environments that I’ve observed, in contrast, come up short on the frictionless and freeform criteria. They make it far too difficult for prospective users to contribute, and they persist in slotting people into pre-assigned roles based largely on the formal org chart. In many cases they also impede emergence by having many small and mutually inaccessible environments, instead of one big one. The tendency to build walled gardens is evidently a deep-seated one, and one that should be questioned far more often than is currently the case.

Do these three criteria —  freeform, frictionless, and emergent — make sense to you and resonate with what you’ve seen and learned about the new technology-facilitated modes of collaboration? Am I leaving out any critical properties of effective technologies for supporting Enterprise 2.0? Do you agree that an enterprise’s chances of success rise significantly if it creates digital environments that have all three? Leave a comment, please, and let us know.

  • sengseng

    While I agree against creating “walled gardens” I think the ability to “pull the curtain” – via a private channel or backroom (ie in Twitter, the DM) – is necessary. An integrated and seamless ability to have some sense of security in the audience can increase the level of disclosure which in turn can foster a higher level of collaboration that can benefit the community writ large.

  • http://twitter.com/dagb dagblakstad

    I found this post excellent, and one I will revisit when developing new software. It will be a great help both for procurement and development of Enterprise 2.0 software.

    It could possible be mentioned explicitly the importance of the architecture and design of Web 2.0 software although it can be inferred from the post.

    I do not interprete the post to drop security. Sometimes there will be a conflict between openess and the need to keep things secret. But this is not typical anyway for the types of services this software is used for. If it is important to keep (some) things secret, the tings that is not secret can be open for collaboration. Often the perceived need for secrecy is not real, but promoted be people that thinks openness can hurt them.

  • http://www.capital-punishment.net/ Saqib Ali

    Excellent blog post Prof. McAfee. In addition to linking, tagging, friending, and following (in emergent), I would like to propose the ability to produce and consume RSS feeds. It seems like a basic thing, but too many so-called enterprise 2.0 products that lack the ability to produce “meaningful” RSS feeds. And most of the e2.0 wiki style collaboration suites have very limited RSS consumption capabilities. A full-features AJAX based RSS feed reader (something like Google Reader) should be part of any e2.0 suite. Getting email notification via email is sooo 2008. Users need to now switch to RSS feeds and RSS readers, but for this to happen the e2.0 platform first need to produce meaningful RSS feeds and at the same time provide a way to consume them.

  • http://twitter.com/JoeSchueller Joe Schueller

    Interestingly, the degree to which the platform emerges as freeform, frictionless, and emergent depends on how well or poorly those “highly configure” for their environment and their goals with the platform.

    In an IT management culture that expects and exacts control, insists on immediate pay-out numbers and enforces every existing policy, then you can pick the best platform in the world, but you're still pretty much screwed.

    I'd add OPEN to your list of evaluation criteria. The ability “mash” data from the platform in and out of other enterprise content/data repositories via well known and easy to use API's is critical. Without it, you face an uphill battle from all your existing IT projects. With it, you look like help. They don't have to put in their own “2.0” features, they just leverage your platform's capabilities in their apps. I guess this is the developer's version of frictionless.

  • cl

    Great post and key themes to hit on. I like the terms and the agree with the role the environment plays in contributing. That said, the challenge in reality is to balance real world concerns of liability and complex org structures within a single environment. Legitimate use cases exist for walled gardens of collaboration as well as wide open spaces. In extremely large organizations, this is a constant struggle.

    Finally, IT management matters here too. Few would yet consider enterprise 2.0 core to their success (yet is a key word here) and thus the temptation to keep costs down.

    Established vendors make people feel comfortable and are making headway in allowing us to leverage existing investment, but still come up short on the usability end. So, is cheap, configurable, and easy to use too much to ask???

  • http://aboveandbeyondkm.com/2009/01/km-should-be-an-absent-clockmaker.html KM Should Be An Absent Clockmaker | Above and Beyond KM

    […] run itself.  I found myself thinking about this as I read Andy McAfee’s excellent post, I Know It When I See It.    In his post he describes the criteria by which he determines  whether a collaboration […]

  • http://www.socialmedia-academy.com Axel Schultze

    Excellent observation, needless to add any more appreciation. !!!!
    But I’d like to change the question or better said ask those who ask it to rethink their question. Feature list, security, manageability, with or without this or that tool… Tech people (I'm one of them) are conditioned over 30 or so years to think in those terms. We are conditioned to launch, deploy, integrate… We are conditioned to value software or any other tool by putting it in the perspective of our company and then we try to “see” how our teams would deal with it, adopt it, how it would help further optimize business processes, how the reporting would do for the leadership team and so forth.

    B U T

    The whole point about social media lays in SOCIAL and MEDIA. If we think SOCIAL in a business context we think customers, partners, alliances, suppliers, vendors and many other constituencies of our respective market. If we think MEDIA we think in ways how to connect, interact, communicate, respond – faster, better, more meaningful. We think in overarching attributes like Trust, Advocacy, accelerating skill development or co-creation. This all has one mutually beneficial purpose: Businesses want to create the products and services their customers are interested in, in order to be more successful. Customers want exactly the products and services THEY need to be more successful. Social Media provides the bridge to cross that chasm.

    And if that is the PURPOSE of social media in and for the enterprise, the more important question in the first place is what enterprises need to DO to follow the changing behavior of their customers. The resulting question is what the tools need to do in order to achieve those objectives. And then, and only then, features and brands will emerge as a SOLUTION for said challenges. Hence the question “what kind of tools ….” Is the wrong question and we should help those people understand that the starting point is neither the tools nor their capabilities in security or manageability but what goals and objectives need to be achieved and who are the player in that concert. Once objectives are identified, that target audience is understood, the features will unfold, and security and manageability requirements establish themselves.

    Unlike any other emerging technology before back to the beginning of our industrialization, this technology [Social Media] is no longer a tool only for the “company” it as a tool collectively used by an entire eco system – and if not we are back to internal IT, which Social Media most certainly is not.

    My 2 Cent

  • http://elliotross.wordpress.com Elliot Ross

    If a tree falls in the forest…..

    Freeform as a concept is fine.

    But can it exist without enough structure to generate value?

    If six folks in marketing have blogged and Wiki'd their way to an amazing idea, until that can be structurally harvested,

    … does it make a sound?

  • gyehuda

    I agree with what you wrote — especially the parts you stressed. But I think there is more to the qualification that would help people call it out when they see it. Part of the “I know it when I see it” argument is that I know how it makes me feel (but I might not want to express that feeling explicitly). The feeling is a reflection of a change. In the case of Enterprise 2.0, I think there is a feeling of a power shift or improved flow of information. And this is much like the Theory X vs. Theory Y (http://en.wikipedia.org/wiki/Theory_X) recognition of different roles that managers and workers play in work situations. The difference really makes a difference. The moniker “2.0” implies a new stage, not just emergent, but transformative. If the use of the tool does not really change the workplace, I'd say it falls short. I'd say for something to earn a “2.0” badge, it has to make a notable change (improvement). A blog that provides real insight, and when read – people consider new ideas. A wiki that helps bring transparency to projects. A social network that allows workers to discover people they should be working with. Social bookmarks that transform the act of “searching” into “finding”. etc. Whether from the large familiar vendors, or the smaller innovative ones, the goal is to demonstrate value to the business, otherwise they won't spend the money. Onther way that E2.0 differs from W2.0 — Since someone else has to be convinced to spend the money to support the purchase of these new tools, by being convinced of the improvements to workplace efficiency.

  • http://blog.tarn.org/2009/01/28/challenging-the-orthodoxy/ Challenging the orthodoxy « Enlightened tradition

    […] of how wikis might work. In this connection, I was very taken with Andrew McAfee’s simple approach to evaluating Enterprise 2.0 products: I usually dodge questions about specific vendors and their offerings, and instead answer how I’d […]

  • http://sloanreview.mit.edu/improvisations/2009/01/28/what-defines-enterprise-20/ Improvisations « MIT Sloan Management Review » What defines Enterprise 2.0

    […] offering an “Enterprise 2.0″ product? Andrew McAfee of the Harvard Business School has an interesting blog post on that subject this […]

  • http://aboveandbeyondkm.com/2009/01/if-you-love-somebody-set-them-free.html If You Love Somebody, Set Them Free | Above and Beyond KM

    […] reading Andrew McAfee’s discussion of the key criteria by which to judge whether a collaboration software deployment meets the requirements of E2.0 and Alexander van Elsa’s post on how old-fashioned business models are holding back Web […]

  • http://www.deskaway.com Priya

    I think these points slowly kick in. initially a newbie to web 2.0 would take up a frictionaless service like a messenger, tweetdeck as mentioned. only later will he start going for more emergent stuff like subscribing, using google reader etc..


  • http://sis.net.in/2009/01/30/a-restaurant-or-a-web-20-service/ A restaurant or a web 2.0 service?

    […] I Know It When I See It […]

  • Gloria

    Hi Andrew,
    I found your post on Web 2.0 collaborative criteria very interesting and your definations especially helpful. I have found the strongest resistence often comes from organizational bosses who fear they will lose control of information. Issues like: lack of trust, loss of control of information, liability, etc. are frequently mentioned. Web 2.0 is functioning as the great equalizer and minds are flocking to it, “freeform', 'frictionless' and ideas are 'emerging 'as multidisciplinary minds meet and discuss ideas. The purpose of introducing Web 2.0 was to help keep up with the information age and encourage collaborative work, I believe it is doing just that, in fact, if an organization does not encourage the criteria that you have so aptly defined, I think knowledge workers will collaborate outside of the organization and that is knowledge lost to a firm.

  • tunghoy

    I read of your mnemonic of SLATES, to describe what web 2.0 features are. Which is how I ended up here, since I use web 2.0 tools for site development.

    As I understand it, the first S stands for search…..but there doesn't seem to be a search feature on your new site. Since there is a search feature built into WordPress, someone probably removed the PHP tag from the template.

  • http://itsinsider.com/2009/02/02/enterprise-20-meets-reality-tv/ ITSinsider » Blog Archive » Enterprise 2.0 meets Reality TV

    […] that have incorporated what we know to be Enterprise 2.0 technology and business practices as defined by Andy McAfee and Dion Hinchcliffe. I’m also particularly interested in enterprise […]

  • http://www.accmanpro.com/2009/02/05/what-if-2/ What if? | AccMan

    […] a particular social media project will lead to favorable outcomes. Way more. And it’s hard. Even Andy acknowledges: Emergent is both most intuitive of these three terms and the hardest to pin down. It really does […]

  • http://enterknowl.blogspot.com Simon Carswell

    I think one of the problems with of Sharepoint is that it attempts both to be a structured, workflow type system at the same time as a Web 2.0 collaborative one. Whilst it works tolerably well in structured mode, it fails all three of your test for a Web2.0 / Enterprise 2.0 platform.

  • jaycross

    Andy, I would throw in another aspect of frictionless: does it fit with the way I already do things? I'm a lot more likely to use a technology if it doesn't require adding yet another app or destination to my work process. For example, I'm more likely to read an RSS feed that shows up in my email than one that requires me to go to a separate reader.

    As for the point of the overall exercise, whether an app is 2.0ish is secondary to whether it works well for the business. In addition to freeform, frictionless, and emergent, that takes the ability to establish one's identity through personal profiles, archives of past activity, and the like; it addresses the nodes as well as the connections.

  • http://www.tccgd.org/andrew-mcafee-i-know-20-when-i-see-it.html/ Guide to Study » Blog Archive » Andrew McAfee: I know 2.0 when I see it

    […] it 2.0? Harvard B-School’s  Andy McAfee looks for three things: Is it freeform?  How frictionless is contribution? And is it […]

  • RTodd

    I suppose my thoughts on this definition are around the basis of the construct. That is to say, do we define Web 2.0 / Enterprise 2.0 applications as an open inclusive definition or one of closure and exclusion. For example, one the first comments by Mr. Carswell is that Sharepoint is excluded because it imposes structure. While this is true, isn’t it true that blogs do the same thing with structured fields for the title, comment, tags, and date of publication. Sharepoint allows the user to utilize this structure but also allows for organizational customizations. So what happens to products that pass two of the three tests? Do we exclude products that are performing the basic function of collecting that unstructured conversations which is what E2.0 tools are doing? I guess I am still old school in that I define Enterprise 2.0 technologies as those that enable, encourage, and facilitate Many to Many Communications. Reguardless if they are collaborative or social, if they pass this basic test.

  • http://www.avandialawyer.net avandia lawyer

    well after reading you blog i should say I Know It When I read It!!!

  • http://blog.enterpriseupdate.org/2009/02/andrew-mcafees-three-prongs-of-enterprise-20/ Andrew McAfee’s three prongs of enterprise 2.0 @ Enterprise Update

    […] I need to noodle down on his dimensions myself, but for now I thought I’d link you over. Take a look and let Andrew (or me) know what you think! Trendspotting Andrew McAfee, enterprise […]

  • http://algarabias.wordpress.com/2009/02/10/criterios-para-identificar-un-entorno-de-aprendizaje-20/ Criterios para identificar un entorno de aprendizaje 2.0 « algarabias

    […] en su blog dedicado a la relación entre las nuevas tecnologías y el mundo de los negocios. El post de McAfee establecía tres criterios que, desde su punto de vista, tienen los entornos 2.0. Según él, estos […]

  • http://blogit.helsinki.fi/tyonimi/ Jere Majava

    I think your list of criteria works well also as a general definition of social software. It's kind of ironic that often the versions of social software that are targeted specifically for enterprises have “features” that make them less freeform than the original ones: many levels of user privileges, hierarchies and predefined categories instead of tags, etc. This in turn both causes friction for the use of software and dampers emergence.

  • http://www.online-pharmacy-adviser.com/avandia-rosiglitazone.html buy avandia

    Avandia (rosiglitazone) for Diabetes, Type 2: I have been taking 4mg once in morning for last 8 or 9 years without any problems. Weight is staying stable within 5 lbs /- for the last 8 or 9 years.

  • http://www.breastpumpdeals.com/medela-freestyle-breast-pump.html freestyles

    This is really very interesting post, I have learned a lot, keep interesting posts like this one coming, bookmarking you straight away, will be back again.


  • http://www.tourtravelchina.com/beijing-tour.html Beijing Tour

    It could possible be mentioned explicitly the importance of the architecture and design of Web 2.0 software although it can be inferred from the post.

  • medelabackpack

    Hi Ross, it's an interesting question from a genius mind of yours!


  • http://andrewmcafee.org/blog/?p=928 How Beautiful it is, and How Easily it can be Broken : Andrew McAfee’s Blog

    […] deployed. They had better be appropriate to the goal(s) of the initiative, and they had better be freeform, frictionless, and emergent. If the tools offered are clunky or annoying, they will be tossed aside and people will just keep […]

  • http://fredzimny.wordpress.com/2009/07/07/how-beautiful-it-is-and-how-easily-it-can-be-broken-andrew-mcafee%e2%80%99s-blog/ How Beautiful it is, and How Easily it can be Broken : Andrew McAfee’s Blog « Fredzimny’s CCCCC Blog

    […] deployed. They had better be appropriate to the goal(s) of the initiative, and they had better be freeform, frictionless, and emergent. If the tools offered are clunky or annoying, they will be tossed aside and people will just keep […]

  • pixbook

    Ways to make money

    Great article. Thanks for sharing this.

  • http://www.mbopmegashop.com Online Shopping Mall

    Ha ha Avandia, you're so right!

    Mbop Megashop

  • http://www.mbopmegashop.com Online Shopping Mall

    Ha ha Avandia, you’re so right!

    Mbop Megashop

  • http://www.ticketloot.com/george-strait-tickets George Strait Tickets

    I think the ability to “pull the curtain” – via a private channel or backroom (ie in Twitter, the DM) – is necessary. An integrated and seamless ability to have some sense of security in the audience can increase the level of disclosure which in turn can foster a higher level of collaboration that can benefit the community writ large.

  • http://idotamaphack.blogspot.com/ Ronnel@Dota Map Hack

    Now I know the meaning of emergent, frictionless, freeform thank you very much for this article. thumbs up!

  • http://mytrusted-best-ptc-sites.blogspot.com/ Best PTC Sites 2011

    I still developing my website 2.0 through wordpress, it seems this will help me alot, Thanks Sir Andrew for very informative article.

  • http://restaurantcitymagics.blogspot.com/ Restaurant City

    I also prefer wordpress 2.0 so i enjoyed reading your aticle Mr.Mcafee thanks a lot….

  • Emily Brown

    Choosing specific vendor for your apps or software are actually not bad, as long as you know the vendor that much about the quality of their products….but sometimes we need to seek for anther options and get updates   for you to have the best of what you needed….it’s simply like loving  Bruno Mars’ songs while keeping updates with Justin Bieber‘s hottest release…. to keep yourself updated with the current trend.

  • http://emttrainingauthority.com/ Manni

    freeform, frictionless, and emergent makes sense on what I’m trying to do right now.

  • arun baby

    thanks for the valable information

  • arun baby

    thanks 4 the info

  • Jaanus Ringmäe

    Nii ja naa

  • Stox

    I’m almost finished with Enterprise 2.0. Excellent read, btw. One of the case studies mentions a company that used Facebook as their ESSP and encouraged their employees to contribute work-related content via their personal profiles, which naturally included their personal content as well. Having heard (a very little) about Google’s “Circles” construct, it seems to me that this would be a much better method of encouraging collaboration, even though it would divide content into separate “gardens” as mentioned above. The positive effect of this division would likely be trust. Once contributors are confident that they can filter their content in the same way they filter their verbal speech (you wouldn’t say the same things to your boss as you would to your sarcastic college buddy over a six-pack) they will be more likely to contribute work-related content. So, in certain cases, I think some carefully crafted divisions would be beneficial to the E2.0 idea. The effect on the whole enchilada (enchidata?) would have to be considered in the light of searchability.

    Btw, can you elaborate on the difference and definitions of ‘searchability,’ ‘discoverability,’ and ‘emergence’? It seems like ‘searchability’ and ‘emergence’ have a strong definition, but also that ‘discoverability’ seems to be a duplicate of either or both?

    Thanks in advance.

  • Sasa Garelovic

    Excelent post.

  • Sasa Garelovic

    I think neobux is the best ptc ever.

Previous post:

Next post: