Enterprise 2.0 = Emergence Software
There is something very interesting happening in the field of enterprise technology. I called part of it Web Office. Ismael Ghalimi called it Office 2.0. Ross Mayfield calls it Social Software in the Enterprise. Dion Hichtcliff calls it Enterprise Web 2.0.
Most famously, Dr. Andrew McAfee of the Harvard Business School called it Enterprise 2.0 in an article published in MIT-Sloan.
Collections of many Davids triumphing over a few Goliaths
The best book I have read in the last few years is by Steven Johnson. It is called “Emergence: the connected lives of ants, brains, cities and software“.
As an economist, it appealed to me right away. The idea behind the notion of emergence is simple: Individual agents, self motivated and operating according to simple rules, can suddenly produce patterns. In certain circumstances, these patterns can evolve into behavior that is intelligent.
In the free market economy, those individual agents are entrepreneurs, companies and consumers. In a democracy, those agents are voters.
Johnson’s stories about ants are amazing. With no central plan, and no ant in charge (the queen never orders anyone to do anything), the ants bump into each other randomly. But they have standards of interaction. Ants sample the other ants they meet to see what they are doing. They use pheromones as a standard way of communicating. If an ant meets too many Ants on Midden Duty (ie clean up), and that ant is on Midden Duty, it will switch over to Scavenger Duty. Out of this system, the optimal number of ants are always working on the right thing for the colony.
In McAfee’s article, Enterprise 2.0: The Dawn of Emergent Collaboration, McAfee says “When I use ‘Enterprise 2.0′ as an adjective, I mean “supporting of emergent collaboration.”
It takes only three simple notions
- - Many self motivated individual agents
- - Standards for interaction
- - New robust small scale technology used by the agents
It is decentralization: Niel Robertson’s quick summary of Enterprise 2.0 is “IT without the CIO”.
I believe what McAfee is saying is that everything new and interesting in the Enterprise isn’t necessarily emergent. If IT builds an AJAX application that must be used by end users to account for their time, there is nothing emergent about that system. Therefore it isn’t Enterprise 2.0.
Enterprise 2.0 is about decentralization of responsibility. This requires a completely different way of managing people. That is also why a Harvard Business School professor is so interested in it.
If you want bottom up corporate intelligence, you also have to give people new types of technology. Most importantly, that technology has to help them build their own solutions. You need to expect your knowledge workers to be creative. You need to expect them to be Innovation Creators.
That is why I agree with McAfee that MR Rangaswami at Sandhill.com has got it wrong when he says
“Enterprise 2.0 is the synergy of a new set of technologies, development models and delivery methods that are used to develop business software and deliver it to users.”
Sorry MR, but Enterprise 2,0 isn’t about building solutions for end users. Enterprise 2.0 is about building tools that end users can leverage to build their own solutions.
Out of those highly customized end-user built combinations of people, process and technology, will emerge better business practices. Better because they will be more intelligent, more flexible and they will generate more long term competitive business advantage because they will generate more innovations.
To be truly useful, these tools have to plug into the back end of any corporate entity. Critical features will include audit trails, access control, version control, authentication, provisioning and backup. The best Enterprise 2.0 systems out there will have thought through these issues.
An Aside to the Lotus Notes Crowd
Lotus Notes tried to do all of this but (at least up to version 6.5) ultimately failed to create something that was easy enough for all knowledge workers to use as a platform to build their own solutions. I only have experience with 6.5.
That’s why people make good money being Domino Developers. Making a business application in Lotus Notes is not drag and drop simple. You have to code. That is too much of a barrier for most end users.
However, new Enterprise 2.0 tools such as teqlo, iUpload, Blogtronix, SocialText, Mobilize, BEA’s Project Builder, Microsoft’s Excel Services, Atlassian and Itensil can all be used by knowledge workers to build various kinds of enterprise class solutions.
None of these systems can be used to solve every single problem faced by a knowledge worker, but the point is that knowledge workers can use every single one of these tools to build emergent solutions.
The critical differentiating factor is ease of use.
These simple Enterprise 2.0 tools are also why I am not too concerned about Lotus Notes applications with 100,000 lines of code. Porting the application might be hell. But rebuilding it with teqlo could end up being trivial.
When I was in grad school, a computer science Prof told me about the angry reaction his fellow IT engineers had when he first started showing people a drag and drop IDE. “That’s a con job…. it can’t be that easy” they said about dragging and dropping a UI
element onto a form. Until that point, people had to hand code the co-ordinates of what element went where. The amazing thing about these new Enterprise 2.0 tools is that they really are drag and drop easy.
Lotus Notes Options for the Future
Please note that I am sure that IBM has lots of options with Lotus Notes / Domino. Further, given all the brilliant people that work for IBM, I am sure that IBM is working on the ease of use issues. However, just like Microsoft, IBM has its work cut out for it, because the companies I mentioned above are going to use Port 80 to sell directly to business end users. SalesForce.dot has proven that this approach works.
Technorati Tags: Enterprise 2.0, Enterprise+2.0, innovation, Enterprise+Irregulars
Nice article Rod,
Under Enterprise 2.0 and the dynamism of knowledge workers implied, how is an enterprise supposed to balance the needs of its workers with its requirements to submit itself to governmental regulation?
In my industry, P&C insurance, we have compliance requirements with Health Insurance Portability and Accountability Act (HIPPA), Gramm-Leach-Bliley Act (GLB), the Patriot Act, Sarbanes Oxley Act (SARBOX), various other federal and state regulators, our own trade industry mechanisms, etc. Most of these regulations deal with privacy or transparency or accountability for our employees’ actions. A huge amount of our IT effort is applied to produce privacy controls, audit logs, and control measures to meet these requirements.
In another industry where I previously worked, big pharma, their systems, even non-production systems, must under go extensive validation processes prior to their usage under Food and Drug Administration pharmacuetical Good Practices (GMP, GLP, GDP, GVP and GRP.)
It seems to me that your take on E2.0 requires a more ad-hoc approach to the creation of these systems which lies directly in the face of the regulatory demand for extensive structure…how do we meet both goals at the same time?
Andy,
Thank you for the question. I am most familiar with SOX, so, if it is OK with you, I’ll confine the answer to that.
Today, most “knowledge work” gets done with MS Word, Excel, PowerPoint and Email.
The problem is that there is very little audit trail of who made what changes.
An enterprise class wiki, for example, records every single change people make. Thus, if a wiki was used to build financial reports, for example there would be a great trail of everything, making SOX compliance much easier to demonstrate.
Today, most 10Ks or 10Qs financial reports are built with extracts from legacy systems such as the G/L plus lots of email, Excel and word docs floating back and forth.
Itensil takes the Wiki thing one step further, but letting end users design formal processes on top of the Wiki.
I write a page, with maybe an Excel file, and then forward it on to the next person as “To Do”.
Rod
Rod, you have an interesting idea, and I agree that sometimes the barrier to entry is too high. How many times have we heard that users don’t know what they want until they see something, and I know I often ask a user to draw out what they want on a piece of paper or a whiteboard so I can see their point of view.
The issue as I see it with allowing ad hoc application creation is you end up with everybody creating their own silos of information. Jane in Accounting creates a document library database for her sales vs. budget spreadsheets, Debra uses iUpload for her customer service teams, and Bill in marketing is using Salesforce.com to track leads. Eventually you reach an inflection point where people want all this tied together. Creating something coherent out of it becomes an issue for IT. Why? Because they’re the programmers. It doesn’t matter that there was no formal design or review process, this is Enterprise 2.0 baby, we don’t need no steenking oversight!
I think a more reasonable approach is to involve more end users in the design and specification processes to help ensure their needs are met. I also think more attention needs to be given to making systems extensible and customizable by end users. That would address many of your complaints about end users having inflexible systems while still allowing a company to maintain some level of control.
It does require breaking away from a strict waterfall model and embracing a prototyping, agile or extreme programming model. And it also requires a forward-thinking management and development team who are willing to approach every project with the idea that the end user is the customer and the goal is to solve their problem. It’s not about the technology, it’s about the solution.
P.S. In Notes (prior to 7.02) a blog-style database is as simple as File > Database > New (or just Ctrl+N) and then select the Discussion template. If your primary intent is setting up a review cycle for attachments, you can use the Document Library template. Either of those will have at least 75% of what you’re describing, and both have decent web frontends.
P.P.S. I think you meant Salesforce.com instead of Salesforce.dot.
You are saying that the integrity of public companies’ 10K & Qs rely on spreadsheets?
Coming from a manager at E&Y (a big four ACCOUNTING firm), I find that to be an amazing statement.
Even my little private $25M company has an accounting system with more integrity than that. Our accounting system cost us $26K for the license, probably $50K for the implementation, and $6K/year for maintenance. Our individual company trial balances are produced directly from those companies’ GLs, and the consolidated financials for the holding company are an automated summation of those trail balances (though a tool called FRx that MS now owns) of those subsidiary companies. Once the controller and CFO ensure that the trail balances balance, we close the period and produce a controlled set of financial statements. If the trial balances don’t balance, it is generally pretty easy to figure out why due to the nature of the integrity of the systems.
Now your telling me that companies that spend MILLIONS of dollars for consolidated financial reporting systems rely on spreadsheets? Spreadsheets have absolutely no integrity because they typically rely on a human to enter a number.
To extend your example, in order to have the type of integrity my current SMB1.0 company has in its financial documents, my E2.0 Wiki would need to pull its data from my back-end system, understand the chart of accounts, and understand the relationships between our subsidiary GLs (we have three companies, one of which operates under Statutory Accounting Practice SAP instead of GAAP so we also have a fourth GL that makes a reconciliation between SAP and GAAP so that we can produce a GAAP consolidated statement) because in E2.0 these types of systems should be modular/plug and play for our knowledge workers to assemble themselves so that we can complete our month end close faster or with more integrity?
Let’s learn from history, else we’re destined to repeat it. To take the examples you bring up, way before Web 2.0 — in fact, way before the web full-stop — there were application “platforms” out there expressly designed to let end-users do their own thing.
Amusingly, one of these was called “Lotus Notes” — read up on Ray Ozzie and his vision for this thing he created (itself based on a system called PLATO) if you don’t believe me.
Now of course, and as you point out, if you want a decent functional Notes / Domino application, you need to engage the services of a developer — it’s gone way beyond the “end-user-tweaks” development model.
How do we ensure that this doesn’t happen all over again with the so-called “Enterprise 2.0″ stuff? Don’t forget that Notes and co. became what they are due to feature requests from end-users — I can’t see how you avoid this.
Don’t get me wrong, I’m not dissing the new wave of enterprise-focussed apps (I use — and love — Confluence for example). I just don’t think it’s the sweetness and light you portray.
Rod, I’m not sure that Enterprise 2.0 is the right term for what you’re advocating. How about Anarchy 2.0?
It seems like you think users don’t need a CIO and that they can do just fine developing their own applications – given the right tools. Users can buy the tools over the web and then design and run the applications over the web. Where’s the data? Is it safe? Is it backed up? What about the additional data charges or bandwidth charges in the small print? What if the hosting company goes bust? Will the price go up next year?
Now multiply that by ten because different departments will decide on different tools. Now wait until one of the new “super users” leaves the company. How well is the application documented?
The old legacy systems weren’t that cool but at least they talked to each other. Procurement didn’t have to type all the requisition data into their orders because Engineering had already done it. Now that Engineering have gone out and bought their new system, it doesn’t talk to the Procurement system “because it’s a legacy system”. Maybe that would have required coding and Engineering can’t do it.
And why should they? They’re Engineers. There’s a reason why good Lotus Notes developers earn a lot of money and it is because they have experience. They know what works and what doesn’t. They wouldn’t dream of trying to design a bridge (even if they had a good finite element analysis tool) because they are not Civil Engineers.
The problem is that IT may now not have the skills either for the myriad of new tool interfaces or APIs.
Now I agree 100% that e-mailing spreadsheets around for update by multiple people is bad. There are better ways. Lotus Notes in an Enterprise can be good or it can be bad. It depends very much on how the company designs or buys applications. Do the users have a say in the features that their application needs? Do they get involved in testing? Do they get enough training? But that is the same for any software.
Saying that “Lotus Notes sucks” won’t make your company more successful. Constructively trying to improve bottlenecks in your company’s applications just might. You never know, IT might even let you loose with a copy of Designer!
I think the CIO job is safe for a good few years yet though. Someone needs to keep a grasp on the overall strategy and keep the auditors happy.
There’s a reason why good Lotus Notes developers earn a lot of money…
They do? You know, that’s the umpteenth time in the past few days I’ve heard about “expensive” Domino developers.
I don’t know any Notes / Domino developers who earn a lot of money. Sure, maybe back in the heyday of IT contracting (late 1990s here in the UK), but not now.
Anyway. The other points raised here are good ones: chaos might be good for the power users, but what happens when everything has to be linked together? Well, allow me to blaspheme: I don’t think it’s an issue.
RSS is the new protocol for intranets, didn’t you know?? And of course, most of these new-fangled Web 2.0 things come with lightweight APIs over http. Integration at the data / service layer isn’t as tough as you’d think with these disparate tools. The real issues come about with that old nightmare: authentication / single sign-on.
Rod,
Regarding your comment above:
“Today, most “knowledge work” gets done with MS Word, Excel, PowerPoint and Email.The problem is that there is very little audit trail of who made what changes.”
Your suggestion of using a Wiki as a solution would require users to round-trip their existing documents from Microsoft Office (where they have been created) to a Wiki (converted to a web format) for auditing and possibly exported back to a Microsoft Office document at the end of the document lifecycle. The Wiki part requires users to step into a new environment with severely limited functionality compared to their existing Office authoring environment (and of course, the lack of offline access).
As compared to this, wouldn’t it be far easier to simply add auditing, access control and transparency to your existing Office documents without requiring users to abandon these application - we have made a first attempt here - Live Documents - perhaps you can take a look and offer feedback.
Cheers,
Sumanth
Thank you everyone for the great comments on this post.
Charles,
You are right that the enterprise has to give structure. I could not agree with you more. If you dig through my white paper called Turning Knowledge Workers into Innovation Creators, you will find a description of the kind of structure I am talking about. Here’s a link of the related blog post: http://www.innovationcreators.com/2005/12/conclusion_innovation_in_an_em.html
For exmaple, in enterprise blogs, it is a good idea to have “blog types” such as People Blogs, Project Blogs, Client Blogs, Product Blogs. The result are defined just enough to give people the rules of the road, and let them create from there.
To Rob Wills point about Anarchy 2.0, there will always be a very important role for the CIO and the IT team.
I just think that this technology means that the role is changing. Instead of being in charge of a hiarchy of defined solutions, the CIO takes on the role of the standards used to govern interactions. Specifically things like what access control system to use, what microformats to use and how to define interactive processes. And of course, the CIO and his or her team need to provide the infrastructure upon which all these fancy new solutions are going to be built.
BTW - I never said “Lotus Notes Sucks”. I did say that Lotus Notes tried but ultimately failed to create something that was easy enough for all knowledge workers to use as a platform to build their own solutions.
New technology, such as Teqlo, has a good chance of solving the problem. The potential result will be very powerful.
Finally, Andy Broyles, I would like to follow up on your question about 10Ks and 10Qs. First, I am not an accountant. I do not have a CPA. And, on this blog, I can not and do not speak for my employeer.
However, I can point you to the 10Ks and 10Qs for the top US banks. They are all available on Yahoo! Finance.
The reports are usually between 80 and 120 pages long! Given that the reports have been made public, and are signed by the CEO or Controller, they generally look to be SOX compliant.
In each of these reports, you are absolutely right that the Balance Sheet and Income statement information comes from a highly structured General Ledger.
However, that information takes up on 3 or 4 of the 80 to 120 pages. The rest include things like votes for Board Members, lots and lots of text describing what is going on in the bank, regulatory and agency capital requirements, etc.
When it comes to things like tallying up votes for board members, in my experience, most banks use spreadsheets.
Spreadsheets are not inherently SOX violations. It is just that you need lots of checks and balances to make up for a lack of a wiki style audit trail of every change ever made to anything.
Thanks for the clarification, your response makes perfect sense.
Have you ever been exposed to Quickplace?
The reason I ask is that I am attempting to bridge my knowledge and experience in building collaborative applications and the concepts presented by Enterprise 2.0.
It seems to me that a lot of what you are saying about the ‘theory’ of Enterprise 2.0 is very familiar to me in the product Quickplace.
The basic premise behind Quickplace is that people often need to form ad-hoc ‘places’ for collaboration. Often, these places have commonality (project management, litigation management, recruitment process management, etc-see the link provided); and at the same time the needs are often one-off ventures that need specific implementations of forms, documents, processes, etc.
The best part of these ‘places’ is that they are very easy to establish/deploy. There are templates for specific places available, some even meeting regulatory requirements for problems like New Drug Application (NDA) processing and Rapid Response Management (for groups like FEMA, state/local agencies or NGOs.) After selecting a template, you fill in two text fields, enter a name and email address and password and you are set. It even has an offline work capacity and supports the import/management of MS Office suite files. For those who need additional automated processing, there is also a scripting capability called ‘placebots’ that can be scripted to perform various activities on specific events.
I created a trial Quickplace for you to visit if you are interested…you should be receiving an email invitation shortly.
The downside is that it is Lotus Domino based and is fairly ‘old’ technology in web years…It was initially released before 2000 and is on its 7th major version release.
For a variety of reason, I think we got off on the wrong foot. I am not trying to pick a fight by making this statement, I am truly trying to understand Enterprise 2.0 and what it may mean to my organization…any help you prefer to lend in educating me is greatly appreciated.
Rod - I think you’ve written a very cogent post with an exciting vision. Much of it applies to me personally in my work creating knowledge for my little company, but less well when I think of applying it to the organizations we support with online learning and knowledge sharing solutions.
A couple of the challenges we face with this bottom-up approach are that (a) the huge majority of workers we support are “information workers” charged with intelligently implementing guidelines and policies, not “knowledge workers” charged with creating new such; (b) the appetite for the vast majority of those we support is for easy to new tools, not new and creative ones; and (c) we believe that there is a large hidden cost to the multiplication of tools.
We’ve written a few blog posts on this entitled “Ants & Enterprise 2.0″ in our blog - I’d be interested in your comments there. http://blog.collabhost.com
I think that Andrew McAfee is actually referring to emergent structures, structures of tasks or other items that emerge from the collaborative behavior of many people, a team, for example. Here’s where decentralization comes from.
You’re right, when you say that the best Enterprise 2.0 tools should plug into the back end of any corporate entity. They should also support or integrate with some existing processes, like email communication, bridging the existent gaps in present collaboration models.