The future of IT Performance Analysis and Reporting

As IT/Computing moves through successive "event horizons" to becoming Ubiquitous, Universal and Invisible [like petrol, tyres, home telephones, plastic bags and ATM's], I think we need to monitor, model and report "Performance" in 3 areas:
  • Organisationally: For every dollar spent on ICT, what revenue and profit does it bring?
  • Technically: For each IT system and Computing Application, how does it scale, where are its bottlenecks and 
  • Business Systems and Apps: What are our ICT, staff and resource input costs? Where our bottlenecks? What investments are needed to optimise throughput, reduce costs and increase revenues?

The key points here are:
  • Without the Business, there is no ICT.
  • Without ICT, there is no Business.
  • Neither the Financial or Technical sides can be considered in isolation.
  • 'Business' isn't just about Finances, it's about managing and optimising many other dimensions: Staff, Demand Forecasting, Production Planning, Resource Allocation, Materials Purchasing, Training and Assessment, Marketing and Sales, ...
  • The Business is responsible for integrating the Financial and Technical Sides.
  • If systems don't have data-capture/'sensors' built-in, then post-hoc capture is not reliable and will be expensive.
All Businesses, especially large and Government Agencies, are now completely dependent on their ICT systems, and becoming more so.

Ignoring the importance of managing this "new" class of asset and active cognitive tool (before, people were the only cognitive tools in a business) will only weaken businesses and lead to problems.

If businesses or Government Enterprises can't answer the following question, what are the Management being paid for?:
If we need to cut costs, are we better off reducing staff and increasing ICT, or cutting all budgets similarly?
Not having a functional model of the input to output Transformation matrix means Managers can only guess at what to do. Because they aren't at the coal-face and don't perform the daily income producing tasks, they will have a very poor idea of actual Processes and Procedures.

Without good data and a good-enough model of "how things really get done", they are pretty much guaranteed to make sub-optimal or worst-case trade-offs. It's not like their jobs could be on the line or anything...

No comments: