Over half of major information technology projects fail. Corporate America spends tens of billions, yes BILLIONS, of dollars on these projects resulting in huge losses. These results span all industries, crossing geographic boundaries and applying to all corporate operations from front office to back office.
Is this a technology problem? No, technology is just a tool set like any other. It's how the tool set is put to use that defines the boundary between success and failure. Before spending any more scarce budget dollars on software, examine the "worst practices" defined below.
Forget strategy and process
Any major software application selection and rollout will impact both what people do and how they do it. Software is too often viewed as a panacea; just select the right package and all the current problems will go away. In reality, process changes will invariably be needed to take full advantage of the power offered by the new software.
For example, say a company decides to deploy Microsoft Office 2003 within a document production team. The software contains powerful new features for document control and revision tracking. In order to fully leverage the benefits provided, the writers must change the way they manage document production. Without process changes, the new software will provide little or no improvement.
Minimize planning, just do it
Most companies spend large amounts of time and money planning the move to a new facility. They will define needs, create blueprints, hold departmental meetings, publish guidelines, and work on contingency plans. Yet major software deployments are often done with little such planning.
As the number of people using a new software application grows, so does the amount of planning needed. For very large deployments involving hundreds of people across multiple departments, a phased implementation plan is best. Break up the plan into a sequence of smaller, more manageable implementations that can be completed every 15, 30, 60, or 90 days. Keep the plan flexible. Never advance to the next phase until the current phase has reached a stable state.
Worry about integration later
Integration of major software applications across departmental boundaries is a major industry problem. For example, the finance department is using Great Plains while the sales department uses Goldmine. Customer data is stored in both. Customer orders must flow from sales to finance. Making the integration happen can be a major hurdle if not planned from the beginning.
There is no killer application that solves all or even most integration problems. Large-scale implementations always require some level of customization. This frequently leads to problems that put vendors at odds with customers. Tackle these issues early and factor the extra time and cost into the overall plan.
Let the software manage the data
Obsolete, inaccurate or wrong data residing in corporate databases is an insidious hurdle that is often overlooked. Data is the raw material that leads to useful information. Typing errors, spelling mistakes, transposed numbers, and outdated content can act as infections that spread throughout the data center.
Remember the old saying from the days of mainframe computers, "garbage in, garbage out"? It's still true today. Dirty data will severely limit the value derived from any new software application. Some cleanup tasks such as correcting zip codes and area codes can be automated though many others such as the spelling of names call for manual inspection. Evaluate the cleanliness of the data before making a final commitment to the deployment cost and its timeline.
Assume the entire staff will get on board
People naturally resist change. It's human nature. A major software rollout can have a significant impact on the daily lives of the people that use it. Top performers will ask, "Why should I be forced to change my work habits? Those habits have enabled me to excel." On the other hand, marginal performers will fear loss of their jobs. If all employees are not convinced of the benefits of the new software, passive resistance and low employee-adoption rates will result.
Make the end users part of the planning and implementation process and they will react much more positively. Communicate openly about the what's being done and why. Solicit feedback on the plan and listen to the concerns. Above all, provide adequate training and support.
Ignore accountability, everything will be fine
The business workers assume that the IT department is accountable. The IT department assumes that the vendor is accountable. The vendor assumes nothing.
Senior staff is often unwilling to assign accountability to project leaders or program managers. Yet senior staff rarely has the time needed to supervise the project. Thus those in control are not accountable while those accountable are not in control. Create an implementation team with clear senior executive leadership, give them control and hold them accountable.
Maximizing the value of major software investments is difficult under any circumstances. Companies that avoid the pitfalls described above will be much better positioned to fully leverage software technology and improve their bottom lines.
Vin D'Amico is Founder and President of DAMICON, your ADJUNCT CIO. He is an expert in using open source software to increase worker productivity and reduce IT costs.
This article appeared in Vin's monthly Virtual Business column for the IndUS Business Journal in June 2004.
To learn more about how DAMICON can help your business, please take a look at our service programs.
This column appears monthly in the IndUS Business Journal.