I touched on this topic a while back, but thought it would be good to refresh it. In short, this is my view of why so many IT initiatives fail—gathered from personal experience plus various consulting engagements.
Overly optimistic project budget estimates
Ok – I am a bit guilty of occasionally fitting a project to a budget (with scope reductions or careful selection of resources), rather than the other way around. But I always added contingency to the project budget to make up for my optimism.
Key Lesson: For large IT projects, ask for three independent estimates of budget and schedule, including one from an outside consultant. Average them and add 10-15% contingency into the budget, and add additional time to complete the project as well.
Best Practice: Call your peers in the industry (or an external consultant) and benchmark the IT project costs and schedule with actual implementations.
Ineffective Project Management
I continue to be amazed by clients who commission a large IT project and then simply trust the vendor (or the senior manager leading the project) to deliver. And before anyone sends me emails saying “What? You don’t trust your people?” I would quickly add that the most effective approach to project management is “Trust but Verify.”. How? It’s easy:
- Request the baselined project plan and post it in your office. Be sure the project plan lists all vendor dependencies.
- Request copies of weekly status, including having the project executive come into your office to post any changes in the baselined project plan.
- Hold monthly executive steering meetings with key leadership, including your internal business clients, vendor(s) and your entire leadership team
- Do not accept any schedule slips or budget increases without serious due diligence and explanations – from your project executive as well as any vendors that are involved.
Key Lesson: Inspect and monitor, trust but verify, and request periodic independent project quality reviews (performed by external consultants or other IT leaders not involved in the project).
Best Practice: On mission critical projects, consider engaging with external consultants to perform independent project or program management throughout the project.
Lack of Vendor commitment and executive engagement
Most IT projects have several vendors who provide critical deliverables and/or support. Too many projects fail because the vendor(s) are not aware of the criticality or do not provide appropriate resources.
Key Lesson: Meet with your vendor account salesperson AND the vendor’s director or VP for service delivery. Review the key project vendor dependencies (refer to your wall chart with the project plan) and ask how the vendor is planning to support the project. Invite both vendor executives to the monthly steering committee (See item 2c above).
Best Practice: Review the vendor(s) statement of work (SOW) to verify that it’s complete and matches project plan and related expectations.
Lack of Business commitment and executive engagement
Almost every IT project has an external or internal business customer – the group(s) that will be using the IT system(s). IT projects typically fail when the business customer says “You did WHAT?“ or “That’s not what I need” during the final acceptance testing. A modified version of the tree swing sequence of activities.
What IT Architected:
What IT and Vendor(s) Developed:
What IT Operations put into Production:
What Business Customer Wanted:
- Start with Process – begin with understanding the affected business processes. Inspect ”As-Is” processes and clearly develop “To-Be” processes. Obtain signoff by the VP representing the business customer organization(s).
- Only when Step A is complete should you develop systems requirements. Obtain signoff by the VP representing the business customer organization(s).
- As soon as Step B is complete, begin developing test cases for IT testing (integration, end-to-end) as well as user acceptance testing (UAT), including stage gates for specific go-live criteria. UAT leads should be designated as full-time roles and come from the affected business organizations.
- Hold at least 3 dry runs for implementation/cutover, with full data conversion. After each dry run, personally review what went well and what areas need improvement.
Best Practice: Use an independent external consultant to write test cases and perform IT testing and assist with UAT. This will provide a valuable second set of eyes on all systems functions and enable you to have an advisor throughout the final development and implementation phases.
Let’s just go live – we can tune performance later
In aviation, they call it ”Get-there-it is.” It is the tendency (after months or years of project planning, development and testing) to proceed with the implementation in spite of warning signs such as failure to meet specific go-live stage gate conditions. At least one of my clients proceeded with a systems implementation – in spite of our warnings that performance was not meeting stage gate requirements. Sure enough, Monday morning call center response time was horrendous and emergency actions to reboot servers and tune systems and databases were necessary. (See http://pm.withflyingcolors.biz/component/content/article/4-landing/17-get-there-itis-are-you-immune.html for interesting reading on this topic)
Key Lesson: Develop and obtain business sign-off on specific criteria for ”go-live,” including number of existing defects (e.g., zero Priority 1, certain number of Priority 2, certain number of Priority 3). Hold executive conference calls leading up to and throughout cutover weekend to monitor status. Review progress verses “go-live” criteria. Hold firm on “no go” decisions, if criteria are not met.
Best Practice: Review stage gate criteria in monthly executive steering committee meetings. Confirm that all participants (including vendors) have the same definitions of priority 1, 2, 3 defects and reporting standards. Not long ago, we found that a vendor had a different definition of priority 1 and 2 defects from that of the client. It took days of negotiations to get everyone on the same page. What was even worse was this vendor reported test case execution status differently from that defined in the project management office standards. Again, it took days to get to the bottom of this discrepancy.
I hope you will join me and will pass on the link to your friends and networks. Please subscribe, send me feedback, and check back next week for the next installment. If nothing else, I promise the international travel tips will be extremely useful!
Today’s International Travel Tip: Getting that upgrade (Continued)
The executive summary: Keep calling your airline’s frequent flyer line.
The more expanded explanation: Most airlines, such as United, allow their most valuable customers to request upgrades weeks in advance. If the upgrade is not confirmed immediately, you are placed on a wait list for the upgrade. Twenty-four hours in advance of the flight, United takes you off the automatic upgrade list, relegating upgrade processing to airport personnel at checkin.
But wait – a little known fact – if you are a Global Services member (United’s highest and most valuable tier) you can call United’s Global Service line – and request they attempt to ”open a seat” to accommodate your upgrade. For me, this has worked more than 50% of the time. Of course, success depends on the specific flight, passenger load and a couple additional factors that are not well publicized.
On long-haul international flights, most airlines (including United) use 3 pilots (a captain and two first officers). Shortly after takeoff, one of the 3 pilots goes to sleep, to be awakened prior to landing. The operational aspects are clearly understandable: after 10 hours at the controls, pilots (even with autopilot) may be fatigued. The rested relief pilot takes command of the plane for landing.
So – where does the relief pilot sleep?
On United’s older configurations of 777 and 767 aircraft, the relief pilot takes one of the first class seats. A series of heavy blue curtains are setup to screen the seat for two reasons:
- Provide maximum darkness and sound isolation conductive to sleep, and
- Prevent passengers from becoming alarmed to see a uniformed pilot asleep during the flight.
On newer configurations of United’s 777 aircraft, the relief pilot sleeps in a compartment adjacent to the cockpit, directly across the corridor from the 1st class lavatory. New configurations of the 767 continue to have the blue curtains surrounding a 1st class seat.
Independent of the relief pilot sleeping arrangements, there seems to be some operational leeway regarding the use of 1st class and business class seats for crew rest areas during the flight.
A 1st class seat is often reserved for a pilot (up until the last moment). Gate personnel ask the pilots if they wish to use the seat. If the answer is no, that seat goes to the first person on the waiting list.
In addition, certain business class seats are also surrounded by blue curtains for the use of flight attendants during a long flight.
In summary, keep checking with your airline to A) verify your position on the upgrade list and B) appeal for a manual action to get your upgrade processed.