Mike Brown asked a question on Twitter this morning that piqued my interest. My knee-jerk reaction was to say “bugger all”, but the real answer is much more nuanced. My reply to Mike was this:
Consulting companies make their money by selling themselves as unbiased, unimpeachable, process-driven companies who can work with all parties to resolve questions. In relation to software and web development, those questions might include:
- Which software should we buy, and how much should it cost?
- Which partner firm should we engage in for a tender?
- What do we actually need in the first place, and how much can we afford?
- Can you help us manage the implementation using an approved process framework?
- What went wrong and who is to blame?
In a lot of cases, the way that consultancies go about answering these questions are pretty brainless, and designed to ensure that no one can be blamed for missing the important bits. An uncharitable way to describe this would be ass-covering.
These processes might range from simply undertaking deep business requirements analysis (i.e. asking everyone what they want and then writing it down so that each need/want has a reference number), through to writing RFPs and tender documents, scoring systems for procurement, SWOT analyses, risk matrices, and all the other stuff that anyone involved in “Enterprise IT” will be familiar with.
None of this is particularly scandalous. You could argue that the amount of money spent on consulting firms does not account for the value gained, but that’s a different debate.
What I do find vexing is the fact that software procurement, development, and maintenance is (or should be) an absolute core requirement of almost any modern government department. As such, the trend to outsource all aspects of the process bother me greatly. What we end up with is a world where all of the enterprise IT expertise is centralised in these consulting firms, who in turn can extract near-monopoly rents from the government because each department is forced to use the consultancies to cover their asses.
You see, if you did not use a consultancy in your procurement of a system (e.g. Novopay), when the shit hits the fan, you have no one to point the finger at.
Another issue is that current consultancy culture is inextricably tied to Big IT. The outputs of the consultancy engagements are almost without fail gigantic tomes with massive requirement lists that can only be fully covered by Oracle, SAP, and their ilk. [Ed: Which as Norm points out below, are in turn implemented by the same consultancies who help with the procurement.]
And lastly, the trend toward consultancy by-default means that Enterprise IT skills are sucked directly into consultancy companies, meaning the skills required to manage these projects in-house are unavailable – especially at the salaries paid by government departments. So yes, as of right now, these consultancy firms are “necessary”.
In a perfect world (well, my perfect world at least), the money spent on consultancy should be directed into hiring and training public servants to understand the nuances of large project governance in a way that means the day-to-day users of the software are deeply engaged in the procurement and development of said software. It would also be better spent on breaking down large software implementations (and “Enterprise Transformation Projects“) into bite-sized chunks that can be implemented in isolation.
This in turn would give smaller local IT firms a look-in. It’s a win-win-win: we get public servants who are educated and engaged in IT process; better opportunities for local IT companies; and direct accountability.
Which is why this will never happen.