| You asked about summary of things I’ve learned about IT. I wrote a bulletin a number of years ago for my consulting clients. I went back to read it nearly 20 years later, and it still rings true for me. So here is what I said (with a few additions in red).
Information Technology is often too complicated and large to be useful, especially for the smaller enterprise. The reality of the Web-world is that you must be fast and pragmatic. The following are a collection of insights -- proverbs if you will -- about the business of IT gleaned from technology and business authors, and thirty years in the industry
- The 50% rule: 50% of what you need to know will be discovered during the system development and implementation. Build a prototype or paper model you can show users early.
- 80% solutions: Build systems that meet 80% of the needs. The remaining 20% can wait, and often prove unnecessary.
- Don't pave the cow path: Automating an old, broken process only creates a broken system. Reengineer the process first. (from Michael Hammer)
- Cut the steps in half: When reengineering a business process, look to cut out at least half of the original steps.
- One approver: A process needs only one approver or sign-off. Everyone else can check the reports.
- Do the minimum for the maximum value (MVP 1): Put something small in users hands quickly that delivers meaningful value for them. (from Extreme Programming).
- Chunk it down: Break a project into smaller phases that still deliver value and can be done in 60 days or less.
- Off the shelf: Don't program a system you can buy. The vendor usually has more resources than you do to develop, maintain and enhance systems. Many customers usually means packaged systems that represent a collection of best practices.
- Custom applications eat more resources the older they get: 80% of IT time is usually spent maintaining old code. Get rid of it. (from Barry Boehme)
- Fail fast, fail cheap: Based on the principle that implies that innovation comes from many, small experiments, most of which fail. Better to fail with small, fast projects. (from Tom Peters)
- System integration over system building: invest in the integration of off-the-shelf systems rather than creating custom systems. MVP 2: Minimum Viable Integration.
- Don't change core code: When implementing an off-the-shelf system, do not touch the core code. The costs of upgrades will kill you.
- Change the process not the code: For off-the-shelf systems, with many customers, the systems usually represents the best practices. Change the business processes to fit the systems rather than the other way around.
- Small is beautiful: Related to chunking down and 80% solutions, small systems, or phases, have less risk, deliver value sooner, and provide the learning necessary to do the next phase better.
These are my top 14 take-aways from a lifetime of managing IT. There are many more things I learned, especially the soft skills. That's one reason this collection of letters is more than one.
|