Today’s tech procurement rule changes is a sensible attempt to get much more value out of the Commonwealth’s $6.5 billion annual technology spend and making it easier to contract with big government. The vast majority of this eye watering figure is on keeping existing systems running, with about $1.5 billion earmarked for new infrastructure.
The strategy — crafted by ex-management consultant, Minister Angus Taylor — comes out of a much awaited PM&C review and picks up many of the practices of the UK Government to drive costs down — the limiting of individual contracts to $100 million being a straight lift from the mother country playbook.
How this works in practice is to be seen. While 96% of fed IT spend is on contracts of less than $1 million, around a quarter of the value goes to five big players — HP, IBM, Telstra, Lockheed Martin Australia and Optus Networks. These vendors service Canberra’s big five — Health, Human Services, Defence, ATO and Border Force — most who have large locked in multi-year contracts. This is big system stuff involving critical data and delivery, and no tier one secretary is going to spend too much time mucking around with nice-to-have edge innovation.
The transformed Digital Transformation Agency is the new investment and assurance police officer on the beat. There are about 17 mega contracts it will be overseeing, in a much tighter process to try to stop the duplication and multiplication of applications and systems. It is notoriously difficult to run a ruler over large ICT projects from a distance, so how that works in practice is to be seen. It is not an original idea, but a whole of government cost envelope for ICT would introduce some much needed internal contestability and force agencies to talk to each other, before they duplicate yet another application.
The push to turn the notoriously clunky procurement system into a window to tap market innovators, sees the government also adopt similar practices to the private sector — the e-procurement catalogue for commodity services, dynamic pricing that means tax payers are not locked into yesterday’s higher prices and a long overdue standardisation of terms and conditions.
Similarly in a world where agencies are being told to start small and iterate, there are measures to enable quick proof of concepts to be stood up. The feds have broadly adopted a NSW government program that encourages agencies to test their ideas before betting the house on bigger, riskier transformations. A gov pitch option will also enable smaller providers to showcase their technology smarts.
Net net this is expected to increase SME involvement from around $1.8 billion or 28%, to 2.45 billion or 38 %. That is very ambitious and compares with a bench mark of say 33 per cent SME involvement the UK government has set for all procurement. The biggest winners will be Canberra’s booming small vendor ICT community. They will be heavily courted by the big techs as they search for partners to meet the higher SME targets. But if the SME’s are simply reselling kit from the big US techs (no Chinese tech in Canberra thanks), then this just becomes a rort.
The dirty little secret is there is a distinct shortage of quality channel partners in the national capital, which is one of the unspoken reasons why some traditional big vendors such as Salesforce, have barely a presence in Canberra. The software as a service model demands local providers who can professionally integrate and customise, and for large scale deployments around core infrastructure such as a CRM, there is a very real paucity of quality and tested providers who can work at the scale government’s require.
There is also a focus on outcomes based procurement, rather than forcing agencies to prescribe up front their requirements. This acknowledges the profound shift in procurement thinking caused by the emergence of cloud based subscription services. The national government has been notoriously slow adopting SAAS services — in part because of major delays in getting cyber security approval from Defence’s zealous Australian Signals Directorate.
We are not talking small fry here — think Google and other titans. Microsoft broke through this log jam recently with commitments to host locally — but only after an protracted period getting applications approved, that had long ago been given a tick by the rigorous US FedRAMP security clearance program. Ironically this is occurring as the new chief of the Australian Cyber Security Centre, Alastair McGibbon, is pushing all agencies to move off their porous legacy infrastructure to modern cloud systems, which typically have modern security capability built in.
There is a push to improve procurement skills and of all the initiatives this is perhaps the most important. The more honest vendors like IBM are up front about how tech is being commoditised and commodified. But the shift to subscription services can often see large blow outs in implementation costs as agencies are socked with multiple licence fees, endless integration charges and high cost maintenance and support plans for software that is sold as plug and play.
This requires real market savvy and for a host of reasons there is naivety and softness in Canberra’s procurement community that big players like the banks, airlines and telcos long ago addressed, by appointing the meanest chief procurement officers they could find. That of course is hyperbole, but in my opinion the number of Canberra CPO’s with top grade commercial experience could be counted not on one hand, but on one finger.
Which brings us to the elephant in the room. Taylor notes in the government’s formal response to the report: “The Taskforce found that a culture of risk aversion in procurement has undermined the freedom to innovate and experiment, to take risks and fail, to try again and ultimately succeed. If we are to reward the innovative and entrepreneurial spirit, a new procurement culture is necessary.”
But as close observers of the federal procurement system know, the real risk is not so much a culture of risk aversion, but rather unthinkingly treading a well-worn path. Cleaning up the supply chain and procurement is very useful hygiene, but the big issue is for what purpose. To date the digitisation push has largely seen a long overdue clean up of front end web sites and some welcome automation of forms and point solutions. But in the main it is paving the cow path stuff, rather than really rethinking the process and approach. Digital is IT with real world purpose and this takes vision and leadership at the business end of agencies.
What has been lacking is clear business outcomes, developed and owned by these agencies. An example is the billion dollar welfare system rebuild, called WIPT. This massive and much needed infrastructure overhaul should be underpinning what we want our welfare system to be. And while there has been some attempt to simplify the business rules that try to make sense of our highly targeted welfare system, the project is largely replicating the current spaghetti of benefits and morass of rules. This drives massive system cost, and essentially replicates the post-war analogue benefits framework.
Cleaning this up requires big political commitment. In a tight fiscal era where we can’t any longer simply pay off the losers from any large scale simplification this means Ministers and departmental leaders have to step up to the plate to drive foundational change. Not much sign of that.
Nor of any strategic vision how the system is to be designed for a period of large scale cognitive applications and the big disruptor of them all, the fast coming 5G-powered Internet of Things. The previous big government tech review was a decade ago and missed the mobile and cloud revolutions. When you are spending over $6 billion a year you want to have a very solid horizon scanning process that brings together the big commercial tech providers and government to plot out an aligned strategic path way that enables government to move from being a perpetual laggard to a leader.
Government is essentially the same worldwide, so done well this opens up huge economic opportunity. Deploying government as a service is absolutely something that could be done as a global offering. Our passport management system for example is a service solution many offshore governments could use, and turns a cost centre into a revenue play.
Finally at a time where every galah in the pet shop is squawking about data and the need to apply it big time into systems, services, policy and regulatory design, there is a desperate need for some deep thinking about what the underlying operating system begins to look like. In one scenario government could rapidly become a pure data play, supporting a vibrant eco system of providers in an Uber/AirBNB type model. As we redo the plumbing for a digital era, we still seem to be relatively clueless about the difference between doing digital, and being digital.
Over to you Gavin Slater.