The ERP for IT fallacy

There is a school of thought that takes the idea of ERP systems (Enterprise Resource Planning: monolithic integrated systems for running all of a business from operations through to HR to financials) and applies it to IT. The suggestion is that these concepts are fractal: if they work for the organisation as a whole then they can also work on a smaller scale for the IT department and the IT systems and services.

Recently the idea of ERP for IT has had some attention again. We need to stamp it out. Tweet this

These massive federated ERP systems built on common data are only suited to huge organisations, and we all know the many instances where implementing them has all but brought down the company. They only pay off when the processes are well defined and understood, and the procedures are performed very often, and are repeatable and stable.

Making that kind of investment is even harder to justify at a smaller scale in only one department, such as HR or IT. I don't see integrated business systems as a rationale for ERP for IT. Au contraire! it seems to me to be a salient warning.

Automation only pays off - only provides a return on investment over its lifetime - if the system being automated is large scale, has a high frequency of repeated simple transactions, and is stable enough to keep the cost of modifications down. That's not IT. Our IT industry's faith in automation strikes me as naive, almost delusional. Sometimes automation has a positive business case; often it doesn't. Tweet this

Our IT systems are at least as complex as any enterprise. Our systems are volatile: we struggle with higher rates of change than any other sector or industry. The only thing that holds our IT business together is human expertise and flexibility. The idea that integrated automation is going to work for the IT subset of organisations when it is a dodgy proposition for the organisation as a whole is worrying to me.

I wrote of the Five Percent Club for CMDB: the idea that only five percent of organisations are complex enough and rich enough to warrant a CMDB and be able to run one. I think ERP for IT is a concept that is an order of magnitude more complex and expensive than CMDB.

Concepts like CMDB and ERP - in fact all the pro-automation ideas - come from an industrial mindset, of repetitious routine work. So too do the concepts of Lean and Six Sigma and TQM, and large chunks of ITIL, all the process-centric bits. Service industries don't work like that: all services are variable, all services involve humans with all the unpredictability that entails. And now we may be moving from a service economy into a knowledge economy, which introduces even more uniqueness, variation, unpredictability. Trying to use Japanese car-making ideas in IT only works in certain contexts, certain subsets of our world. Tweet this ERP posits total integration: the enterprise as a mechanised automated factory. That doesn't work in many industries - as the world has seen - and it won't work in IT.

The software vendors love the idea of ERP for IT, for its vendor lock-in, its out-of-the-box tech solution to a non-technical problem, its marketing appeal, and its limitlessless up-sell and consulting revenue potential. ERP for IT doesn't have ITIL fueling it like CMDB does, so it is a low level threat right now, but it could run out of control, e.g when vendors are done banging the drum of service catalogue. Chokey the Chimp will keep a close eye on this trend and stomp on it at every opportunity.

My thoughts. What are yours?

N.B. The brilliant Charles Betz has a website ERP4IT. I am generally humbled by Charles' thinking. In this one case I beg to differ :) Organisations the size of Wells Fargo - where Charles did much of his ERP4IT thinking - just might be in the 5% Club for ERP4IT but most of the planet isn't. Visit his site for a rich source of exciting new ideas in IT... and some discussion of ERP.