FROM
https://www.abc.net.au/news/2025-11-24/bom-website-approved-...
"... "The $96.5 million that we're talking about was not just the front end of the website, the tip of the iceberg that the public sees, but the back end, which sees data flowing from tens of thousands of pieces of equipment in the field, to the supercomputer that does all the modelling, right through to systems that actually forecast the weather and put it through to the website," he said.
"So every bit of that chain had to be hardened and made secure to stop a future attack taking down the whole website."...."
I guess some of the endpoints could be supplying critical data . Interference via hackers to these endpoints data could have serious consequences. I have no specific idea actually, but with experience as electronics engineer / embedded programmer .... there probably be some such scenarios .... hence endpoints would need security upgrades .... But I really have NO IDEA about any specifics ...
Hopefully they'll go through all billed invoices with a microscope. My guess is that this will reveal outright fraud from the consulting firm(s), in the form of overbilling in hours.
Even if they bill $500/hr, and they billed 24 hours a day, that would come out to $4.38m / year for each consultant. That's a 11 member team billing 24 hours a day, all year round, for two years straight.
And if they billed more realistic hours, said team would blow up by many multiples. But of course, billed hours is not the only thing consulting firms will charge.
EDIT: For comparison, the website www.yr.no/en, has I believe 10 - 12 devs working. Maybe they've grown since the past years.
I doubt that the consulting firm seriously overbilled.
To my knowledge rather consulting firms are great at selling the necessity of lots of consultants or consultant days:
Just let the customer talk very openly about their wishes for the project, and you immediately get an insane scope explosion for the project, i.e. it "needs" an insane amount of consultants over many years to implement all these wishes.
To increase the bill, every highly qualified consultant that is necessary for the project "needs" a lot of support personnel (senior consultants) so that the senior consultant can 100 % concentrate on their work (otherwise the customer would pay insane hourly rates for highly qualified experts to do "grunt work" - no customer would "want" that). This way, you sell a huge number of senior consultants (this is rather some low rank) to the customer.
And, by the way: since of cause many consultants you sell to the customer shall be highly qualified experts in their discipline, and the project trivially consists of a lot of disciplines, the number of subject-matter experts that can be sold to the customer can be increased by a lot. In some ordinary software project, you would simply use a small team of good generalists (jacks of all trades, master of none) who can do most things in the project, but of cause, as a consulting company, you rather sell the customer "some of the greatest experts that money can buy" (without mentioning that these are insanely expensive and not really needed for the project).
That's how you do it; scamming or billing unrealistic hours is for amateurs.
> To my knowledge rather consulting firms are great at selling the necessity of lots of consultants consultant days:
Just let the customer talk very openly about their wishes for the project, and you immediately get an insane scope explosion for the project, i.e. it "needs" an insane amount of consultants over many years to implement all these wishes.
"Oh yeah, we can do that!" Boom, there's a team...somewhere...working on it. It's a line on an on-site project manager's status report.
You just do it at enterprise scale with all the people needed to make it enterprise legible... and a couple of setbacks and change orders later and you're at 2.5x the original budget!
“It said the cost breakdown included $4.1 million for the redesign, $79.8 million for the website build, and the site's launch and security testing cost $12.6 million“
So 95% of that wasn’t from the cost of the website redesign. “$79.8 million for the website build” included the actual weather modeling and suddenly that number looks way more reasonable.
Classic case of project scope creep stretching what redesign means until you’re including a supercomputer as part of the bill of materials.
AT $250 an hour and 8 hours per day / 2000 hours per year, that's almost ~50 people years, which likely means a team of 10-12 devs working on it over 18 months with another 1-3 design and product and project people in the way making things look good until the bill arrived. Accenture is good at that. [0]
My experience while working in consulting - 1-2 dev, 2 testers, 2 analysts, subject matter expert, scrumbag, product owner, platform owner, designer but she’s always very busy, and some other people that rhyme with management. So about 10 people managing 1 persons work which in principle I’m fine with, except you need to sit just insane amount of meetings and meeting with more than 2 people is horribly exhausting for me.
See, usually you don't have 11 developers coding 24/7. What you usually have is project managers, account managers etc and then a few people who code every now and then. Then you have licenses and support costs.
You can't just code the website, zip the code and mail it to the client. They have many stakeholders like this person needs to be able to show this that persin needs to be able to access this etc because they are running a business or service with than many people. Then you will have requirements like blind people should be able to use that and someone should be able to monitor all that. For each complication you will use specialized tools and do integration, i.e. Adobe will sell you one thing Oracle will sell you another thing and you will have to have people overseeing all these integrations and requirements etc.
That's why you have thousands of employees in tech companies with seemingly a simple product that you can fully code in a week(at least the user facing part of it).
This clashes with my experience; At least in my field....PM can basically only come from an individual who has architecture + development experience and thus is higher paid than any dev who is working on their product.
> You can't just code the website, zip the code and mail it to the client.
The suggestion that the only alternative to paying $96 million AUD ($62 million USD) for a website is getting one that was "coded, zipped, and mailed" is absurd.
> That's why you have thousands of employees in tech companies with seemingly a simple product that you can fully code in a week(at least the user facing part of it).
I've worked at Salesforce, Facebook, and Adobe. I couldn't code even the thinnest sliver of a vertical slice of any of their products in a week.
For most software I've seen, it's a handful of (really complicated) requirements that balloon the cost by orders of magnitude.
Making a piece of software? Easy! Integrating with one external provider? Okay... yeah we can do that. Integrating with an arbitrary and ever-growing number of third parties?
> The suggestion that the only alternative to paying $96 million AUD ($62 million USD) for a website is getting one that was "coded, zipped, and mailed" is absurd.
You can zip and mail every software, given that the mail server(s) accepts the mail, including software that they spent 96$ million on.
Just a hypothetical. If you have a team of 11 devs billing $500 / hour, every hour of the day, all year round, that comes out to a hair over $48 million a year. Do that for two years, and you have the $96.5m bill. Not necessarily rooted in reality.
Ok, here is another realistic hypothetical: a team of 10 devs billing $500/hour, plus extra "package" fees for subject matter expert review, machine learning experts advice, senior partner reviews, focus group experiments, A/B test monitoring, regulatory compliance lawyers, all coming at extra cost. You will find that they can milk that cow legally in much more imaginative ways than your calculation.
Exactly this. At least in the US, consultancies that contract with the gov’t can keep a small full time staff in order to qualify for small-business preference and keep their overhead low, and then depend on an army of subcontractors for large projects.
> "It is unbelievable a private consultancy was paid $78 million to redesign the website," Mr Littleproud said.
This is the crux of the issue. If you have outsourced software engineering competency, yet one of your core missions is maintaining a large pile of software, then this is the inevitable result.
Of course one really ‘unbelievable’ thing is that this infrastructure upgrade contract (including the website) was actually initiated and approved by the previous Government (since voted out to opposition) that Littleproud was part of back in 2017…
My primary competition is guys who are good at marketing, sell expensive packages, and then have someone in the Phillipines or Vietnam do the actual work for a tiny fraction of what is paid.
My primary source of business is customers who paid a lot for they and didn’t get what they asked for and then the vendor blames it all on their subcontractor, or expects more money at astronomical rates. For example $200 an hour for basic WordPress customisation.
It's extraordinary how far departments (even large companies) will go to avoid in-sourcing work. $AU96M is a small team of developers hired, paid and pensioned for decades.
Anyone rubber-stamping that sort of invoice deserves jail time.
This is a huge issue in my experience. I've done some work as a government contractor and often the govt offices have been so heavily outsourced there's nobody left with any technical knowledge at all to oversee the projects. Like the most technical person literally doesn't know what a git repo is.
Even if you hire good contractors who work in good faith, the inability to have remotely technical discussions leads to all kinds of miscommunication and mismatched expectations.
Building and maintaining more in-house developers would be vastly more efficient for the govt, but so many people have a religious level hatred of public employees and glowing respect for private ones. So we not only end up in the current situation, it's actively being made worse in the US by the current administration.
> but so many people have a religious level hatred of public employees
Are you sure this is the issue here? Seems there are lots of regulations around how much public employees can be paid (usually too low for skilled IT), how easy it is to move or fire them, etc. Sounds much easier to just hire a contractor.
Ok great, in-source the project but we’ve been promised a yearly headcount increase of exactly 0. Any hires above that require the minister to sign off.
In your team you have, 3 data scientists that have never worked on a software project, an intern who likes computer games and a PM that used to work in the tax department.
This is a 12 month project, everyone needs to also do their own job and if it’s late, they’re coming for all of us in next years budget.
What would you do? The correct answer is to pay an external consultancy to take the heat and an external team to get the right people needed to get it done.
> promised a yearly headcount increase of exactly 0
That's the silly thing that seems to cause so much trouble. The conversation should be about budget rather than headcount, and the department heads given flexibility on how to manage their budget. There probably is some reasonable amount of budget to bring consultants in for advice on industry practices, but as mentioned elsewhere in the thread, projects like these should generally be managed in-house, and used to build up organisational knowledge, which will be crucial for effective long-term maintenance.
On that note, one of the best uses I've seen of consulting companies, is to have them help define the hiring plan for implementing the project, and sit in on interview panels, to help put in-house leaders in the right mindset of how to assess the competencies that they themselves are lacking.
Opex vs capex budgets is the issue then. The finance people have too much power.
For the in-house experience, keeping a core focus is important for any business. “Are we a website company now?” isn’t something I want my government weather department asking. They’re always going to be worse than industry and one website company per gov department isn’t a slope I’d like to slip down.
Also, sometimes actually defining at the boundaries the problem helps to solve it ie when the internal team creates a spec doc for the ext dev team.
IMHO, the reason government projects end up like this is derisking, pearl clutching, being difficult pricks to deal with and then being uncompromising on process change.
All of the worst aspects you could hope to find in the enterprise at the same time, and then they have a monopoly on being the government.
I’m trying to make the point that it is rational thought that leads to these situations. But the constraints put in place, often never having known of or considered that projects like this need to be done, cause decisions to be made that from the outside just look batshit insane.
I’ve been through this myself as the software developer who’s contract ends on Tuesday but with a company wide hiring freeze causing the general manager to have to call the global dept manager, someone in charge of 25k people, to personally sign off on extending the summer intern into a full timer.
As for not building the website, that’s fine but it will be more expensive tomorrow.
Another POV is that if you had to cost out a finished product ahead of time, that’s what it costs.
Startups and small teams fail before they’ve spent $100m on product development. The scenario you describe never happens.
You could argue instead for incremental development, but if they’ve already decided they need this whole stack of products (it’s $100m for a backend, supercomputer code, etc) why delay and make it more expensive?
I don’t know a lot about weather websites. For video game engines, I know Unity has put more than $3b of product development into the engine. That’s a lot of money. And they’re still not “finished.”
Anyway, the worst part of the discourse is the jail time thing. Pray tell, who the hell is going to feel passionate about programming for the government if their peers are threatening jail time over budgets? One of the reasons we get to enjoy high standards of living in the west is that we pay for consultants, like expensive secondary medical staff, programmers, etc, instead of only tangible things. In your world, the only big government expenditures that are permissible are roads and shit. It is the very same energy as tariffs, this belief that the only valid labor is manufacturing. It’s dumb.
> "A complete rebuild was necessary to ensure the website meets modern security, usability and accessibility requirements for the millions of Australians who reply on it every day," a spokesperson said.
I guess those "usability and accessibility requirements" don't include being usable without JavaScript.
Analysis using BuiltWith shows that the site is coded with PHP, Perl and Java. Smells like real enterprise decisions right there. I’m no expert but I’d be guessing that the Perl is likley the remains coming from the old code base. That’d be fair yeah? Haven’t heard of many people coding websites with Perl in 2025.
I'm of two minds about this comment. A glance at the website suggests it has a lot of content and a full overhaul for 4.1m AUD (2.6m USD) might not be that that high of a price.
But the problem is with the assumption that the website needs a full overhaul. So often a full overhaul is where projects go to balloon in cost by 20x. An outside agency sells the leadership on a big picture full of fluff about "modernization" without any connection to real improvements.
A better approach would be to determine the most important weaknesses of the existing website, and incrementally improve them. But big organizations struggle with this. Government agencies are probably even worse than big corporations, but big corporations are terrible too.
Agreed. I'm a regular user of the BoM website, and from my perspective the old version was absolutely fine. I wasn't one of the people instantly panning the redesign, but after using it for a while I haven't found positives to outweigh the minor annoyance of the change, let alone justify the expenditure. I can totally believe there were some accessibility issues that I was oblivious to, but it's hard to imagine they couldn't have been fixed in a much narrower, cheaper way.
(It was slightly weird that the old website didn't support https -- but on the other hand, I can't really think of a realistic case where that mattered. And I reckon they could have sorted it out for closer to $0m than $100m.)
This is the type of thing that requires everybody who signed off on this and their bosses to be jailed for fraud. Also, the billing company names involved in the billing should be jailed for fraud too. Jail them for five years each by setting them as an example.
I’ve worked on major website redesigns, and there’s so much more to it than just refreshing a few screens. A big redesign is kind of like making a 90min film that costs over $100 million. What you see on the surface is only a small part of the effort. Behind every large website there are teams of experts and consultants. $96 million sounds like a lot, but so is the amount of work required to build or rebuild a site at that scale. It’s easy for people who aren’t familiar with the process to guess what it "should" cost, but without understanding the complexity, those guesses are usually way off. And of course, technology and tech workers are expensive. Sometimes even the initial consulting can run close to a million+ dollars.
Every now and then, we see a post on HN where someone claims they've recreated a well-known website in a short time. But they really haven’t. What they’ve done is copy a handful of screens and maybe a couple of data flows that work for a few test users with no fault tolerance, security, or scalability. Building or updating a hardened, website is far more difficult and far more expensive than anything a small team can spin up quickly. It's expensive and time consuming. Was there some small waste? Yes, I almost guarantee it. But fraud? $96m is a lot of money but without understanding the work that's gone into it, there's no way to know if there was fraud in the cost.
> It said the cost breakdown included $4.1 million for the redesign, $79.8 million for the website build, and the site's launch and security testing cost $12.6 million.
There's a few comments speculating about fraud, but they're way off on the timeframe. I was approached about this project like 7-8 years ago. It's probably been in development the whole time.
1k * 250 * 8 * a team size of 20 is about 40 mil in salary for engineering, could be low, add on their $12M testing, $4.1M just for the design (vintage Deloitte), some cloud cost blowouts and a bunch of dickhead managers and scrumlords, plus the putrid enterprise-grade 3rd party map/data system they've gone with, I bet that wasn't cheap. All up it's in the right ballpark for a typical well-intentioned trainwreck consulting project.
Wouldn't be the first project to blow out because of a bunch of enterprise Typescript, Java and C# devs that can't deliver anything.
I didn't know that about the UK...and my instinct immediately was to think that the UK meteorologists are some awesome badasses walking around with military uniforms and some cool weather patch on their shoulder! :-)
Half-seriously, it does kinda send a signal that such a function for government (meteorology) is so essential, that it stayed lumped in with another important function of government (military/defense). I think its not a bad idea! I admit to not knowing any details at all for how its actually run in the UK...but i contrast that with the severe gutting of budgets of essential agencies in the U.S....and yet again, feel envious of other countries. (Well, maybe not envious of whomever approved the contracts for the AU BOM website, but still envious in other areas.)
when will someone find the $300M bill to update the FAAs NOTAMs website which consists of a giant database with some simple web pages on top?
google pso + cgi federal overloaded the migration with 300+ people and 4 layers of subcontractors to deliver a sub-par infra stack in order to make some media splash before DOGE (rip) could derail it - yes that is why they shifted the db choice last minute to ship quicker versus ship best practices.
NOTAMs is the most important tool used daily by airlines, pilots, airports, the military, and every other industrialized nation in the world. and its now a rube goldberg hodgepodge to avoid an audit from a team that just got canned....
disclosure - i am a pilot who also worked on the migration before being canned for vocally pushing back against the infra architecture decisions that will result in hackable, or offline NOTAMs, halting air traffic worldwide.
I remember that a, few years back the phrase "They have not yet realized that they're a software company now." was used to describe the necessary change in perspective for many companies.
I guess while some companies have caught up, most nation states still have to make this realization.
how much money did corner petrol stations make in profits around the country in one day ? How about the refineries and distributors of Oil and Gas ?
Regarding the entire project -- you cant change what you don't measure? media poking fingers at the expenses of anti-climate pollution projects seems to be a political move these days
Honestly, with an LLM, I can do it by myself for $96,000 - maybe less. I had a brief look at the website.
Downvote me if you want. But I just built a small business website for a relative in about 5 minutes using Vercel's v0. All I did was upload the logo design, gave it some details about the business and it spit out a fantastic professional looking website in about 1 minute. Made some changes to it and pressed a button to publish with a custom domain and it went live. The entire process took 5 minutes.
I'm sure I can make a weather website with a map for $96k.
People rely on this for safety as well as economic activity like agriculture etc.. As bad as the site redesign is, we even more don’t want a crappy vibe-coded site when incorrect weather warnings could kill a lot of people or cause economic damage!
- That "weather website" has to serve all of Australia.
- It's got to be usable on big screen desktops, tablets, smartphones.
- It has to have an uptime of what I estimate to be 99.99%. As the article says, farmers will pitchfork you if you can't tell them when rain will hit their fields.
- It has to be slinging dynamic image data to (about) every visitor.
- The data comes from somewhere. You're lucky if they have that under control already. Probably not.
I came up with these aspects, not knowing anything about what the "Bureau of Meteorology" actually needs in a website. It's just common sense speculation.
This seems like a classic case of the whole "once we have all the data building this is easy". Yeah that's the easy part, that's why building a weather dashboard is one of the classic beginner projects.
The issue is aggregating the data. Same for supermarket systems, transport and logistics, and other systems tied to external data that requires physical interaction.
FROM https://www.abc.net.au/news/2025-11-24/bom-website-approved-... "... "The $96.5 million that we're talking about was not just the front end of the website, the tip of the iceberg that the public sees, but the back end, which sees data flowing from tens of thousands of pieces of equipment in the field, to the supercomputer that does all the modelling, right through to systems that actually forecast the weather and put it through to the website," he said.
"So every bit of that chain had to be hardened and made secure to stop a future attack taking down the whole website."...."
And so clearly was not only for design
I guess some of the endpoints could be supplying critical data . Interference via hackers to these endpoints data could have serious consequences. I have no specific idea actually, but with experience as electronics engineer / embedded programmer .... there probably be some such scenarios .... hence endpoints would need security upgrades .... But I really have NO IDEA about any specifics ...
Hopefully they'll go through all billed invoices with a microscope. My guess is that this will reveal outright fraud from the consulting firm(s), in the form of overbilling in hours.
Even if they bill $500/hr, and they billed 24 hours a day, that would come out to $4.38m / year for each consultant. That's a 11 member team billing 24 hours a day, all year round, for two years straight.
And if they billed more realistic hours, said team would blow up by many multiples. But of course, billed hours is not the only thing consulting firms will charge.
EDIT: For comparison, the website www.yr.no/en, has I believe 10 - 12 devs working. Maybe they've grown since the past years.
I doubt that the consulting firm seriously overbilled.
To my knowledge rather consulting firms are great at selling the necessity of lots of consultants or consultant days:
Just let the customer talk very openly about their wishes for the project, and you immediately get an insane scope explosion for the project, i.e. it "needs" an insane amount of consultants over many years to implement all these wishes.
To increase the bill, every highly qualified consultant that is necessary for the project "needs" a lot of support personnel (senior consultants) so that the senior consultant can 100 % concentrate on their work (otherwise the customer would pay insane hourly rates for highly qualified experts to do "grunt work" - no customer would "want" that). This way, you sell a huge number of senior consultants (this is rather some low rank) to the customer.
And, by the way: since of cause many consultants you sell to the customer shall be highly qualified experts in their discipline, and the project trivially consists of a lot of disciplines, the number of subject-matter experts that can be sold to the customer can be increased by a lot. In some ordinary software project, you would simply use a small team of good generalists (jacks of all trades, master of none) who can do most things in the project, but of cause, as a consulting company, you rather sell the customer "some of the greatest experts that money can buy" (without mentioning that these are insanely expensive and not really needed for the project).
That's how you do it; scamming or billing unrealistic hours is for amateurs.
> To my knowledge rather consulting firms are great at selling the necessity of lots of consultants consultant days: Just let the customer talk very openly about their wishes for the project, and you immediately get an insane scope explosion for the project, i.e. it "needs" an insane amount of consultants over many years to implement all these wishes.
"Oh yeah, we can do that!" Boom, there's a team...somewhere...working on it. It's a line on an on-site project manager's status report.
You just do it at enterprise scale with all the people needed to make it enterprise legible... and a couple of setbacks and change orders later and you're at 2.5x the original budget!
Yeah but in this case it was 23x over budget
That's some Oracle grade consulting right there.
“It said the cost breakdown included $4.1 million for the redesign, $79.8 million for the website build, and the site's launch and security testing cost $12.6 million“
So 95% of that wasn’t from the cost of the website redesign. “$79.8 million for the website build” included the actual weather modeling and suddenly that number looks way more reasonable.
Classic case of project scope creep stretching what redesign means until you’re including a supercomputer as part of the bill of materials.
> * “$79.8 million for the website build” included the actual weather modeling*
I don't think that's a safe assumption to make.
Per https://www.abc.net.au/news/2025-11-24/bom-website-approved-... updated modelling, including the supercomputer to run the model, was included in the bill.
Did they get a new satellite too? /s
AT $250 an hour and 8 hours per day / 2000 hours per year, that's almost ~50 people years, which likely means a team of 10-12 devs working on it over 18 months with another 1-3 design and product and project people in the way making things look good until the bill arrived. Accenture is good at that. [0]
0 - https://australiatimes.com/australia-s-bureau-of-meteorology...
My experience while working in consulting - 1-2 dev, 2 testers, 2 analysts, subject matter expert, scrumbag, product owner, platform owner, designer but she’s always very busy, and some other people that rhyme with management. So about 10 people managing 1 persons work which in principle I’m fine with, except you need to sit just insane amount of meetings and meeting with more than 2 people is horribly exhausting for me.
See, usually you don't have 11 developers coding 24/7. What you usually have is project managers, account managers etc and then a few people who code every now and then. Then you have licenses and support costs.
You can't just code the website, zip the code and mail it to the client. They have many stakeholders like this person needs to be able to show this that persin needs to be able to access this etc because they are running a business or service with than many people. Then you will have requirements like blind people should be able to use that and someone should be able to monitor all that. For each complication you will use specialized tools and do integration, i.e. Adobe will sell you one thing Oracle will sell you another thing and you will have to have people overseeing all these integrations and requirements etc.
That's why you have thousands of employees in tech companies with seemingly a simple product that you can fully code in a week(at least the user facing part of it).
Sure but PMs are billed lower than devs from my experience. You might have 1-3 on this project.
Tangential, but I wonder.
This clashes with my experience; At least in my field....PM can basically only come from an individual who has architecture + development experience and thus is higher paid than any dev who is working on their product.
Maybe I need a different job.
> You can't just code the website, zip the code and mail it to the client.
The suggestion that the only alternative to paying $96 million AUD ($62 million USD) for a website is getting one that was "coded, zipped, and mailed" is absurd.
> That's why you have thousands of employees in tech companies with seemingly a simple product that you can fully code in a week(at least the user facing part of it).
I've worked at Salesforce, Facebook, and Adobe. I couldn't code even the thinnest sliver of a vertical slice of any of their products in a week.
For most software I've seen, it's a handful of (really complicated) requirements that balloon the cost by orders of magnitude.
Making a piece of software? Easy! Integrating with one external provider? Okay... yeah we can do that. Integrating with an arbitrary and ever-growing number of third parties?
Gulp.
> The suggestion that the only alternative to paying $96 million AUD ($62 million USD) for a website is getting one that was "coded, zipped, and mailed" is absurd.
You can zip and mail every software, given that the mail server(s) accepts the mail, including software that they spent 96$ million on.
Why people are so obsessed with yr.no? I know it’s popular in NZ too.
Where did you get 11 members from?
Just a hypothetical. If you have a team of 11 devs billing $500 / hour, every hour of the day, all year round, that comes out to a hair over $48 million a year. Do that for two years, and you have the $96.5m bill. Not necessarily rooted in reality.
Ok, here is another realistic hypothetical: a team of 10 devs billing $500/hour, plus extra "package" fees for subject matter expert review, machine learning experts advice, senior partner reviews, focus group experiments, A/B test monitoring, regulatory compliance lawyers, all coming at extra cost. You will find that they can milk that cow legally in much more imaginative ways than your calculation.
Exactly this. At least in the US, consultancies that contract with the gov’t can keep a small full time staff in order to qualify for small-business preference and keep their overhead low, and then depend on an army of subcontractors for large projects.
Usually there are hard limits around doing 51% of the work yourself, so you can only sub out half of it.
> "It is unbelievable a private consultancy was paid $78 million to redesign the website," Mr Littleproud said.
This is the crux of the issue. If you have outsourced software engineering competency, yet one of your core missions is maintaining a large pile of software, then this is the inevitable result.
Of course one really ‘unbelievable’ thing is that this infrastructure upgrade contract (including the website) was actually initiated and approved by the previous Government (since voted out to opposition) that Littleproud was part of back in 2017…
The private consultancy likely outsourced pieces of the work to (far) lesser-paid subcontractors, too.
I would imagine the margins on that project to be astronomical.
My primary competition is guys who are good at marketing, sell expensive packages, and then have someone in the Phillipines or Vietnam do the actual work for a tiny fraction of what is paid.
My primary source of business is customers who paid a lot for they and didn’t get what they asked for and then the vendor blames it all on their subcontractor, or expects more money at astronomical rates. For example $200 an hour for basic WordPress customisation.
It's extraordinary how far departments (even large companies) will go to avoid in-sourcing work. $AU96M is a small team of developers hired, paid and pensioned for decades.
Anyone rubber-stamping that sort of invoice deserves jail time.
This is a huge issue in my experience. I've done some work as a government contractor and often the govt offices have been so heavily outsourced there's nobody left with any technical knowledge at all to oversee the projects. Like the most technical person literally doesn't know what a git repo is.
Even if you hire good contractors who work in good faith, the inability to have remotely technical discussions leads to all kinds of miscommunication and mismatched expectations.
Building and maintaining more in-house developers would be vastly more efficient for the govt, but so many people have a religious level hatred of public employees and glowing respect for private ones. So we not only end up in the current situation, it's actively being made worse in the US by the current administration.
> but so many people have a religious level hatred of public employees
Are you sure this is the issue here? Seems there are lots of regulations around how much public employees can be paid (usually too low for skilled IT), how easy it is to move or fire them, etc. Sounds much easier to just hire a contractor.
Ok great, in-source the project but we’ve been promised a yearly headcount increase of exactly 0. Any hires above that require the minister to sign off.
In your team you have, 3 data scientists that have never worked on a software project, an intern who likes computer games and a PM that used to work in the tax department.
This is a 12 month project, everyone needs to also do their own job and if it’s late, they’re coming for all of us in next years budget.
What would you do? The correct answer is to pay an external consultancy to take the heat and an external team to get the right people needed to get it done.
> promised a yearly headcount increase of exactly 0
That's the silly thing that seems to cause so much trouble. The conversation should be about budget rather than headcount, and the department heads given flexibility on how to manage their budget. There probably is some reasonable amount of budget to bring consultants in for advice on industry practices, but as mentioned elsewhere in the thread, projects like these should generally be managed in-house, and used to build up organisational knowledge, which will be crucial for effective long-term maintenance.
On that note, one of the best uses I've seen of consulting companies, is to have them help define the hiring plan for implementing the project, and sit in on interview panels, to help put in-house leaders in the right mindset of how to assess the competencies that they themselves are lacking.
Opex vs capex budgets is the issue then. The finance people have too much power.
For the in-house experience, keeping a core focus is important for any business. “Are we a website company now?” isn’t something I want my government weather department asking. They’re always going to be worse than industry and one website company per gov department isn’t a slope I’d like to slip down.
Also, sometimes actually defining at the boundaries the problem helps to solve it ie when the internal team creates a spec doc for the ext dev team.
IMHO, the reason government projects end up like this is derisking, pearl clutching, being difficult pricks to deal with and then being uncompromising on process change.
All of the worst aspects you could hope to find in the enterprise at the same time, and then they have a monopoly on being the government.
> one website company per gov department isn’t a slope I’d like to slip down.
But one software department per government is very reasonable and likely needed anyways.
I'm not defending the ridiculous politicking about government hiring. I agree, it's a blocker to rational thought.
But there is a third option: don't build the bloody website.
I’m trying to make the point that it is rational thought that leads to these situations. But the constraints put in place, often never having known of or considered that projects like this need to be done, cause decisions to be made that from the outside just look batshit insane.
I’ve been through this myself as the software developer who’s contract ends on Tuesday but with a company wide hiring freeze causing the general manager to have to call the global dept manager, someone in charge of 25k people, to personally sign off on extending the summer intern into a full timer.
As for not building the website, that’s fine but it will be more expensive tomorrow.
Non-tech firms are allergic to making software.
I don't know why. They have seemingly no problem keeping their accounting department in house, even though they're not an accounting firm.
If a core part of your business is software, you are a tech company. Congratulations. Yes banks, that means you too.
Another POV is that if you had to cost out a finished product ahead of time, that’s what it costs.
Startups and small teams fail before they’ve spent $100m on product development. The scenario you describe never happens.
You could argue instead for incremental development, but if they’ve already decided they need this whole stack of products (it’s $100m for a backend, supercomputer code, etc) why delay and make it more expensive?
I don’t know a lot about weather websites. For video game engines, I know Unity has put more than $3b of product development into the engine. That’s a lot of money. And they’re still not “finished.”
Anyway, the worst part of the discourse is the jail time thing. Pray tell, who the hell is going to feel passionate about programming for the government if their peers are threatening jail time over budgets? One of the reasons we get to enjoy high standards of living in the west is that we pay for consultants, like expensive secondary medical staff, programmers, etc, instead of only tangible things. In your world, the only big government expenditures that are permissible are roads and shit. It is the very same energy as tariffs, this belief that the only valid labor is manufacturing. It’s dumb.
The new site post re-development: https://www.bom.gov.au
The old site provided with HTTPS: https://reg.bom.gov.au
> "A complete rebuild was necessary to ensure the website meets modern security, usability and accessibility requirements for the millions of Australians who reply on it every day," a spokesperson said.
I guess those "usability and accessibility requirements" don't include being usable without JavaScript.
Analysis using BuiltWith shows that the site is coded with PHP, Perl and Java. Smells like real enterprise decisions right there. I’m no expert but I’d be guessing that the Perl is likley the remains coming from the old code base. That’d be fair yeah? Haven’t heard of many people coding websites with Perl in 2025.
https://builtwith.com/bom.gov.au
It has data-drupal attributes in the source. Not sure where the Perl and Java things would have come from.
I cannot fathom this. What an egregious waste of (assumed) public money.
Surely someone can request to see where this went? Even the original figure of $4.1m is insane.
Why? Its 4.1m AUD. Given the salary of devs and the scope of such a website, the original budget seemed pretty optimistic.
Because the ridiculous scope creep perhaps? And spending $96M of government money on an website (still with large faults that were backed out)
This was Accenture and Deloitte - not some backyard dev shop.
This is pretty standard for Accenture and Deloitte.
I'm of two minds about this comment. A glance at the website suggests it has a lot of content and a full overhaul for 4.1m AUD (2.6m USD) might not be that that high of a price.
But the problem is with the assumption that the website needs a full overhaul. So often a full overhaul is where projects go to balloon in cost by 20x. An outside agency sells the leadership on a big picture full of fluff about "modernization" without any connection to real improvements.
A better approach would be to determine the most important weaknesses of the existing website, and incrementally improve them. But big organizations struggle with this. Government agencies are probably even worse than big corporations, but big corporations are terrible too.
Agreed. I'm a regular user of the BoM website, and from my perspective the old version was absolutely fine. I wasn't one of the people instantly panning the redesign, but after using it for a while I haven't found positives to outweigh the minor annoyance of the change, let alone justify the expenditure. I can totally believe there were some accessibility issues that I was oblivious to, but it's hard to imagine they couldn't have been fixed in a much narrower, cheaper way.
(It was slightly weird that the old website didn't support https -- but on the other hand, I can't really think of a realistic case where that mattered. And I reckon they could have sorted it out for closer to $0m than $100m.)
>It was slightly weird that the old website didn't support https
It always did if you linked to https://reg.bom.gov.au , eg https://reg.bom.gov.au/products/IDN60801/IDN60801.94768.shtm... https://reg.bom.gov.au/products/IDR711.loop.shtml
> Given the salary of devs
Devs are shocked to hear that their 500k salary makes the project cost more than a 50k they think it is worth.
Hahahaha, right. I wish. No developer in consulting is getting paid $500k.
Jesus, people hear one tidbit about prestige staff engineer positions at Microsoft, Meta, and Google, and then assume we're all getting those numbers.
5 person team, 5 overheads (dealing with the government no less), 200k/yr each that’s 2m a year + 50% margin would easily get you a 4m burn rate.
This is the type of thing that requires everybody who signed off on this and their bosses to be jailed for fraud. Also, the billing company names involved in the billing should be jailed for fraud too. Jail them for five years each by setting them as an example.
Absolutely crazy. I'll negotiate on your behalf next time; just give me 10% of what I save you.
Yes, but what special relations do you have with the purchasers?
Should've made it a nice round billion and sprinkle the necessity of AI to usher in the new era of meteorology
I’ve worked on major website redesigns, and there’s so much more to it than just refreshing a few screens. A big redesign is kind of like making a 90min film that costs over $100 million. What you see on the surface is only a small part of the effort. Behind every large website there are teams of experts and consultants. $96 million sounds like a lot, but so is the amount of work required to build or rebuild a site at that scale. It’s easy for people who aren’t familiar with the process to guess what it "should" cost, but without understanding the complexity, those guesses are usually way off. And of course, technology and tech workers are expensive. Sometimes even the initial consulting can run close to a million+ dollars.
Every now and then, we see a post on HN where someone claims they've recreated a well-known website in a short time. But they really haven’t. What they’ve done is copy a handful of screens and maybe a couple of data flows that work for a few test users with no fault tolerance, security, or scalability. Building or updating a hardened, website is far more difficult and far more expensive than anything a small team can spin up quickly. It's expensive and time consuming. Was there some small waste? Yes, I almost guarantee it. But fraud? $96m is a lot of money but without understanding the work that's gone into it, there's no way to know if there was fraud in the cost.
My recollection was ( could be wrong) - It was said in media that "security testing" was 12 million AU dollars
From the linked article:
> It said the cost breakdown included $4.1 million for the redesign, $79.8 million for the website build, and the site's launch and security testing cost $12.6 million.
There's a few comments speculating about fraud, but they're way off on the timeframe. I was approached about this project like 7-8 years ago. It's probably been in development the whole time.
1k * 250 * 8 * a team size of 20 is about 40 mil in salary for engineering, could be low, add on their $12M testing, $4.1M just for the design (vintage Deloitte), some cloud cost blowouts and a bunch of dickhead managers and scrumlords, plus the putrid enterprise-grade 3rd party map/data system they've gone with, I bet that wasn't cheap. All up it's in the right ballpark for a typical well-intentioned trainwreck consulting project.
Wouldn't be the first project to blow out because of a bunch of enterprise Typescript, Java and C# devs that can't deliver anything.
Welcome to the Aussie tech industry.
> complete rebuild was necessary to ensure the website meets modern security... requirements
> launch and security testing cost $12.6 million
What are the challenging security concerns for a weather website? And why would testing alone cost $10+ million?
In the UK, the Met Office is part of the military, for historical reasons.
I didn't know that about the UK...and my instinct immediately was to think that the UK meteorologists are some awesome badasses walking around with military uniforms and some cool weather patch on their shoulder! :-)
Half-seriously, it does kinda send a signal that such a function for government (meteorology) is so essential, that it stayed lumped in with another important function of government (military/defense). I think its not a bad idea! I admit to not knowing any details at all for how its actually run in the UK...but i contrast that with the severe gutting of budgets of essential agencies in the U.S....and yet again, feel envious of other countries. (Well, maybe not envious of whomever approved the contracts for the AU BOM website, but still envious in other areas.)
It looks like it was part of the MoD but it's now part of DSIT:
https://en.wikipedia.org/wiki/Met_Office
There were none. There is a separate sub-site for military consumers. It has its own challenges, but they are not primarily related to security.
when will someone find the $300M bill to update the FAAs NOTAMs website which consists of a giant database with some simple web pages on top?
google pso + cgi federal overloaded the migration with 300+ people and 4 layers of subcontractors to deliver a sub-par infra stack in order to make some media splash before DOGE (rip) could derail it - yes that is why they shifted the db choice last minute to ship quicker versus ship best practices.
NOTAMs is the most important tool used daily by airlines, pilots, airports, the military, and every other industrialized nation in the world. and its now a rube goldberg hodgepodge to avoid an audit from a team that just got canned....
disclosure - i am a pilot who also worked on the migration before being canned for vocally pushing back against the infra architecture decisions that will result in hackable, or offline NOTAMs, halting air traffic worldwide.
I remember that a, few years back the phrase "They have not yet realized that they're a software company now." was used to describe the necessary change in perspective for many companies.
I guess while some companies have caught up, most nation states still have to make this realization.
Is the consulting company actually named somewhere?
If you are a government procurement manager, you certainly would want to know where the biggest kickbacks come from.
No need for a meteorologist, I can tell you $96M bill is sus.
They really messed up cause I would have done it for $95.99M
how much money did corner petrol stations make in profits around the country in one day ? How about the refineries and distributors of Oil and Gas ?
Regarding the entire project -- you cant change what you don't measure? media poking fingers at the expenses of anti-climate pollution projects seems to be a political move these days
Honestly, with an LLM, I can do it by myself for $96,000 - maybe less. I had a brief look at the website.
Downvote me if you want. But I just built a small business website for a relative in about 5 minutes using Vercel's v0. All I did was upload the logo design, gave it some details about the business and it spit out a fantastic professional looking website in about 1 minute. Made some changes to it and pressed a button to publish with a custom domain and it went live. The entire process took 5 minutes.
I'm sure I can make a weather website with a map for $96k.
People rely on this for safety as well as economic activity like agriculture etc.. As bad as the site redesign is, we even more don’t want a crappy vibe-coded site when incorrect weather warnings could kill a lot of people or cause economic damage!
Fixed price? Shake on it LOL
- That "weather website" has to serve all of Australia.
- It's got to be usable on big screen desktops, tablets, smartphones.
- It has to have an uptime of what I estimate to be 99.99%. As the article says, farmers will pitchfork you if you can't tell them when rain will hit their fields.
- It has to be slinging dynamic image data to (about) every visitor.
- The data comes from somewhere. You're lucky if they have that under control already. Probably not.
I came up with these aspects, not knowing anything about what the "Bureau of Meteorology" actually needs in a website. It's just common sense speculation.
This seems like a classic case of the whole "once we have all the data building this is easy". Yeah that's the easy part, that's why building a weather dashboard is one of the classic beginner projects.
The issue is aggregating the data. Same for supermarket systems, transport and logistics, and other systems tied to external data that requires physical interaction.
The Australian government eagerly awaits your expert advice.