$96.5m? The BOM website blowout and the cost of outsourcing

$96.5 million? The BOM website blowout shows what happens when we outsource what really matters

BOM

As someone who gets extremely tetchy when business stakeholders demand answers as to why technology costs explode during a project, I am loathed to comment to the Bureau of Meteorology (BOM) website rebuild scandal; but even I am tilting my head at the $96.5 million price tag for the re-launched BOM website.

Look, technology builds are often under estimated and when you are dealing with new technologies, a cost blowout of about 20 per cent ought to be expected. And when you have one or two of the most expensive consulting firms involved, you can expect a cost increase of approximately 10 fold; after all someone must pay for all those boozy lunches and expensive suits.

But $96.5 million?

WTF? What did they do? Gilt the code with platinum? Seriously?

According to my trust website page counter tool, the BOM website consists of 502 pages and more than half of those rendered dynamically. A simple division tells me that each page was built at a cost of $193,00.

The maths is definitely not mathing. 

Explain to me like I am a five-year-old how the User Experience (UX) re-design cost $4.1 million and it had to be rolled back within 10 days because of complaints from users?

User testing anyone?

Even if they made it 100 per cent Web Content Accessibility Guidelines 2.1 compliant – it still should not have cost them more than $1 million in development costs if the front-end developers were properly trained on accessibility tooling.

However, we do know that it was not extensively tested because complaints from the public meant they had to revert back to the old design within 10 days of launch.

The BOM website is not a fluffy eCommerce site – it is a mission critical source of truth that people across the country depend on. Its radar readings are essential for farmers planning their crops, and data from the BOM is used for everything from planning wedding photos to understanding fire risks. This digital asset is not for shits and giggles, peoples’ lives and livelihoods depend on the accuracy and accessibility of this data.

And my first problem with the whole re-design starts with the consulting firms. Why were ‘for profit’ corporate parasites who send profits overseas within 10km of the BOM? Just why? Do we need more proof that consulting firms are con artists? When do we finally put an end to governments across Australia being taken for rides by these criminals in expensive suits?

The BOM is an essential service and must be developed and maintained by public servants; not by grifters motivated by sales numbers. The BOM needs people who believe in its mission to serve Australia, not the boards of multinational companies.

I have worked in and around large technology transformation projects most of my career; and I can confidently attest that government projects run by expensive ‘consultants’ are the worst of the worst. They are hotbeds of passive aggressive poseurs with little world experience of actual technology delivery. 

I know of technologists – good people-  who refuse to work in organisations for at least two years after a consulting firm has been hired to ‘fix’ things. The ‘fixes’ often result in large redundancies, middle management destruction and corporate bloodletting that create toxic cultures akin to Chernobyl 2 days post meltdown. People who remain are often deeply traumatised and disenfranchised and become shadows of their former selves.

I have long argued for the nationalisation of technology delivery in Australia – where we set up a national technology delivery service capability across Australia. This organisation can partner with consulting firms for sure; but their core responsibility needs to be in the protection and development of IT systems that manage sovereign technology.

Without depth of knowledge about the whole BOM clusterfuck, here is how I would have personally run the BOM transformation program of work:

  1. Scaled agile with progressive releases of functionality without the big bang that gave us the big bust
  2. Established at least four interdependent Agile Release Trains 
  3. Establish at least four squads within each release train, which gives us a total of 16 squads in total
  4. Each squad is stacked with people to release functionality – complete with technical lead, front-end designers, UX architects, testers and product owners. Each squad costs around $2 million per year to run. A total annual cost of $32 million is still 30 per cent of the $96 million cost
  5. Start engaging with the people who use the BOM website
  6. Publish the features list and have the people who use the asset vote on priority of the features.

All of the above would have come with the added benefit of the government owning the capability and consultants not walking out with knowledge that belongs to the nation.

×

Stay Smart!

Get Women’s Agenda in your inbox