Enterprises are rich in both context and control, while new social media sites start completely flat and without either. By embracing and extending our strengths, enterprises can take a shorter journey to successful and mature social media than the consumer models that inspire us.

Context and control

Most enterprise applications are built around two things:

• Context - Who are you? What is your job? What are you trying to do?
• Control - What are you allowed to do? What do we want you to focus on?

Public social applications start with neither, but work hard to build both over time:

• Context - Who are you? Who do you know? What do you like to do?
• Control - We prefer our members to participate and communicate in this manner.

Revolution and Enterprise 2.0

Practicioners attracted to Enterprise 2.0 typically start through a desire for revolution. We want to break out of the context and control forced on us by existing applications and cultural norms. We want new models to emerge, new opportunities to connect with others, new methods of working.

The best examples of these models are available in the consumer space, so our natural tendency is to recreate those tools inside the firewall. This also sits well with our desire for new models and new power. But, it is in direct conflict with the existing norms, powerbase and way of working.

We need to heed lessons on change from Machiavelli:

And let it be noted that there is no more delicate matter to take in hand, nor more dangerous to conduct, nor more doubtful in its success, than to set up as a leader in the introduction of changes. For he who innovates will have for his enemies all those who are well off under the existing order of things, and only the lukewarm supporters in those who might be better off under the new. This lukewarm temper arises partly from the fear of adversaries who have the laws on their side and partly from the incredulity of mankind, who will never admit the merit of anything new, until they have seen it proved by the event.

We need to embrace and extend.

Embrace and extend

Revolution towards a consumer equivalent involves the complete destruction or ignorance of existing context and control the organisation has created. This is a huge leap of faith and not one taken easily by any established organisation.

It's also pointless. Mature social media requires high levels of context (more than enterprises have already) and at least some level of control / agreed behaviour.

A better approach is to embrace the existing context in your organisation. Seed your tools with information that we know to be relevant and expected. For example, when building an internal Twitter, automatically have everyone following their bosses, peers and/or direct reports. When implementing a wiki, setup areas for each existing part of the business. Don't waste people's time and energy requiring the recreation of structures that we already know, use and respect.

Don't give up all aspects of control either. Some groups should have closed membership. Some areas should be locked for editing. That's OK, the world isn't flat.

But, extend both of the models above. Allow new context to develop without intervention. Remove controls that stop the development of context. Expect new controls and conventions to form within this community.

Conclusion

The context and control inside organisations is closer to the norms of mature social media applications than it is to new tools. By embracing the strengths of enterprise structures and extending by allowing users to create new context in emergent areas practicioners can drastically reduce the barriers to approval and adoption.

Successful Enterprise 2.0 uses context to seed flat tools, not to control them. Successful Enterprise 2.0 accepts controls on existing areas, but frees the organisation to create new areas, context and information. Successful Enterprise 2.0 knows that the world isn't flat.

Here are the slides I presented at the Enterprise 2.0 Executive Forum today.

Interested readers, may also like to see my detailed posts on these topics:

• Our Intranet, the Wiki: Case Study of a Wiki changing an Enterprise.
• Jitter: Experimenting with Microblogging in the Enterprise.
• Building Enterprise 2.0 on Culture 1.0.
• Clarify. Simplify. Implement.

Introduction

In February 2008, Janssen-Cilag Australia & New Zealand launched an internal microblogging platform called Jitter. Combined with our intranet's people search capabilities, this formed an interesting enterprise hybrid of Facebook & Twitter style capabilities. This People Search with Jitter solution received Highly Commended in the 2008 Intranet Innovation Awards.

While our intranet wiki JCintra continues to be highly successful, we wanted to keep building our culture of collaboration by capturing and highlighting the flow of ideas. We also wanted to make it easier for our field force to participate and collaborate.

This post is an overview of our approach and outlines some of the lessons learned for others to consider as part of their journey.

People Search with Jitter

This is the home page of the People Search component. Note the simple search box, followed by a list of recent/common searches and then a random face from the organisation.

On the right hand side you can see Jitter posts integrated with the main site news feed. The last 3 posts are shown as a group, and are injected into the news feed based on the latest post timestamp of the news / Jitters.

Searching for a name (e.g. Nathan) shows results from first or last name matches. This quick view allows immediate use of the telephone numbers etc, and incorporates information from our local company system (Juice) and other operating companies through integration with the Outlook Corporate Directory they populate.

Search results are immediate (no Enter click required) and use an AJAX component to prevent the need for full intranet page refresh.

Users may choose to search for a team name (e.g. Information), which returns a picture wall of faces from matching teams.

Note that team and individual results may be mixed together depending on the search term and matches.

Simple page displaying all information for Nathan Wallace. The latest Jitter post is integrated as a status message.

The organisational hierarchy is displayed, including peers, direct reports and his manager. Clicking on those faces navigates the hierarchy. Green arrows show if a team is present under that person.

SMS sending is integrated into the People Search. Messages can be addressed to individuals or entire teams.

If the sender has a mobile, the message appears to have come from their number. If not, there is no reply number, but instead a short text based name is shown on the recipients phone as the sender.

SMS costs are billed to the senders cost centre through the Juice system.

Users can post to Jitter by clicking "Update status" in the Jitter section of the news feed, clicking "update" in the Jitter section of their People Search profile or by sending a text message to the designated mobile phone number.

Posting is done inline, fast to complete and published immediately. Note that SMS following is also available in the system for real time notification of new posts.

An archive of previous Jitter posts is available for browsing.

Adoption and business impact

So far, 59 different people have contributed a total of 306 posts to Jitter. We’re excited that about 17% of people have tried posting, but disappointed that posting remains so infrequent and experimental. Here are some examples:

• State of our public hospitals June 2008 report now available at: http://www.health.gov.au/ahca
• Neuro Specialist Team your cycle meeting accreditation guide can be found on the Topamax specialist team private space. Please review prior to cycle!!
• Suffering from glute meltdown...
• Whoo-hoo! This weekend Trudi became the new Australian Swing Dance Balboa champion.
• XYZ still appears to be down - I have requested ETA on when this may be back and will notify the business when I have an answer.
• Please come and get some friday snacks from my fundraising box in downstairs kitchen. These are to buy new toys and equipment for my baby's daycare.
• Dear colleagues, I need a lift to Gordon or Pymble for the next 3 days. Anyone live that way ?
• Jitter, Jotter, Blotter, Blatter, Matter! Does it?

Jitter has settled into a pattern as our informal news channel. It’s used for public congratulations, for sharing links and for short news flashes. This is a communication need that is infrequent, but not served by email (too intrusive) or JCintra news (too formal).

As a comparison, our SMS message sending tool has seen 104 users send 1852 messages to 5162 recipients. It is commonly used for announcements to the field force and individual messages from office based assistants to travelling executives. Usage has continued to grow each quarter since it was launched.

Lessons Learned

The flow of news on JCintra has been hugely successful and filled a natural need for the organisation. But Jitter wasn’t responding to a need, it tried to create demand. Open collaboration and idea sharing are common organisational goals, but that doesn’t mean there is latent demand among the people of the business for the tools that enable it. With any new organisational capability, always stay focused on end users and helping them to solve a problem.

While Jitter is a highly flexible tool that people are already using for a wide range of purposes, we didn’t do enough to position this new communication medium or to demonstrate the business value. People didn’t know how to use this new tool. Some feedback was negative, but overwhelmingly people asked “What do I post to it?”, “What’s the business value?”. Without clear answers, people just waited to see what others would do.

People have no idea what Twitter is. People have no idea what microblogging is. Most people don’t know what wiki’s, blogs or social networks are either. When explaining Jitter, one user was even worried that this meant that all the SMS text messages they sent to anyone would now be published on the Intranet. These technologies are natural and well known to people like us, but for the vast majority of people in the world they are new, confusing and weird. Remember to design your solutions and train people as though your mum is the key user!

Microblogging is particularly difficult to position as a business tool since it’s so hard to say anything worthwhile in so few characters. For an organisation starting the journey of sharing ideas and thoughts, blogging may be an easier starting point. Posts can be more serious and business like. Blogs are better known, and at worst look more like normal web pages. Authors can craft and position their entries to meet the political challenges and communication realities of the enterprise. Even if your organisation is ready for fast thoughts and short posts, authors can evolve towards really short blog entries.

Conclusion

In a recent post on microblogging in the enterprise Ross Dawson said "It's a learning process. We must discover what a whole array of new communication technologies allow us to do as organizations. We don't know yet. But we do know that they might make a massive difference to how effective we can be. So those who are the first to work it out will be ahead. No doubt about it.".

At Janssen-Cilag, we’re a step or two closer to working it out.

Introduction

A smooth new starter process and good day-to-day management of equipment are essential to the reputation and operation of all IT departments.

Traditionally, this is a never-ending cycle of standards, policies, delivery and backend information collection (CMDB). Frustrated users end up expensing high cost, non-standard equipment and building shadow IT solutions.

Determined to find a better way, we started a project with 3 aims:

1. It should be easier to buy something internally than externally.
2. Equipment orders and new starters are always urgent, so make that the default case.
3. IT needs macro-level control of standards and procurement. Users and managers need micro-level control of equipment and cost-benefit decisions.

With one person working on the process and another working on the system development, within 4 months we launched Juice. This case study presents our journey and results 12 months after launch, recognised by our Gold award in the Business Solutions category of the 2008 Intranet Innovation Awards.

Before Juice

High level standards were in place (e.g. Dell D630 laptop), but low level customisation was allowed (e.g. 4GB RAM instead of standard 2GB) leading to a wide range of parts and supported systems. With standards control and provision approval, ad-hoc decisions were continually required by IT as to what equipment should be supplied to individual users, for example, does this user do enough travel to justify a light-weight laptop? (Of course, in the end users work out to play the OH&S card and get what they want.)

Processes were designed for the best case (knowing weeks in advance about new starters), not the usual case (they start tomorrow). New equipment orders were placed as required, with a lead time of weeks to delivery and deployment. In addition, equipment was owned by and depreciated to individual cost centres throughout the business. This combination fostered a high level of ownership, but lead to protectionism, non-compliance, misallocation of costs and wasted resources as equipment sat in cupboards as departments tried to ensure a smooth entry for new starters.

This ad-hoc ownership and provision meant that broken equipment endured long downtime periods, particularly for remote users, as it was returned, fixed and then sent back. For new starters, equipment like PDA’s may even passed directly from ex-employees to new employees with no review, training or quality check.

Things worked before Juice, because everyone cared about the end user and worked hard to deliver. But, everyone was frustrated, spot fires were common and a lot of excess effort was required to make things run smoothly.

Managing equipment as kits

All equipment is now defined in kit form. For example: Dell D630 Laptop, BlackBerry 8100 Pearl, Lenovo 20" monitor docking station.

Each kit may have a number of accessories. For example: The Dell D630 Laptop includes a mouse, battery, carry bag, power pack and network cable. Whether they are new or previously used, shipped kits always include exactly those accessories. We chose to remove printed manuals, software CD's, etc.

Lost or broken accessories may be individually replaced through single-click ordering. No approval is required as they are cheap and considered essential to the functioning of the kit. If a kit is returned incomplete, we simply charge the user for any missing accessories.

Each kit has a key item (e.g. the laptop itself). If that item is lost, stolen or broken then the entire kit is replaced. For example, we will send a completely new laptop kit including all accessories and ask for the old laptop / accessories to be returned. This ensures that we are always shipping kits complete and tested.

Where possible we use a "swap & repair" model to deal with problematic equipment and minimise downtime. For example, if a user is having trouble with their PDA we simply configure a new one and send it out immediately. The old PDA is then returned, repaired and reused in the future by another user.

This kit-based approach is highly flexible and allows the system to move beyond IT. We also track corporate credit cards, phone numbers and storage lockers as kits in Juice.

The financial model

We decided that departments must control their own costs, while IT provides standards, service and guidance. If IT is a gatekeeper to individual needs and owns costs we'll always be crushed by low level decisions, limiting the business with strict policies and eventually worked around through shadow IT or creative business / safety / any reasoning.

Instead, IT must control the standards, options and service. While users and managers must own the cost and have the freedom to make personal equipment choices.

As such, in Juice all equipment is purchased, owned and depreciated by the IT department. Based on the cost and useful life of each kit, they are leased to users at a fixed daily rate. This has a number of advantages:

• Equipment can be optimally moved between departments and users. (Items owned by individuals are heavily protected and difficult to relocate.)
• Equipment moves from a hidden, uncontrollable cost (depreciation) to a controllable IT cost charged monthly. Managers can now engage in a conversation on how to reduce their IT costs or get better value for money.
• Equipment is only required for active users & a buffer, so less capital is tied up in equipment for vacant positions.
• Users only pay for a kit while they are using it. This encourages the reuse and return of equipment, and encourages IT to maintain the lowest possible inventory that meets service standards.

If a kit is lost, broken or stolen the residual value is charged to the owner's cost centre. While harsh in some cases, this provides encouragement to properly care for and safeguard equipment at all times.

If a user chooses to replace or upgrade a kit, they pay half the residual value of the kit they currently own. (This is a deterrent to upgrading early, but also discourages people from just reporting the kit as lost or broken.)

In the end, IT owns macro-level standards, cost control and service levels. Users own micro-level decisions, needs and cost-benefit choices.

Approvals and workflow

New kits (e.g. I need a data projector), kit upgrades (e.g. I want a lightweight laptop), or kit replacements (e.g. I lost my PDA) require manager approval. This single step workflow is fast, effective and completed by the cost centre owner.

Replacement accessories do not require any approval, they are considered essential to keep the kit in good working order and charged immediately to your team cost centre. For transparency, these charges do appear in the managers monthly bill.

Users and managers have freedom, transparency and accountability for their purchase decisions and costs. IT has no involvement in these approval or workflow stages, we are not a gatekeeper. Our job is to maintain an appropriate menu of equipment, agree recommendations of standard kits per role and to provide high quality equipment, service and training.

Onboarding & Offboarding

Without good people control, you cannot have good asset or equipment control.

Juice synchronises each night with our HR system, detecting any additions, reductions or changes to the organisation structure. Juice also stores a position hierarchy for the business, including predefined employee types with standard equipment. We synchronise the HR changes with the position structure to create appropriate orders in the system.

As a result, we have no new starter forms for IT. This is a huge improvement from forms requesting accounts, ad-hoc requests for security access and equipment orders done on a person-by-person basis. For managers, the work has dramatically reduced. For new starters, the process is far more reliable.

Juice also stores cost centre and team information, inherited down through the organisational tree. With this information, Juice simplifies approvals and allocates charge backs.

Order fulfilment, stock management & asset tracking

A key insight during this project was the importance of separating procurement (purchasing equipment) from fulfilment (distributing equipment). In many cases, Juice is actually solving a fulfilment problem since equipment is reused in the constant turn of employees.

The fulfilment team configures equipment, fulfils orders, checks returned equipment and orders new stock. This is a high touch process with considerable side activities to streamline the business. For example, we consider the redirection of phone numbers and the handing out of phones with a full battery charge to be a standard part of our service. The impact on users of this is remarkable – they can walk away with a phone that is immediately ready for business.

Separating fulfilment from procurement allows us to aim for a 2 business day turnaround on all orders. Kits are ready for immediate shipping and stock is ordered in batches using a simple two-bin Kanban system with visual cues. (Tracking stock levels in Juice is tempting, but much more complicated than visual cues.)

An order is only complete after the recipient marks it as received in Juice. This ensures that the delivery loop is closed and the user is satisfied.

Tracking kits in Juice gives us a full inventory of all equipment, along with its current owner and history of movement. This can be used for asset tracking, identifying troublesome equipment and timely upgrade of old equipment.

Business impact

The first impression on new starters for IT and the business as a whole has been transformed. On their first day we can now hand them a computer, mobile phone, mobile broadband, etc all configured and in working order. This truly shows that we are an efficient, organised company who value your arrival and expect you to be productive. In feedback to HR, basic logistics has gone from the number one frustration of new users to the item they are most likely to raise as unprompted positive feedback.

For managers, the new starter process now requires no forms and works well regardless of the managers proactivity or experience in onboarding. This is a huge relief and reduced burden on their time.

For day-to-day equipment purchases the adoption of Juice has been phenomenal. Within a year we’ve gone from no visibility or control of purchases to more than 180 orders being fulfilled each month. That’s 180 times each month that users can stay focused on their core job rather than worrying about equipment choices, shopping trips or doing nothing and remaining less productive.

Beyond the volume of orders, a great measure of success is the requests we’ve received to expand the catalogue into other areas. Not only did users as for data projectors to be added to the catalogue, but soon after we were asked to add a standard offering for data projector screens. High usability and simple ordering is actually driving users to request standard offerings!

For IT, Juice has increased our accountability through transparency, but it has also made it much easier for us to deliver high quality equipment on time. The process is clear, the orders are standard and our metrics are defined. Beyond that, and perhaps unexpectedly, we’ve also found the transparency has made our users more patient when we have stock delays or high demand. They can see their order is in the system and can trust that it won’t be missed or forgotten.

In short, Juice increases both control and freedom. Users have the freedom to choose which equipment best suits their needs, and control over their individual costs. IT has control of equipment standards (the menu), purchasing decisions and quality delivery.

Integration & next steps

The Juice model of ownership, approvals and cost control is highly extensible and flexible. We’ve continued to evolve the “kits” on offer and extend the system to cover non-IT aspects of our employees needs.

Juice is the database driving the people search capabilities on JCintra.

Juice is used for billing of consumable costs such as the SMS send feature of JCintra and our mobile broadband billing. In time, we will integrate per page costs for printer use and mobile phone charges.

We’re continuing to improve the processes around Juice, streamlining various activities. We automatically order name plates for head office staff and manage storage locker information for shipping of materials to field staff. In time, we will automate the ordering of things like business cards.

We will also continue to explore the capability of Juice to make other usually hidden information visible to end users. We imagine permissions and rights like signing authorities or system access controls tracked as visible assets in Juice - automatically triggering their setup for new users or revocation for leavers.

Conclusion

Juice was a low cost project that has had a large and growing impact on the day-to-day productivity of our business. The Clarify, Simplify, Implement approach pushed our team to convert our delivery process, remove the barriers for users and transform the financial model.

With Juice, users and manages have the freedom to choose the service mix they need, while IT controls the provision, standards and delivery of high quality equipment. Jobs are simpler, roles are clearer and we don’t need to fight for freedom or control.

Clarify. Simplify. Implement.

Working through a wide range of projects, our IT team has settled into a consistent project methodology: Clarify, Simplify, Implement.

Clarify: Work with key stakeholders to understand drivers behind the process. Question motives and key assumptions. Turn over all the rocks to see what lies underneath. (In traditional software terms, this is requirements gathering.)

Simplify: Relentlessly question, review and challenge the processes and solution being developed. Drive for consistency. Search for well-known models or applications you can copy. Don't be afraid to change basic assumptions, where simplicity can be enhanced. Always challenge the value of edge cases and try to eradicate them. Work hard to remove every single process, click, page view, icon, etc until you have something so simple that it feels right to everyone involved. (This is the primary value adding activity for IT.)

Implement: After the requirements are clear, and the solution distilled to its simplest form, start implementing. Do not start with a preconceived solution. Continue to loop through clarify and simplify while performing the implementation. (Use your preferred development methodology, provided it supports constant change and rapid prototyping.)

Tough love adds value

Consultants can gather requirements, and programmers can deliver code from anywhere in the world. But, tough love is only available from those you know and trust. This is the advantage and importance that internal IT teams offer.

External service providers make money through complexity and increased scope. It's in their interest to understand your desires, validate them and then do more work to deliver the wish list. IT needs to reject this model, and help prevent the organisation from becoming as complex as it constantly tries to be.

Simplicity is making hard decisions up front so users can save time and effort in every interaction for all time. Assumptions must be challenged. The status quo should not be accepted as always correct. Trade offs must not be avoided.

Tough love and simplification change IT from a human power tool into a true business partner who provides both leadership and support. Tough love is different to just being tough, it includes love. IT should never be a blocker and will occasionally need to be forgiving. IT should be open in communication and have the best interests of the business at heart.

Design reviews: Brutal refinement and pixel-perfect goodness

An essential part of Clarify, Simplify, Implement are design reviews. These form the ongoing basis for a loop of improvement beyond the initial pass of requirements gathering, simplification and implementation.

A design review includes the application/process owner, the key implementation team and a set of trusted peers. They systematically move through and challenge every process, screen, button, decision, layout and definition. Pixel alignment is important. Removing every excess user decision and superfulous design element matters. Entire pieces of the process or application may need to be redesigned or thrown out. Consistency is critical.

Design reviews are hard and tiring, but ultimately hugely rewarding. Project deadlines and a desire to move onto new problems make it hard to continually refactor your solution design and implementation. It's tempting to stop at good enough, when great is just around the corner. Hours spent discussing alternative user interfaces and nitpicking over definitions can seem like wasted or unproductive time, particularly when your not sure if anyone will notice.

Design reviews take good solutions and make them great.

Zero training

Every user is time poor. They have no interest or time for attending training sessions. Training is the first and biggest hurdle to adoption of your new system and process. While complexity exists and training is required, users can always reject or work around the process with a politically acceptable excuse - "It's too hard".

Our aim, through simplification, is to make people's life easier, reduce the burden on their time and remove all the excuses. The reward is adoption, engagement and relief that that finally it's been done the way everyone always thought (individually) it should be.

Staying simple

After launch, when everyone loves the new system because it just seems to easy, is when discipline becomes truly critical. Feature requests, small changes and extensions will flow from users and every single one "should be easy to add". The hard part is deciding which requests are worthy while ensuring that the system remains simple and consistent.

Clear, simple solutions challenge traditional project economics

With a robust process of clarification and simplication, two things happen:

1. The implementation phase is much easier. (e.g. multi-step, parallel workflow problems become one level approval).
2. The solution becomes very agile and iterative, since it's only through the project process that new clarifications and simplifications become apparent.

Traditional enterprise software projects start with large, feature rich solutions that cover the complexity of features and organisational behaviour that appear to be "requirements". Clarify, Simplify, Implement refuses to engage in projects until the status quo has been challenged leading to changes or understanding.

For example, we recently set ourselves the IT procurement challenge that "it should be easier to buy something internally than it is externally". On our journey to achieving this, the obvious first step was inclusion of a shopping cart (we were using Amazon as a benchmark). But, when we saw it working we realised that using enterprise context (e.g. cost centres tied to individuals using single sign on) we didn't even need the complexity of a cart. One click ordering is now the default process.

Project economics and style change to become:

1. Focused on user experience. All projects and features must provide a significant improvement to the user experience or process. If the cost of implementation outweighs that improvement, then keep looking for a simpler approach that is not so expensive to produce.
2. Just in time development and design. Accept that clarity and simplicity are a journey, no one has the vision to see that far in advance. Be disciplined enough to realise that sometimes small feature additions need large architectural change just to keep the overall application as simple and consistent as possible.
3. Optimised for lifetime value. The cost of an application must include the cost to end users of training, inconvenience and usage. For example, the cost of implementing single sign on must be compared to the cost of X users performing Y logins over Z years.

Small, simple projects are fast to prototype, easy to justify and responsive to business needs. Combining Clarify, Simplify, Implement with an iterative improvement process like the Continuous Application Release Cycle sets a journey of positive dissatisfaction and continuous improvement that will quickly change your organisation for the better.

Take the time to write short letters

Lack of time, politics and ego drive enterprises towards complexity. Complex solutions reflect our perception of the difficulty of our jobs, they reflect the important differences of every department involved and are an inevitable result of looking for quick wins by not challenging ourselves upfront.

As Mark Twain once wrote "I didn't have time to write a short letter, so I wrote a long one instead".

Unfortunately, most project teams take this approach, saving on delivery time and hard conversations and effectively hiding lifetime project costs in lost productivity, frustration and training courses.

Clarify, Simplify, Implement challenges this process and demands the writing of short letters. Users will thank you for it.

Here are the slides I presented at the Enterprise 2.0 Executive Forum this week.

Interested readers, may also like to see my detailed posts on the same topic:

• Our Intranet, the Wiki: Case Study of a Wiki changing an Enterprise.
• Building Enterprise 2.0 on Culture 1.0.
• Clarify. Simplify. Implement.

Introduction

The Continuous Application Release Cycle is a simple process for providing predictable, stable releases in a rapid and sustainable way. It is not a detailed methodology for release planning, development or testing.

Agile methods, open source development and online applications (often in perpetual beta) have established the power of fast iterations and release of minor versions. Release early, release often! I've successfully used the Continuous Application Release Cycle with high velocity applications ranging from publically downloadable software (Sauce Reader) through to internal enterprise web applications.

The Continuous Application Release Cycle

Adopting this cycle brings certainty and momentum to end users while ensuring continuous improvement and low risk release management for developers. Developers move seamlessly from one version to the next, with only a small cross-over for bug fixing during the Beta period.

Release planning is a high level determination of the features and bug fixes that will be incorporated into the release.

Feature development & Bug fixing is the rapid development of wide-spread code changes to implement the planned features.

Design review & Feature freeze is a formal examination of the design and implementation of features for this release. Ensure they are neatly integrated into the application and do not compromise its integrity or simplicity. Build a final design / development plan to finish integration of the features, or delay their release until a future version.

Final development, Help & training and Bug fixing is a finalisation phase, with consolidation and review of the code base. New capabilities may be added to finalise features, but major surgery should be avoided. Help and training materials for new features are constructed in parallel.

Beta launch puts the application through standard test suites and launches to the Beta user group.

Beta testing & Bug fixing has Beta users installing, trying and testing new application features. Appropriate bugs may be fixed with careful control and testing.

Launch puts the application through standard test suites and launches to all users on production.

Production is wide-spread use of this application version. Bugs and feedback are collected for integration into the next version.

The CARC in practice

The length of the cycle varies depending on the type of application, but I've found that monthly releases work well in practice. The development phase lasts about 4 weeks, with 1 week of Beta testing before launch. Each version is in production for about 4 weeks.

The fast development cycle means each release has fewer changes, facilitating a short testing cycle and removing the heavy crunch that typically accompanies software releases. For developers, the key delivery date is the release to Beta testing. With experience, the release to production becomes increasingly routine.

Regular, planned releases keeps developers close to customer needs and allows rapid response to application problems or competitive features. End users enjoy a sense of momentum from the application, and become increasingly engaged as their suggestions, feedback and problems are quickly addressed.

Introduction

JCintra, our Intranet Wiki, has seen incredible levels of adoption and participation, with a positive impact on the way information flows in our organisation.

Over 18 months, JCintra amassed 23,335 content contributions from 239 (~70%) people. The number of contributions per month continues to increase steadily.

But, JCintra continues to function as an incredibly easy to use Intranet, rather than as a genuine Wiki. In fact, 85% of our 3000 pages only have one contributing author. (Interestingly, this behaviour occurs even at Atlassian, who build Wiki software as their business!)

This article documents our cultural journey so far, and outlines our ideas for driving the next phase of change.

Technical and Cultural Maturity

What does success look like?

Decisions about information sharing in organisations like Janssen-Cilag are complex. Some information should be open, but isn’t. Some information needs to be closed and controlled. Some ideas should be discussed in the open, while other ideas need to be carefully communicated.

Success is defined by what we do, not what we have the opportunity to do. Implementing a Wiki isn’t success, building an organisation that will take collective ownership and collaboratively edit content is. Technology creates opportunity for changes of behaviour and helps shift the conversation away from excuses (it’s too hard) to reasons (it’s too risky).

Frankly, at Janssen-Cilag, we don’t yet know exactly how we should be communicating and collaborating. But, we do know that the steps we’ve taken so far have improved communication, increased our flexibility and given people the power to run with ideas. We want to continue this journey, pushing more power to the edge of the organisation.

The Enterprise Collaboration Maturity Model

All knowledge work is either individual or group based, and it is always performed in an individual, shared or open environment.

The Enterprise Collaboration Maturity Model depicts these work models, and incorporates the cultural journey that enterprises take to reach each stage. Currently, Janssen-Cilag provides an open Wiki (high capability maturity) but primarily uses it as Groupware (medium usage maturity).

To continue our journey, Janssen-Cilag needs to become comfortable with the idea that published content is not finalised. Specifically, we need users to:

1. Make contributions in an open space that are not policy or announcements.
2. Edit work or information that is owned collectively.

Successful Enterprise 2.0 style collaboration requires both technical and cultural maturity. While technology opens immediate potential, organisations must grow towards new patterns of usage and collaboration.

The two cultural barriers to collaboration

There are dozens of reasons and millions of excuses as to why people won't share knowledge; but they all fall within two areas:

1. Sharing knowledge adds more work (“I don’t have time to share”); and
2. Sharing knowledge increases personal risk (“I don’t want to share”).

These negatives cannot be eradicated, but they can be minimised.

Reducing additional work

Collaboration and knowledge sharing take time. The technical process takes time, but more significantly, wording your thoughts takes time.

Tools for collaboration must do everything possible to reduce the friction of contributing. It needs to be so easy to use, that you can literally laugh at anyone who tells you it is too hard (in a nice, let me show you, kind of way). In practice this means single sign on, one-click editing and instant gratification on saving. Hurdles like slow technology, login screens, workflow approvals or training kill collaboration before you even start.

The time taken to correctly phrase thoughts and distil ideas is unavoidable, but can be minimised by changing our expectation of shared content away from “finished product” towards “work in progress”. Publishing information early and often (rather than infrequently and completely) moves authorship away from essays and succinct conclusions towards sharing of insights and decisions. The ultimate method for sharing without increasing work is to move the work in progress into an open environment (share everything by default).

Policy opportunities exist to move (but not reduce) the work of sharing knowledge. For example, information is shared verbally on the condition that the recipient will publish it for wider consumption. He who asks, documents. A solution like this rewards the giver with time, builds knowledge on-demand and provides learning reinforcement for the recipient.

Reducing the personal risk of sharing knowledge

Collaboration and knowledge sharing increase personal risk by creating a published, traceable flow of inputs (My mistakes are permanently recorded!) and making past information less valuable than new ideas (What if they don’t need me anymore?).

Risk can be offset by increased rewards, such as recognition for contributions or performance objectives based around knowledge sharing. In practice however, these are hard to implement or judge.

In fact, most people are comfortable with publishing or sharing "finished product". At Janssen-Cilag we've seen this through high usage of news announcements and publication of documents. Unfortunately, most knowledge work is a constant work in progress without a clear end-point and thus never reaches the point of being shared.

The solution is to encourage content contributions that are finished enough to be low-risk publishable, but are not so big as to never reach completion. Encourage people to contribute to a flow of insights and decisions that are made as part of larger projects. Adding to the flow of information (I'm adding to the discussion) is far less risky than publishing final knowledge (I own the final decision) or changing existing content (I'm changing the company position).

Own the flow and the stocks will come

News announcements have been the most successful part of JCintra. Open for publishing by anyone in the organisation, they have replaced email for announcements ranging from major organisational change through to baby announcements.

Through this flow of news, JCintra has become the trusted source for the latest information. "Did you see the announcement on JCintra?", is not an uncommon question around the office. As a result, users also expect JCintra to have the latest policies and information. By owning the flow of news, we've created the trusted source for information stocks.

There are three critical information flows, each of which creates its own stock over time:

1. The flow of news becomes a stock of facts and decisions.
2. The flow of projects becomes a stock of investigations and outcomes.
3. The flow of ideas becomes a stock of potential and experts.

A focus on capturing the flow has many advantages:

• The system always contains the latest information, building trust and adoption.
• The process is easy to enforce and success is readily measured (by monitoring email announcements, the only alternative).
• Work and risk is minimised for contributors.
• Through search, archived flows become a rich and readily available stock.

Over time, the flow of decisions and insights washes over the organisation, helping each person refine their mental map and build a personal body of knowledge. When new items fit their mental model, they can be increasingly confident and aligned in decision making. When news doesn't fit their mental model, they can seek clarity or raise an area of concern.

Focus on owning the flow of information, then have the patience to watch the stocks gradually compile.

Manifesto for Collaboration Tools

Shamelessly stealing from the Agile Manifesto, I propose the following values for building Enterprise 2.0 collaboration systems:

1. Individuals and interactions over processes and tools.
2. Ease of use over comprehensive training.
3. Flexible tools over completeness.
4. Responding to needs over creating demand.

(That is, while there is value in the items on the right, we value the items on the left more.)

We are building processes and tools to help with collaboration, but should never forget that the main thing is that people actually work together and talk to one another. We don't need to capture every conversation or every piece of knowledge, we just want to strengthen weak ties.

Training in systems is important, but only after we've done everything possible to design for zero training. In an enterprise, your Mum really is the end user; design for her! Always sacrifice features and power for ease of use. The minute you have to train people you will lose them to the "more work" excuse.

It's tempting to aim for tools that deliver exactly what people need in different scenarios. To always take tools that one step further to capture their exact requirements. In reality, people like to push and abuse tools that are comfortable, flexible and part of their every day work (e.g. email, Excel). Wiki's, blogs and search are great examples of simple tools that can be used for a myriad of purposes without needing a million customisations or extensions.

Finally, deliver solutions that meet an existing need. If you build it, they won't come. But, you can build it around where they already are.

Next steps for Janssen-Cilag

At Janssen-Cilag, our Wiki has settled into a steady pattern of news publication and simple intranet editing. It is well established and respected for these tasks. Our aim is to build on the strengths of JCintra, while expanding into new areas of knowledge capture.

First we will make internal blogging available to all employees. Links to new posts will be interspersed with news on the home page, creating a flow of ideas in the trusted location but not taking valuable attention away from the full content news items. The people directory will also have direct links to recent posts.

Second, we'll add a Twitter/Facebook style status capability to the people directory which has a history and can be updated via SMS. This is a powerful micro-blogging solution for our field personnel and will be integrated with the Office Communicator Note field. Recent status updates will also be incorporated into the home page news feed, but in a very lightweight way. (The Jitter screenshot shows our early experiment in this area, which we have decided not to launch but instead integrate into the people directory.)

Finally, we hope to expand our internal project management offering with something in the style of Basecamp, which can create a feed of project related milestone news for the home page.

Overall, the aim is to build on the strengths of JCintra by adding ideas and project milestones to the flow of information that washes past people on the Intranet home page. With time this will build a powerful stock but, most importantly, it immediately provides ideas and stimulation to drive interactions between individuals.

Conclusion

Successful Enterprise 2.0 style collaboration requires both technical and cultural maturity. Janssen-Cilag has adopted an open Wiki with the potential for collective ownership, but usage remains dominated by individual contributions to a shared space. This is reflected in the high usage of JCintra’s news column for announcements and the regular publishing of team and policy information.

To encourage an organisational shift along the enterprise collaboration maturity model, Enterprise 2.0 leaders should focus on capturing the flow of information. Over time, the flow builds not only a stock of searchable knowledge but also a reputation as the source of fresh ideas and trusted up-to-date content.

Building on the success of our Intranet Wiki, Janssen-Cilag plans to introduce internal blogging and personal status updates to encourage the flow of individual insights and decisions.

Note: Here are some of the articles I read while writing this post. Thanks also go to my team without whom this would all be theory.

Over the last few years, user experience has become my primary focus when building processes and systems. But, I never truly understood the importance of interface design until I read this post by Andy Rutledge.

Andy intelligently draws a metaphor where words represent content (what you are saying) and body language represents interface design (how you look as you say it). He also beautifully illustrates that a huge component of all communication is non-verbal, although it is less than the sensationalist 93% that people inaccurately extrapolate from Mehrabian's communication study.

What is your interface design saying?