Source: https://www.abc.net.au/news/2025-11-05/true-cost-of-bureau-of-meteorology-new-website/105948508
Why was the Bureau of Meteorologies 2025 website redesign such a debacle and public relations nightmare? How could almost $100 million be invested into the service bringing in some of the largest consulting firms in the world, only to be met with severe criticism when critical features were no longer available?
Ultimately the BoM’s website redesign was a cock-up of epic proportions brought about by opacity is scope and delivery, lack of single and competent overarching governance of the entire project, and a failure of “bright and shiny” design over actual user requirements.
The story ultimately begins in 2015 when the Bureau of Meteorology site built at that time in Drupal was significantly compromised by Chinese sponsored state hackers. This was no mere cross site scripting hack but the attackers had deeply infiltrated the applications underlying content management system and infrastructure further allowing them access to critical features of the bureaus technology. It’s important to note that The Bureau of Meteorology doesn’t just provide information to the public up large but is a system that is heavily relied upon by not just Australia’s agricultural industry but utilised by the Department of Defense for tactical troop placements and naval maneuvers. Having hacked this important part of Australia’s Defense services was rightly given critical attention by the Australian government at the time it was decided at full overhaul of the bureaus infrastructure and web applications would be required to bring it up to modern levels of security.
Unfortunately this was used as an excuse by the then CEO of the bureau at the time to approach the project with a high degree of opacity, using the terms national security as a means to block the Senate from providing governance to the overall project. Because the project was draped in a high degree of security it meant huge portions of the planning and costing of the project were hidden from standard scrutiny.
The trend of removing technical resources from government departments, to be replaced by global consulting firms, namely Accenture and Deloitte meant that the Bureau lacked the technical expertise required to govern and manage the project, or maintain the project once it was completed. Further, the lack of a single competent overarching governing authority on the project meant that key pieces of the work, being performed by separate siloed companies (Accenture and Deloitte) lead to a lack of sufficient coordination. It appears from initial reports that much of the cost overrun came about due to idle resources within Accenture that were waiting for work to be completed by Deloitte. Unsurprisingly for a government project, Accenture were billing the Bureau for resources on a time and expense basis, but with no provision to place them on hold while waiting for dependent work to be completed. This meant that while Accenture was waiting for work to be completed, they were sidelining their resources, yet continuing to bill for their time. Apparently over the past 6 years there were seven instances of project extension due to dependency delays costing the tax payers millions of dollars in contract costs without receiving anything of value.
Current reports show that the initial Accenture estimates for the project were approximately $31 million dollars, but ballooned out to around $75.3 million following delays and project extensions. The failure was both from a contractual perspective, but more importantly a lack of project coordination, ensuring that the different pieces of the project was scheduled to occur in such a way as not to delay other components. And this ultimately is the consultancy trap.
When the new site was finally launched it was met with a plethora of server criticism from those who had relied on the service for many years. Previously the BoM site had presented high density and complex data in a format that allowed users to ingest vast amounts of data in a simplified and easy to understand format. Users, at a glace could gleen the information they required. The redesign was a departure from the way key users interacted with the site, in a move to dumb down the site and make it mobile first. The problem was listening to technology people who flashed the latest technical buzz words rather than having a keen understanding of the people who relied on the service for their livelihoods.
There was a critical flaw in the decision to redesign from mobile first principals. The nature of the data presented in the old version of the service does not translate well into a small screen. Mobile first is great for e-commerce sites, but is a flawed approach for data heavy services. You wouldn’t expect MYOB to build their accounting software from a mobile first approach, it would lead to a loss of critical information available to the end user.
The redesign of the site, wasn’t just a change to a more modern look, but removed crucial tools used by key stakeholders. One example was the redesign of the Fire Danger Index tool. A tool utilised by volunteer firefighters and rural residents to gauge the daily fire risk in the area. The tool was original built around a precise numerical system from 0 - 100. It conveyed a nuance of important risk information. Replacing part of the scale, 55 (bad) to 95 (catastrophic), with a single generic red badge removed the specific intensity of the threat, which leads to reduction in decision making potential. UI/UX designers were put in charge of the redesign without sufficient communication with the people who use and need the service most.
While there are a number of errors leading to the cost overruns and project failures, ultimately, the government’s decision to strip departments of key technical personnel in favour of an overreliance on big consulting firms is the ultimate failure. It is likely no one within BoM had a sufficient understanding of how to manage and complete a project of this size and scale. I once spoke with a member of parliament who mentioned that government departments have a tendency to bring in the big consulting firms, so when something goes wrong they can push the blame. It appears to have been the case here.
Key Failures Identified:
-
Governance and Opacity: The project suffered from a lack of competent, overarching governance, exacerbated by the then-CEO using “national security” as a means to block Senate scrutiny and ensure high-degree of secrecy around planning and costing.
-
Consultancy Trap and Cost Overrun: The over-reliance on global consulting firms (Accenture and Deloitte) in place of internal technical expertise led to severe coordination failures. A lack of provisions in the time-and-expense contract allowed Accenture to bill for millions of dollars in “idle resources” waiting for dependent work from Deloitte, ballooning the initial $31 million estimate to $75.3 million.
-
Flawed Design Philosophy: The decision to pursue a “mobile first” redesign was inappropriate for a data-heavy service like the BoM, resulting in the removal of crucial tools and the “dumbing down” of complex data.
-
Loss of Functionality: The most cited example is the redesign of the Fire Danger Index tool, where the original, precise numerical scale (0-100) was replaced with a generic red badge for the critical range (55 to 95), removing the necessary nuance for key stakeholders like volunteer firefighters to make informed decisions.
-
Systemic Failure: The ultimate failure is attributed to a broader government trend of stripping departments of key technical personnel, leaving them reliant on consulting firms and lacking the internal capacity to manage a project of this scale.
The lessons from this project are clear and there are a number of things organisation like BoM could and should do to help ensure project success.
1. Strong Internal Governance and Accountability
Appoint a Single, Competent Authority: An internal BoM executive with deep technical and project management experience should have been appointed as the single, non-negotiable Project Director with end-to-end accountability.
Rebuild Internal Expertise: Project funding should have included a significant component for hiring and training permanent technical staff within the BoM. Consultants should be used only for highly specialised, short-term capacity gaps, with a mandate for knowledge transfer to internal teams, ensuring long-term maintainability.
2. Disciplined Contract and Vendor Management
Integrated Team Structure: Instead of separate, siloed contracts with Accenture and Deloitte, a single-prime-contractor model should have been implemented, or a contractual requirement for a truly integrated, co-located delivery team where vendors share risk and accountability for interdependencies.
3. User-Centric and Data-First Design
Prioritise Critical User Journeys: A clear hierarchy of user needs should have been established. For critical services like the Fire Danger Index and core data presentation (relied upon by defence and agriculture), utility and data density must override modern aesthetics.
Continuous User Engagement: UI/UX decisions should not be left to designers in a vacuum. Continuous, mandated engagement with key stakeholders (volunteer firefighters, agricultural industry, military users) throughout the design and testing phases is essential. The process should involve observing how they use the existing system to understand their data ingestion requirements.
Appropriate Design Principles: The “mobile first” principle should have been correctly identified as a poor fit for high-density data services. A responsive design approach that ensures full functionality and data display on desktop while scaling down effectively for mobile is the appropriate technical standard for this type of platform.