"Variety is the enemy of automation...the more variety, the more likely the operation will rely on manual labor and human decision-making."
"The more a lab business has invested in its automation systems, the less likely the business leaders will be to make wholesale technology changes, regardless of the return on investment. In no other area of the organization are financial decisions more likely to be ruled by sunk cost management."
"IT projects don’t fail, people fail."
from: The Three Laws of Laboratory Automation
The #1 cause of failure in implementing LIMS solutions throughout the world is poor project management. Failure rates among custom and "toolkit" LIMS deployments are often quoted at 70% or higher (consistent with failure rates among large-scale application development projects in general). About one-third of all such projects will be cancelled, and more than half will cost nearly 2X the original budget (Chaos: A Recipe for Success/The Standish Group International).
Ten Keys to Project Success
- Dedicate resources. Once each party agrees to the appropriate level of resources, if you re-deploy them at every crisis, the project will get derailed. In most laboratories, there will always be a crisis, especially if you use computers. Implementing a LIMS is not a part-time job, especially if there is an existing LIMS that still requires attention.
- Choose the right person for the job. Most reasonably automated laboratories have about one computer/IT staff for every 20-30 employees, although this ratio will vary considerably (depending primarily on the amount of customization in software systems). The "ideal" LIMS administrator will have a working knowledge of your lab -- familiarity with the workflows, QA/QC requirements, and instrumentation. Teaching someone off the street to understand what your lab does is usually harder than teaching one of your chemists to understand LIMS. If, on the other hand, you turn to the one person who knows everything about your operation (often a QA or laboratory manager), those staff are probably already wearing numerous other hats. It's not likely you can afford to take someone that critical offline, even part-time, for several weeks or months. More than anything, a good LIMS Administrator thinks logically and systematically, with a stong attention to detail.
- Practice real project management. Get the project started on the right foot. Break the project into manageable phases and measure progress against these key milestones. Use your implementation plan as a working document. Insist on, at a bare minimum, weekly status updates and keep the project on your radar screen. Make sure the administrator has the right experience, right tools, and right mind for the job. Strong PM skills are critical.
- Build in redundancy and contingency plans. Staff turnover or reassignment in the middle of an implementation can set a project back months. It is recommended that at least two people attend system administrator training and work together on the project. The "backup" administrator could be a part-time resource for small and mid-size projects.
- Engage the organization. Divide up smaller portions of the project, like data collection activities, across the subject matter experts (SMEs) in the organization. Share in the burden, share in the success. Tap into as many resources and skills as is practical to manage.
- Engage the vendor/consultant. Treat ChemWare's Project Manager as an extension of your management team. Include us in critical strategic decisions. Don't lock your stakeholders in a room for days or weeks discussing test code naming conventions, label formats, and font sizes on reports; there is a very good chance we have examined the issues from every angle and can offer proven, time-tested solutions.
- Insist on a business case behind every key decision. Enterprise software applications are far-reaching in their impact on the organization, but this can be a double-edged sword if the Project Team isn't judicious about managing scope creep. Even more advanced levels of automation can always be added in future phases. Don't think of the implementation as having a finite endpoint; a properly managed Project Team should and will find opportunities for improvement well beyond the original go-live date.
- Draw a line in the sand for dates, scope, and budget. In the two years leading up to the critical Y2K transition, the average length of a LIMS implementation fell to an all-time low. Every implementation should begin with that same type of very real, very critical, "hard stop," and this should be communicated across every department as a top priority for the organization. If senior management wavers on this commitment, other members of the Project Team may be quick to look for similar excuses.
- Spend pennies to save dollars. As new staff join your organization and the Project Team, arrange for formal training. Time is the real enemy of every IT project. Get advanced training for those who are capable and willing to extend the system and understand how to add value to the organization. The quality of the implementation is almost always a direct reflection of the project team staff involved.
- Anticipate and manage change. With very rare exceptions, a well-designed LIMS will not force you to change your operation. The system may force you to reexamine your business processes, and you may decide to implement new best practices. Expect that some people will not want to change. The time to prepare the organization is before ChemWare staff arrive at the site for the software installation, not right before you move into production. Sharing ChemWare literature and newsletters is a good way to help staff see the benefits of change and reassure them that automation improvements do not pose a threat to their jobs.