Guidelines for Knowledge Management

By David Shaw

I read a lot about information and knowledge management. Often it seems airy-fairy. By this, I mean, how do you implement it in the real world? Many people think you just go out and buy a knowledge-management system (KMS).

There! Another check box ticked. And another enterprise system destined for an expensive failure.

So this month I have a few slides for you taken from my workshop on knowledge management. Hope they give you some concrete ideas.

eLearning Return on Investment

By David Shaw

Classroom-based training is still effective for many types of skills that do not transfer well to computer-mediated learning. For other types classroom training has a high cost for instructors, facilities, travel and lodging, and lost time on the job.

Because of the high cost, training is often not done. This has a hidden counter cost for the organization in lost gains in productivity from a lack of skills training, and often an increase in risk and liability when the training is required by legislation or regulation.

As an example, one client was given 60 days by a regulatory authority to deliver mandatory training on violence in the workplace to 2903 employees. To take people off the job and cycle them through classrooms had an estimated cost of $379,100 ($131/employee) including an impact on service delivery. The eLearning cost using an approved off-the-shelf course was $151,200 ($52/employee). One savings was in not having to pay 900 people overtime to backstop those away in class while still maintaining service delivery.

Online learning has come a long way from the early days of computer-based training. Today other benefits of eLearning include peer-to-peer support and collaboration, 24/7 asynchronous availability, onlne resources and web links not feasible in the classroom.

Some of the possible benefits include:

• Improved productivity by increasing skills
• Improved productivity from less time away from the job for training
• Increased efficiency through the  convenience of availability 24-hours a day, 7 days a week
• Increased effectiveness through just-in-time training opportunities
• Centralized knowledge management.
• Built-in trainee enrollment and course management.
• Management of compliance and recertification.
• Increased effectiveness through consistent content, context and interpretation.
• Reduced cost though condensed delivery (e.g., eLearning can take ~50% less time).
• Elimination of travel and living expenses.
• Measurement of employee knowledge retention with pre- and post-assessments.
• Courses can be retaken any time to refresh knowledge.
• Reduced liability.

If you do any significant level of training, eLearning should show a good return on investment (ROI). There are numerous examples of ROI calculators on the web. (Actually most of them are break-even calculators. True ROI requires an NPV or IRR calculation.)

Some of the data you will need is given in this table:

Item	Classroom	eLearning
Number of training cycles per year
Number of trainees per cycle
Trainee Average Hourly Wage
Number of days per training cycle
Classroom cost per day		0
Classroom equipment cost per day		0
Travel and living cost per trainee		0
Material cost per trainee		0
Instructors’ annual salaries		0
Instructors’ travel and living cost per training cycle		0
Cost of back-stopping employees
Number of unique courses per year
Average cost of developing each course
Annual LMS cost	0
Other

Learning Management Systems are a Cloudy Commodity

By David Shaw

Industry statistics show that overall IT departments spend 80% of budget on maintaining core data-centre technology and only 20% on the applications (e.g., an LMS) that deliver the real value for an organization. IT is a cost centre, not a strategic asset.

It’s worse with eLearning. Our experience is that successfully implementing and managing an LMS/LCMS is beyond the capability of most IT organizations. Yes, they can install it and get it running but beyond that they simply don’t have the domain experience. Their experience is in structured data (databases) and they regard courseware as unstructured data. Worse, they believe eLearning is essentially PowerPoints running on a web server; they don’t know how to stream video on the intranet; they won’t support half the file formats common in eLearning courseware; and they’re too concerned about Bring Your Own Device to even think about mobile courses.

Cloud Computing can change this.

Like mobile it seemed that Cloud Computing would always be out there somewhere, always several years in the future. But in the same way consumers pushed mobile computing into the enterprise without IT, they are consuming and promoting cloud computing in many guises and bringing it into rogue business units. Here are just a few examples of Cloud Services that you may have used or heard about:

• Automatic upgrades of software on your computer
• Amazon, FaceBook, Google, Yahoo, YouTube and others reside on vast globally distributed Cloud Computing platform
• Stores like Apple iTunes, Google Play, Windows Store
•  File hosting services like Box, Dropbox, Flickr, Google Drive, iCloud, Instagram, Microsoft (Live) SkyDrive, UbuntuOn
• Free eMail services like Hotmail (Live Mail), Gmail, Zoho Mail, AIM Mail, iCloud Mail, Outlook, Yahoo!, and a dozen others
• Online backup services like IDrive, KinectD, MozyHome, Nero, Norton
• Google Docs, Microsoft Office 365, Zoho office tools
• Google Chromebook cloud computers
• Windows 8 pushing users onto the MS Cloud as the default
• Apple pushing users onto iCloud
• Microsoft’s Lift London studio developing new games exclusively for the cloud
• Wix is a really cool cloud service for developing web sites

What these have in common is that they are a form of Cloud Computing called Software as a Service (SaaS). Other cloud types not discussed here (and potentially of greater interest to the IT department) are Platform as a Service (PaaS) and Infrastructure as a Service (IaaS).

What does SaaS mean for eLearning? There are four basic ways for an organization to operate an LMS or LCMS:

•  Internal Hosted
• Hosted Service
• Shared Service
• Software as a Service

Internal Hosted is a non-cloud single-tenant model.  This is the traditional scenario where an LMS/LCMS is implemented and hosted by an IT department or business unit on local servers. The main advantage is direct control over governance and security. The disadvantages are lengthy and difficult procurement processes, and the need for capital investment that may be hard to get. Capacity is always a step function, so you capitalise servers over the forecast demand or experience loads that cause reduced service from too little capacity.

\

Hosted Service is also a non-cloud single-tenant model. This is the same as Internal Hosted except that it has been outsourced to a service provider who supplies the servers and manages the LMS/LCMS software application. This is sometimes called Managed Service. The servers can be located on premises or externally, depending on the arrangement. The advantage is that you don’t need IT approval for a capital expense because this is usually a leased service. The disadvantage is that you do need IT approval. This model also suffers from over- or under-capacity issues.

Shared Service is a non-cloud several-tenant model. This is an Internal Hosted model in which a few separate business units or organizations share a software application run by a centralised service provider to reduce costs. For example, enterprise web applications like Drupal, Joomla, SharePoint, WebSphere, WordPress and many others can be used to set up more than one web-site domain.

Typically partitioning of user data and content is very weak. Typically the oversight model is weak and bureaucratic. Often the users do not have end-to-end ownership of process. Security and governance, including content approval, are often subject to the business processes and delays of the chief business owner. New functionality, performance levels, service-level agreements and changes to the service are difficult to negotiate. This model also suffers from over- or under-capacity issues. It’s a popular government mantra but only a few practitioners have shown significant cost savings.

Software as a Service is a multi-tenant cloud model. This is an external software application from a service provider that runs on top of a stack of PaaS and IaaS cloud services. SaaS uses hypervisor virtualization to separate tenants. The software and database are architected for multiple tenants with strong partitioning of user and content data.

Security is managed by the SaaS provider. For protection, usually data is fragmented across several geophysical locations and often encrypted. The service-level agreement may be standard and difficult to customize. There is no capital cost – a major benefit. This model matches capacity to demand dynamically, and charges only for usage.

However, under the Electronic Communications Privacy Act of 1986, the Patriot Act of 2001 and the Foreign Intelligence Surveillance Act of 1978/2012 the US government can still access any data without a warrant if it is stored in the USA or if the service provider falls under US jurisdiction. Access does not have to be for reasons of national security – the legislation allows warrantless access for political and economic reasons, too.

Also, as an aside, any email stored in the USA over 180 days is not private or privileged.

In 2013 the FBI’s priority was to get new powers under the 1994 surveillance law called the Communications Assistance for Law Enforcement Act. The FBI wants the power to mandate real-time surveillance of email, cloud services, or online chat providers like Skype. This is everything from Dropbox and online games to Live Mail, Outlook Mail, Gmail and Google Voice.

As a simple matter of due diligence and security from economic espionage data should not be stored with US service providers unless extremely strong encryption is used on the client-side (keys should not be stored in the cloud). This is a thorny issue for service-level agreements.

Some of the benefits of SaaS are:

• Lowering the opportunity cost of running technology. Opportunity cost is the cost of not doing something, thus foregoing future benefits.
• Allowing for a shift of costs from capital expenditure to operating expenditure, which makes it much easier to acquire technology.
• Lowering the total cost of ownership (TCO) of technology (resources are pooled).
• Giving organizations the ability to add business value by renewed focus on core activities by spending 20% on core technology and 80% on better business applications.
•  Enabling IT to focus on the strategic aspects of its role.
• Empowering small- and medium-sized organizations through access to global-scale technology at utility prices.

The National Institute of Standards and Technology (NIST) has defined several essential characteristics for a cloud service:

• On-demand self-service so an end user can sign up and receive services without the long delays that have characterized traditional IT.
• Broad network access via standard platforms (desktop, laptop, mobile, etc.).
• ·Resource-pooling across multiple customers who share bandwidth, load balancing, servers and databases.
• ·Rapid elasticity so the service can scale to cope with demand peaks. This means the service level is always matched to the demand so you don’t pay for over-capacity or lose users from under-capacity.
• Measured service with billing metered and delivered as a utility service the same way we consume electricity and water.

Some so-called cloud services charge by the number of users but we would argue that this is just a form of hosted self-service. A true utility service, e.g., electricity, doesn’t care how many people are in the house.

An LMS is a good candidate for SaaS. At this time there are several hosted self-service LMS providers: Articulate Online, BizLibrary CompanyCollege LMS, Docebo Cloud, ePath Learning ASAP, Epignosis LTD, eFront Enterprise LMS, Zenler Online and probably more.

An LMS meets some of the criteria for a cloud service, such as:

•  It is not a differentiating or core business application
•  It is usually web-facing
• Users need access from desktop, laptop, mobile, etc.
• Demand can be spiky when training campaigns are underway
• Global delivery through content-distribution networks
• It can be managed by business departments

An LCMS is not a good candidate for SaaS. It meets some criteria for not using a cloud service, such as:

•  It is not web-facing (it’s a back-end system to a web-facing LMS)
•  It does not have to scale rapidly to meet demand
• It is a value-added service that differentiates the organization (especially with content re-use)
•  It involves uploading many and often large files that consume bandwidth

Start thinking LMS cloud in long-term strategies (with appropriate security). Think about bringing an LCMS in-house.

Rapid eLearning Prototyping

By David Shaw

 A PDF of this article is available by email from the author.

Rapid eLearning prototyping uses modular design patterns (templates and components) for courses of different but standard types. The goal is to reduce development cycles by 50%. Some examples of components are re-usable HTML objects such as Navigation Object, Multiple Choice Object, Feedback Object and so forth. The business objective in rapid prototyping is:

•  Reduce cost
• Reduce development time
• Better meet client requirements
• Reduce time for future updates
• Create better and more consistent courses
• Allow developers to spend 80% of effort in creative tasks

Like many activities that require some project management, eLearning has traditionally used the Waterfall method. It’s a sequence of mostly linear activities sometimes called, “throw it over the wall” because it does not encourage collaboration. It’s baked in Microsoft Project and most project methodologies.

Figure 1 shows a typical waterfall process for developing eLearning courses. It assumes that all requirements are understood at the outset. Even if this were true (it never is) the development cycle is lengthy so there are many opportunities for requirements to change before the project is finished. Updates for new versions are also time consuming as they have to go over the steps in the waterfall. With the waterfall team members often spend the bulk of their time managing the process and creating the many reports and other documents it requires.

This too has to change.

Fig 1 – Traditional waterfall process

Developing eLearning is a lot like developing a software application, and one thing the industry is lacking is an integrated development environment (IDE) to increase productivity through rapid prototyping or other form of incremental development.

Studio applications like Camtasia or Captivate are a step in the right direction toward an IDE. With Camtasia or Captivate you can create a course and embed graphics and video. If you open media-editing software and revise the graphics or video it will update automatically in the course stream in, say, Captivate. At the end of the process you can publish a course with all of its components and then load it into an LMS for final testing.

Although suites like Adobe’s eLearning Suite or the Lectora set of tools have good integration between tools, an IDE presents a single development environment with a higher degree of integration between tools and the testing environment. When scanning the market for the current set of tools they are still centred on converting PowerPoints and creating Flash.

An ideal IDE would drag and drop course modules in a graphical learning-path or other course-breakdown diagram with all its branches. Clicking on any part would open an editing environment for easy updates of all types of content. Templates with basic elements for re-use, ready-made interface and navigational elements, widgets, media and other objects would be drag and drop.

It’s not an exact analogue but, for example, using an IDE the author developed a smartphone app for both iPhone and Android in less than a week.

However, even without good IDEs we can still make advances through rapid prototyping with small cross-discipline teams. Figure 2 shows one variant on rapid prototyping that the authors have used in eLearning. The goal is to continuously improve the time to completion with successive projects by developing and improving design patterns that can be re-used. A key principle is the separation of roles between Subject Matter Expert and Instructional Designers.

Fig 2 – A rapid-prototyping process

In this rapid approach there are several iterations in each phase and the phases are time boxed to control the schedule. Instructional designers work one-iteration ahead of developers and testers to keep their work-product backlog full. The make-up of the instructional team in each phase is entirely dependent on the context of each project. Sometimes more people are better in a discussion, sometimes less. Either way, daily communication and review within the instructional team is essential. The frequent participation of the Client and Subject Matter Expert (SME) is also essential.

The goal with the time box is to ship a course that meets core requirements. If the course does not meet all enhancement requests, then a follow-on project should be funded to develop the next version of the course. This implies the need to develop and manage a roadmap for the development of the course.

Short iterations may add too little functionality, leading to significant delays in final iterations. Because development is rapid documentation tends to be less than found in a waterfall process. A significant amount of post-project documentation may be required. It is a good practice to keep functional requirements updated and records of screen shots as the project proceeds. Every enhancement, change request and decision should be documented in an email and filed as a PDF.

It may also be difficult to communicate progress and status to management. Including the Client in this process will help alleviate concerns. Because each phase is time-boxed, progress reports can be based on semi-weekly or weekly estimates of effort required to complete versus the budgeted effort to complete.

The following table shows how to calculate and present the variance and percentage complete. A column for cost could be added but note that percentage spent is not equivalent to percentage complete. This method of reporting should also be used for individual tasks. The variance can also be tracked over multiple projects, as a performance indicator in the development group.

Status Report
Work to Date		Days
Actual work in days to current date	A:
Estimated days to complete phase	B:
Total estimated days to complete phase	C= A + B
Budgeted days for this phase	D:
Variance (positive is bad!)	= C - D
Percentage complete	A/C