I’ve thought a lot lately about the hard problems in IT. There was a time that an ERP upgrade was considered about the hardest thing an organization could undertake and certainly those involve a lot of work and time and cost. ERP upgrades also have a lot of risk associated with them and we’ve heard the stories of blown conversions, delayed conversions, cost overrun conversions and even blown quarters at companies that are blamed on the conversion of an ERP environment.
ERP upgrades and similar large-scale IT projects are certainly hard and require a lot of focus and attention to execute those moves successfully. However, they are mostly a very advanced project management project. By that I mean there is a starting point, there are many big and small milestones in between and there is an ending point. Furthermore, there is a specific set of outcomes from the effort that are either met or not. In short, it is a well-defined goal that has an end point. Yes it might take 18 months and 300 people and $15M in costs to complete, but they are rather well-defined.
I think the hard problems today are not well-defined, they don’t have end points and they may not have clear goals. I wrote earlier about the challenges of managing software licenses and if I expand that into full asset management, vendor management and software license management, then this is an area that is hard to define, hard to keep under control and has changing rules (by vendors). It is easy to lose track of some parts of our hardware stack as changes take place over multiple locations across the years. Throw in some M&A activity and it gets worse. Vendors like to change the way they count licenses over time and they like to change their licensing strategies and policies over time. In short, this is a very hard area to keep in control.
Security and protecting a companies IP is another example of a hard problem. The aim is to protect the enterprise and its IP, and while you may think you are doing it well you can’t be totally sure. You have to keep ahead of organized and disorganized threats, you have to support changing end-point platforms, you have to deal with social threats and just plain carelessness. You almost can’t spend enough in this area. The entire universe of security threats is very, very large.
A third area is collaboration related tools and methods. Today, half the software vendors on the planet are jockeying for leadership in collaboration tool sets and methods. An IT shop can create a complicated Venn diagram of the overlap between their different tool sets. Wrong bets can lead a company in a direction that doesn’t end well and might result in expensive conversions later. And it is hard to decide which direction to go and of course, there are proponents of each solution inside your own company. I think that Collaboration is one of the keys to an organizations success going forward as the what got us here won’t get us there. However, it is hard to figure out the road ahead.
The last hard problem I’d mention is around governance of IT and management of trusted agents in the IT setting. In many recent discussions, I’ve been re-learning how hard it is to solve segregation of duties challenges, authorization/re-authorization/review of trusted agents, and most importantly logging of transactions for trusted agents. These are not easy problems to solve in a large-scale IT shop. These are also problems that are hard to define in some cases and you have to think from the perspective of a threat to the Enterprise, much like in the security area.
These ‘hard problems’ are characterized by not being well-defined, having no end point, having changing conditions or constraints and seem to be very complex. These are the hard problems in IT these days and I think they are harder now than they were 5 years ago. We need our some of our best people working in these areas.
I told someone recently I’d rather gnaw off my arm rather than do another ERP upgrade. I’m rethinking my preferences.