The Enterprise 2.0 Recovery Plan

by Andrew McAfee on December 1, 2008

Recent events in the news have inspired a thought experiment: I asked myself what I would do if I were put in charge of IT as part of the turnaround effort at a big US automaker. To be a bit more specific, I imagined that one of the big 3 American auto companies was taken over tomorrow by enlightened and aggressive new leadership whose only goals are to restore the company to operational and financial excellence. This leadership is enlightened (in my book) because it believes firmly in the power of IT to help businesses achieve their goals and differentiate themselves in the marketplace, and will fund and fully support whatever initiatives I propose (this is a complete fantasy for several reasons, but thought experiments aren’t supposed to be constrained by reality.).

So what would I propose?

I’d be guided by a couple facts and a few principles. The first fact is that on day one I would know virtually nothing about the company’s IT environment. I wouldn’t know, for example, what major enterprise systems needed to be deployed, integrated, consolidated, upgraded, etc. I also wouldn’t know about the health, status, and importance of the large projects currently underway. I’d set about trying to learn answers to these questions, of course, but this would be a long, slow process.

My colleagues on the new management team would be similarly in the dark on day one about other critical questions:

  •   Which of our current vehicle platforms under development will be hits in the market?  Which will be duds?
  •   Are our current platform projects largely on schedule, or are they falling badly behind?
  •   Where are our biggest opportunities to cut costs without losing valuable capabilities?
 The second fact (actually more of a very safe bet) is that the company would have a static and fragmented Intranet, and that employees would communicate with each other primarily via email. In other words, Enterprise 2.0 would not be advanced within the company, nor would it be universal.
As I got to work and tried to deliver results and benefits as quickly as possible, I’d be guided by a set of principles, many of which I’ve discussed in this blog:

  1. The company ‘knows’ the answers to our questions. The knowledge required to answer them exists within the workforce. This knowledge is widely diffused, constantly changing, and not contained in the mind of any single person (As Friedrich Hayek pointed out many years ago), but it is out there. Most executives, I’m pretty sure, believe this to be true. What’s frustrating them is that they don’t have great ways to collect and access this knowledge.
  2. Most people want to be helpful to each other, and to the company. I think it’s self-evident that people are largely good; if we weren’t, we would have wiped each other long before now. And we are to some extent wired for altruism and reciprocity. Finally, American carmaker employees have ample reason to fear for their industry, their company, and their jobs, so they have extra incentive to pitch in and help out, and to experiment with new ways to do so.
  3. Expertise is emergent. It’s logical and natural to think that all the good new product ideas come out of the design department and R&D labs, that the folk in the IT department are the best ones to help you with your computer problem, and that the engineers are the only ones who can figure out why the doors start rattling after 5,000 miles on the road. But this is not always going to be the case. The more we look, the more we see that a very effective way to solve a problem is to expose it to a highly diverse set of potential problem solvers, then let them have at it.
  4. People are busy. Most knowledge workers have more than enough to do with their normal jobs, and aren’t going to go too far out of their way too often, even though they do want to be helpful. This implies that any new tools need to be perceived as ‘in the flow’ of work, rather than ‘above the flow.’ There are a few ways to achieve this. One is to make the new tools extremely simple, easy, and intuitive to use, and to ensure that they’re never more than a couple clicks away. Another is to ‘widen the flow’ so that job descriptions include ‘enterprise-level helpfulness / collaboration.’ A series of three posts advocating this is here, here, and here.
  5. Weak ties are strong. Weak-tie networks are great places to look for novel information and introductions to valuable people. And social networking software (SNS) is a great tool for building, maintaining, and exploiting networks of weak ties. Instead of being a time-waster, enterprise SNS would be a powerful resource.
  6. The ability to convert potential ties into actual ones is valuable. At present we rely primarily on human brokers and connectors to introduce us to valuable colleagues. These organizational matchmakers are extremely valuable and influential, and there aren’t nearly enough of them.
  7. Platforms are better than channels , for a lot of reasons. Channels like email hide information; platforms like blogs, wikis, Facebook, and Twitter make it visible, persistent, and widely consultable.
  8. Search is the dominant navigation paradigm. People navigate online content by typing words into search boxes rather than navigating through menus. This implies that we should do everything we can to make sure search works well.
  9. The mechanisms of emergence should be encouraged. For search to work well, online content needs to be heavily interlinked. So people should be given the ability to link to content they find valuable and encouraged to do so. They should also be encouraged to tag, vote, rate, and to all the other things that help identify what a particular piece of content is about, and how good it is. In addition to this explicit work people also vote on and rate content implicitly as they browse through it. 
  10. Anyone can learn the new tools, but they need to be educated, trained, and encouraged. I do think that digital natives use technology differently than us older digital immigrants, but we can learn. The new tools of collaboration don’t require any skills beyond point, click, drag, drop, and type. They do require users to adopt a particular philosophy about sharing information and interacting with each other, and this philosophy can seem strange at first. When I first heard about Twitter, for example, I said something like "What on Earth would that be useful for, and who on Earth would ever want to use it?" Now, however, I’m a fairly frequent user, find it a really novel and valuable resource, and think that it has strong potential within the enterprise (here are my blog posts on Twitter, and here’s a research report on Enterprise ‘microblogging’ from Pistachio consulting ).

So what would adherence to these principles lead me to do? I’d roll out as quickly as possible a single integrated suite of emergent social software platforms (ESSPs) to all employees of the company. This suite would include blogs, wikis (including collaborative document production tools like Google Docs), discussion boards, SNS, a microblogging tool like Twitter or Yammer, a tagging utility, prediction markets, ways to vote on good content (a la Digg) and ways to give praise or good karma to particularly helpful colleagues. Lots of vendors both big and small are working to develop such suites; for now, I’m going to assume that a complete one exists.

As I wrote earlier, SNS helps with principle #5 above, and a blogosphere (broadly defined here to include a Twitterverse) helps with #6. And the whole idea of ESSPs supports #7. To put the other principles into practice, I’d insist that:

  • The tools be trivially easy to use, primarily by copying the look, feel, and user interface of the most popular Web 2.0 resources. Too many ESSPs intended for the enterprise try to reinvent the wheel, and they do so poorly. (helps with principles #4 and #10)
  • All content is cross-linkable, taggable, and Diggable. (#9 and #3 and #8)
  • The ESSPs contain some initial content and suggested structure, but that these are modifiable over time. (#4)
  • There be few initial rules or policy statements beyond ‘use your judgment’ and ‘highlight any behavior you find inappropriate.’ (#2)
  • Most platforms be available company-wide.  I’d probably allow only group collaborative document production tools to have limited membership. (#3)
  • Training on the tools be made mandatory for all employees (#4 and #10)
Of course, I’d also make them as device-independent as possible, and give people the ability to access them from home, the road, etc. Somewhat more controversially, I’d also make E2.0 part of every knowledge worker’s job. A series of three blog posts on this topic is here, here, and here, and generated a raft of great comments. I’d introduce this change by announcing on the ‘go live’ date of the new E2.0 suite that participation will become part (10-20%?) of everyone’s performance evaluation, starting in six months.

But what about principle #1 above, that the company knows the answers to the critical questions it’s facing?  This is perhaps the most important one, yet is not directly addressed above. To start to get answers, I’d set up prediction markets for the biggest projects, both IT and otherwise, within the company, letting people trade on whether they’ll be finished on time, nearly on time, or nowhere near on time. These markets would very quickly provide accurate and valuable information. I’d also set up markets to predict sales volumes and competitors’ moves.
I’d also start asking questions via my own blog, and listen to the comments and responses I got back. I’d work to create an environment in which people feel safe and free to speak the truth.

How would I know if Enterprise 2.0 was working well over time? The accuracy of prediction markets is easy to assess. It’s also easy to conduct surveys and find out if employees like the new tools, and prefer them to previous ways of collaborating. I’d also partner with academics to design and execute research investigating whether various attributes of performance improved after the new tools went in.
But the point of having the trust and commitment of my management colleagues is that I wouldn’t need to justify this kind of expenditure. If they’re on board with the principles then they’re on board with the investment required to put them into practice. And this investment is not huge. I’m pretty sure E2.0 wouldn’t cost as much as the typical ERP project at a car company.

I used Twitter to float this idea before writing this post, and a number of people responded that "IT isn’t the problem at the car companies!"  I totally agree. But technology can be a large part of the cure for what ails them. And I’m confident that the biggest and fastest bang for the IT buck at a US automaker today comes from ESSPs and Enterprise 2.0. 

Do you agree? Or do you think there would be better uses for investment dollars and managerial bandwidth after the hypothetical leadership change of my thought experiment? I don’t, but I’d love to hear your thoughts and reasoning if you disagree. Leave a comment, please, and let us know.

johnny won December 1, 2008 at 8:01 pm

I think your ideas of Enterprise 2.0 are very well put. It seems very likely that the teams that do exist at US Auto are in need of modernization. Perhaps management is holding back the producers but we can’t truly know.

But I must dissent a tiny bit. Realistically, it isn’t likely that the US Automakers are separated by a gulf from Nissan, Honda, BMW, Toyota, Hyundai in terms of enterprise functionality. Surely the management is guided by different principals but can we really argue that the enterprise teams, the knowledge workers are so fundamentally different that a 2.0 solution at GM could produce a Honda Fit killer?

A stripped out, burnt down organization being remade for modernization should certainly incorporate much of your Enterprise 2.0 suggestions but a bunch of Twittering, social-networking knowledge workers will go nowhere without an agenda. I think a great example of this is the Nano from Tata Motors. It was literally conceived as a concept by Ratan Tata and was created when much of the world declared it to be impossible (except C.Ghosn) to by produce a $2,500 car. (

While Tata did indeed use some cross-functional teams and much more integration than what was normal but they nevertheless achieved what many called impossible. The enterprise system may give voice to the knowledge workers comfortably at the middle of an organization but it seems very difficult that such a voice would change or dictate or establish the vision (or lack thereof) of leadership. Ratan Tata, Carlos Ghosn, Peter Schutz indicate that any enterprise organization is secondary to the leadership and their forceful mandate of achieving goals.

Of course if leadership used enterprise 2.0 and noticed that its’ knowledge workers thought the Chrysler Seabring to be the best car in the world then, leadership has an enormously different sort of struggle. (

Fred Johanson December 1, 2008 at 8:48 pm

Given that this entire post is an exercise in fantasy, I can’t help but feel that you are overlooking a critical constraint necessary to ground this experiment in some sense of utility: time. A successful E2.0 recovery plan would surely take years to implement in the best case. By the time, Fantasyland ESSP became the toast of HBR you would be back in the Gulfstream V on your way to DC for another $30 billion.

What are the typical timeframes for major corporate IT deployments? Decades? LetÂ’s assume it would be technically trivial to deploy social software tools on your network, you will still have to worry about how policies (or lack their of) may impact the thousands of suppliers, contractors, and management consultants that rely on your networks. Can my brake pad suppliers vote in the prediction market? Will I let my part-time design contractors Yammer away with my finance wizards?

LetÂ’s focus on our own morale challenged staff though. For this project to succeed, we are going to need wide and rapid adoption across the entire corporation. However, wouldn’t you lose all credibility after announcing that the future of the American automobile depended on blogs, wikis, tweets, tags, and predictions? Working out performance evaluation details is a lot more complicated than assigning percentages. The incentives for the new engineer in Gothenburg will surely be different than the marketing exec in Detroit. How long will this take to figure out?

The biggest and fastest bang for the IT back at a US automaker today – a company-wide email with an easy to use resume template attachment.

Barthox December 2, 2008 at 5:18 am


I’d like to come back on your poin’ #2″Most people want to be helpful to each other, and to the company. ”

I’ve been reflecting on that lately, and a friend of mine who’s a consultant at Deloitte has been confirming my thoughts in a discussion a week or so ago.

“In larger organizations, below the level of middle management, most people are not motivated at all by the succes of the company. They just want to make their job good enough so that they keep it and get paid at the end of the month.”

Now, this is harsh and probably not true for every organization, but I do believe that it is the case … and that is a really strong obstacle to the implementation of ESSPs (I love the acronym / term by the way!).

Julien Le Nestour December 3, 2008 at 9:00 am

Hi Andrew,

Lots of good points to your list, let me add some quickly, I’ll try to do a blog post for a more complete comment:

(I’ve posted a simple framework to support those comments here as a single slide)

– An integrated suite of ESSPs does not exist, at least not a good one.

– I would take advantage of the various cloud options available to switch hosting and drastically reduce costs. Along with this, “outsource” things like email by using Gmail for the enterprise or similar offering. Basically, looking at cutting costs on the infra layer without cutting features or even by adding new ones.

– Would also look at the office suites like Zoho to cut licenses cost.

– Open source: same commitment here, lots of potential cost savings without reducing the capabilities.

– For in-house applications, I would leverage RIA technologies like Air, in order to efficiently build applications which have a good UI, as well as an on- and off-line version.

– I would also focus more on the outside ecosystem of the company: your list is mostly inwardly focused, but lots of potential lies at the edge of the company.

Of course, an ESSPs suite would be at the core, exactly as you exposed it, however I would use the cost savings produced on the other fronts to get buy-in to be ambitious on the ESSP front.

Lee December 3, 2008 at 10:41 am

Hey Andrew,

(isn’t it interesting that comments are usually not prefaced with some form of salutation? but I digress…)

I feel that maybe this post should’ve been called “Twitter saves the Big 3″ or something to that effect 😛 While I always enjoy your thoughts, I think I am (almost) completely over the social media hype. We’ve had discussion forums forever and all the major “new” technologies are at the core not very different, if at all – people post messages. You can call it blogs, tweets, comments, whatever, it’s essentially all the same: people type in text that is associated with an online persona and widely viewable. Woopdeedoo. What seems to me implied in your whole fanatasy rescue plan here is that people by and large don’t have a problem with and actually want to leave electronic messages. And that’s where I think you are indeed living in a fantasy land. Forget about time – which is a great point – but people over a certain age, I would argue, just aren’t down with “socializing” by sitting in front of a computer screen and hacking away at the keyboard. The new generation of people I think will be a lot more comfortable with this, but until they grow up en masse and enter the workforce any e2.0 (god what a stupid name) platform will just fail – which brings actually brings us back to the time issue: maybe in 10, 20 years this could be a lot more successful, but now, I’d bet against it. So if you actually got to be head of IT and proposed to save the automakers by a social networking platform… not even sure what to say because that already sounds so ridiculous.

Social networking may be successful with geeks who _like_ spending all their time ruining their eyes staring at a big lightbulb, but I think the average joe walking down the street – or working at an auto company – would rather minimize their time sitting at the comp and would consider the idea of “socializing” online utterly nonsensical.

The real million dollar question I think is how could computer systems capture human knowledge without the knowledge possessor having to interface _directly_ with that system? Also, how can we reduce the time spend staring at a screen?

And finally, how social is it to sit silently in front of a screen typing in words??? Does that not increase our sense of separateness and otherness from everyone else, thereby strengthening the egoic illusion? But again, I digress.

Let me apologize in advance if I came across as insulting or offensive, this is surely not my intent… I’m just really can’t hear about these “social” utilities saving the day any longer.

– Jaded knowledge worker

Elliot Ross December 3, 2008 at 2:14 pm

Good Afternoon Andrew

With all the difficulty that the Detroit 3 are in, one area where they are often ahead of the pack is in their IT usage.

Ralph Szygenda of GM is the poster child of global IT sourcing.

Chrysler’s IT team did a gun point divorce from Daimlers IT systems at a rate no one thought possible.

The automakers have utilized technology to slash the idea to car concept by massive amounts.

Technology allows an engineer in Detroit, and one in California to review the same CAD files.

GM can tell in seconds when **one** of some 80,000 devices experiences a glitch.

Of the many failings they have, their IT teams have put in their share of improvements.


Elliot Ross

Frank December 5, 2008 at 5:02 am

Well, I did read your fantasy experiment and as experiments go, it seems really unscientific. Where’s the control?

I doubt if social networking could really substitute for a good ERP system. Collaborative payroll is probably still a long way off; even with “digital natives” controlling the company. When I hear that term I get this “Lord of the Flies” sort of image.

No, I don’t believe there will be salvation for the Big Three; but, if there was, it would be rooted in more traditional, “competition from startups” motivating them to further refine their structure, trim waste, lower prices and control greed.

vincent December 8, 2008 at 3:00 pm

Dear Andrew this was a nice idea but i’m not convinced at all with you plan.

Why you take the IT job and not the HR or strategic ? Why you seems to reduce the entreprise 2.0 to this IT approach.

You mess all the sociological approach, how to keep faith, motivation for all the staff in this time of crisis ?

I’m a litlle disapointed by your answer to this challenge

Samuel December 16, 2008 at 7:10 am

Nice post, Andy! While reading this I question myself: Could this approach also apply in general? I couldn’t find a reason why not. So, I think this plan could also be a general e2.0 adoption plan (of course ‘urgency’ helps adoption).

Raj Kumar December 18, 2008 at 3:13 am

At least 3 of the comments above express serious doubts about the ability of ESSPs to deliver a means for success. Borrowing a leaf from the second loop of learning – question the knowledge structure – I would like to go a step further. The self-confessed flight of fantasy by Andrew serves a very important purpose: it illustrates the ESSP delivery untrammeled by limitations. Thus the best that ESSP can do is publicize problems across the enterprise and make the available knowledge accessible. This reveals that its governing philosophy to drive success is flawed.

Success is not a product of the availability of the ‘right’ information (how to overcome biases?) but concerted action on reliable knowledge of patterns of behavior and an honestly shared understanding of reality. This is the only reliable means for protection from mistakes due to the common failure to see beyond assumptions and generalizations.

The decision makers in the auto industry need to go beyond the obvious in their fast paced competitive world for defining problems, solutions and strategy. They are not the first to have suffered for want of reliable means to grasp reality and apply knowledge. Ford had a taste of it when they disregarded the feedback from their own force on why their SUVs were turning turtle. Kodak experienced it when the hubris of being the best caused them to lose sight of the competition. Intel paid heavily for believing a remote error unearthed by an academic could be safely denied.

I agree with the evaluation that free flow of knowledge has the power to emerge the reality. Emergence is an attribute of free flow. However, ESSPs do not have the power to assure the feedback needed to provide this free flow of knowledge. Free flow requires total coordination irrespective of chaos, the assurance of constructive feedback, and categorization to organize content, place it in context and deliver it persuasively with reference to accumulated experience. It needs to be addressed to people engaged with the issue and requires a drive to follow up on action or response and progress the flow. This coordination, feedback, organization and drive requires time and energy that personnel are simply too busy or pre-occupied to provide. Like Fred Johanson has opined, it will take a near impossible culture in an impractical time frame for personnel to achieve this free flow. IT is of no help today because it cannot anticipate the unpredictable flow of Knowledge work to deliver categorization and coordination. ESSPs submit to the prevailing conventional wisdom that personnel must self-organize.

In conclusion I would like to point to AndrewÂ’s blog of March 25, 2007. There he uses Tom DavenportÂ’s gift with words to warn against techno utopianism: “…Most of the barriers that prevent knowledge from flowing freely in organizations – power differentials, lack of trust, missing incentives, unsupportive cultures, and the general busyness of employees today – wonÂ’t be addressed or substantially changed by technology alone. For a set of technologies to bring about such changes, they would have to be truly magical, and Enterprise 2.0 tools fall short of magic.” I would say it is time to get the philosophy right. The need is convergence of technology with free flow and I believe it is possible.

Rob February 5, 2009 at 2:02 pm

Very good ideas.. I always find assumptions the bottleneck to my own IT experiences when going to work in different environments. Get over the, 'this is the way things are done' menatlity and adopt an open question as to HOW things CAN be done.

Good article. I would also say that IT isn't the biggest problem of the automakers, but I agree about ESSP and Enterprise 2.0

Rob February 5, 2009 at 9:02 pm

Very good ideas.. I always find assumptions the bottleneck to my own IT experiences when going to work in different environments. Get over the, 'this is the way things are done' menatlity and adopt an open question as to HOW things CAN be done.

Good article. I would also say that IT isn't the biggest problem of the automakers, but I agree about ESSP and Enterprise 2.0

Ms Office 2007 June 16, 2011 at 4:00 am

Reading hard

Comments on this entry are closed.

{ 1 trackback }

Previous post:

Next post: