management

Yeah, I know those guys, the ones making powerpoints and excel files all day, coming crying to you because they want their stuff done tomorrow, otherwise they won't hit THEIR targets. :P

Ah, yes, Accenture. Technically middling but they're great at the softer stuff – sales, politics and convincing the unaware that they've got the best people for the work.

...even if the backend is a 30 years old AS400 behemoth.

Never really thought of an AS/400 as a "behemoth". They're far smaller than the systems they were designed to stand in for.

Let's pay these 20 people 200k a year (which is a huge salary) to do it.

That really depends where the people you're paying that money are located. Sure, in Cleveland or Arkansas, $200K/yr is huge. In NYC, SF bay area, Chicago and even the DC region (particularly the western half of the region), that hugeness is greatly reduced (each of the named regions have mean household incomes in well in excess of $100K/yr and having stratospheric housing costs).

Let's put 2 millions of related costs,

As noted previousy, Accenture is great at the softer stuff. They love to get face-time with customers. Lots and lots and lots of meetings. What's more, they love those meetings to be face-to-face – even (especially?) if being able to do face-to-face means lots of airfare (and we're not talking tickets on Spirit or even even Southwest; nor are we talking "base" fares) or rides on Acela (and associated hotel and meal expenses). All shit that adds up very quickly. All stuff that, with a team of meeting-goers would easily eclipse $2Mn in the course of 12 months.

Simply put, when customers are willing to pay stratospheric rates, companies are more than willing to structure contracts to milk every last ounce of that willingness. Rarely does it meaningfully trickle down to the people that do the grunt-work. Even more fun is, because there's a ton of money to be made on the lifecycle of a solution, solutions are often sub-optimally implemented so that there's a big, fat recurring-revenue stream for keeping a fragile solution going.

Wow, how did Accenture managed to make a business?
I guess by being crooks and with a little bit of luck.

What about the amount of time and resources invested from across the organization to secure the contract to include salesmen, technical advisers, and legal. I'm not necessarily defending Accenture but there's more to the process than just development.

I'll skip over your calculations and so on (IMHO, they are hugely optimistic, paying those salaries means that company wants to manage a disposable dev department internally - otherwise you'd wrap those devs into a company, where you'd need to add a few people on top of that for management, recruitment, accounting, HR, etc, lots of extra expenses and that company will add a markup for profit on top whereas internalising the process adds a bit less monetary costs but probably more time).

Costs aside, there's one thing I learned working in outsourcing: bottom dollar devs might be technically excellent but they rarely have the grit to stand up to managerial pressure to give bottom dollar estimates to ensnare customers. Plus said management tend to ignore devs anyway in the initial phases of a bid.

I work for a smaller outsourcing outfit and one thing I keep hearing is that we need to find ways to lower estimates, do it simpler and so on. Years in the field and I still don't have a way to convincingly explain the hidden later costs of pushing for bottom dollar (in fact maximise profit) from the beginning.

Customers throw money at such companies because the middlemen know how to make deals sweet and they get the impression that if you have a contract, then things will go as written.

As a dev ... you'd think the easiest way is to not work in such environments, but you do get more than decent salaries and everything is sweet until something like this happens and you get the ire plus the stress of explaining that you tried to explain crap that might go down to managers that don't even speak your language.

I hope Hertz will take it all the way, for whatever good that will do. Accenture has > 10 billion in revenue, 32M is a drop in that bucket. They'll keep doing what they do anyway + likely those 10 billion can buy enough PR and legal support to ensure this case doesn't make a dent in their activity.

I think it's an ethical issue for an awful lot of consultancy companies. Sales and commercial teams get contracts, forcing bad time estimation in order to avoid the contract is awarded by another company. So, the pressure is too high against the developement team, provoking it a burn-down crisis, then, a lot of developers resign their positions. Consequently, this fact brings more time delays with respect to the release date. In addition to this fact, the code will be badly written, buggy code pieces everywhere, bad practices and only black-box testing at most.

I worked for Accenture, but resigned within a year. Way too much politics within the organisation. I would say Accenture is probably okay if you need generic consultants with soft skills. But if you are looking for technology delivery, please consider others.

I've generally heard that Accenture trains someone for a month, then bills then out for hundreds an hour. I agree that it's an ethical issue. If you foresee that you cannot deliver, it's your responsibility to raise that flash asap. That's why people get paid for experience, it gives you the ability to see further down the road than others. Maybe the specs were bad, so partly Hertzs fault, but it's Accentures job to call that out. You claim to be the best, you must deliver the best