How on earth could it get that bad? Even our own highly 'experienced' Capita could manage that shopping list of fail!
Car rental giant Hertz is suing over a website redesign from hell. The US corporation hired monster management consultancy firm Accenture in August 2016 to completely revamp its online presence. The new site was due to go live in December 2017. But a failure to get on top of things led to a delay to January 2018, and then a …
Having been at a very large company that ended our contract in favor of Ass-enter and thus I had to interview with them in hopes of staying with the company, I'm not the least bit surprised.
Ass-enter hires the cheapest IT people who have impressive resumes and shit work ethics, they can find, bogs them down with meetings and metrics, all in an atmosphere of modern tech hipster-ism.
They are IBM's counterpart.
How did Hertz let this happen? Why the bloody hell did Hertz wait so long before pulling the plug on Accenture?
It's common freakin' sense that Hertz should have structured the deal so that there were regular reviews of Accenture's work with milestones set. I suspect that someone high up at Hertz knew someone high up at Accenture and these two golfing buddies put together a love/love contract that was way too friendly to Accenture. So its only after Accenture's incompetence reached a point where the fatcat at Hertz was internally threatened & embarrassed it finally reached point where action had to be taken.
I regret I can only use this icon once - wish I could repeat it ten times for this massive FUBAR.
But, but, but they were practicing agile development!
You know so that they could keep tabs on it to make sure everything was working...
SO Sorry, I couldn't even make light of this with a straight face.
You want something done right, hire the best which means hiring teams of old farts who charge more but do the work right the first time and will not over charge you like Accenture does with blended rates.
There's more I'd love to say but even posting with an Alias doesn't mean I'm anon. Many people know who I am. ;-)
My own guess- from having come across things of this sort (much smaller scale)- is that the decisions and what passed for planning were all done at a level far away from day-to-day service or IT provision by a management that just looked at predicted cost figures and failed to think about what they needed to put in place alongside, or do a sensible risk assessment of the project.
Dunno about the work ethics, but impressive resumes + cheap is the right description.
Thing is, you can bring in all the talent you want, but if you spend their introduction to the company by hammering them into the outsourcing mindset of working fast, cheap and throw away practices out the window, then this is what you get. This is only the higher profile trouble Accenture got itself into.
And they work much like any other outsourcing company. Get some architects develop a plan, hand it over for execution to cheap devs who fail to stand up under pressure and the plan goes out the window. Then throw money and people (cheap, again) at the problem hoping it gets solved. Sometimes it does, often it doesn't.
The sad part is that those are structural problems that smaller companies manage to deal with (to some extent) but bigger outsourcing companies manage to mix in all sorts of failures with entrenched managers and cheap devs that fail to stand up to them,
I bet the ones who had the bright idea are connected, across a few degrees of separation, to the to-be-hired firm. "I scratch your back, you scratch mine."
Accenture, IBM, CGI -- just to name a few -- they are of the same species of predator preying on tech-illiterate clients. IBM with the Phoenix pay system for the Canadian government. CGI with HealthCare.gov. Very lucrative business.
Working on a couple of IT projects with Accenture on behalf of a 3rd party client I can understand Hertz's pain.... Most of their staff is scraping the barrel-quality, although some shine through and mainly thanks to their own excellence. Delays, lousy specs, even not being able to get simple stuff is a day to day occurrence.
Finally... a company that pushes back and files suit for breach of contract / failure to deliver.
Many of us, I assume, have been termed "nay-sayers", "roadblocks", "not team players", et cetera when we found the courage (or gall, as the brass tends to call it) to compare a consulting firm to the emperor in Anderson's prophecy (er, fairy tale).
Glad to see that sometimes even the brass can find that courage. Or did Hertz have a changing of the guard and the currently top doesn't run the risk of being criticized in the selection of said consulting firm, or their failure to hold them to the contract from the start?
How on earth could it get that bad?
All too easy.
Firstly you get your regular progress meetings attended by as high a person as you possibly can. This has two main consequences:
1) The meeting gets attended be second tier first-tier wannabees who see the easiest way up the tree is to YesMan their way up on the coattails of the superiors
2) the more 'important' people attend the fewer tech staff can attend and it becomes a management review.
Management are more easily distracted by shiny things so show them a nice interface. Whereas techs will ask awkward questions about actual function, tests, simulations etc
This goes against the usual rule of thumb for normal development which is never show management a working interface as they will assume that everything is complete. Also, never show coders a fully working back-end as they will assume it all works.
Next, keep fudging until the boss and minions are so far in that they cannot report the issues without landing themselves neck deep in the warm, brown and smelly stuff. Then starts the frantic firefighting and genuine teamwork between client and consultants trying to patch everything and hope they can fix it before anyone notices (similar to the morning after a teenage party when the parents are coming home later in the day and there is a panic trying to prop up three legged horse ornaments or gluing a vase back together with toothpaste).
Then people start looking for new jobs as part of their normal 'career development'. The replacement team comes in, smells something wrong and starts finding vomit in the sock drawer, the whisky bottles all seem to be filled up to the previous level with cold tea* and teenager insists that the TV always had that funny blue bit in the corner.
Oddly enough, over the years, I have been the teenager and the parent and the replacement team member in this scenario.
Once upon a time I got canned as a consultant (“Not a team player”) for pointing out that rewriting/customizing a vendor’s paycheck-printing engine months before go-live was a recipe for disaster. I was pretty damn senior on the tech being used and I could barely follow the vendor’s decades of accumulated logic and optimizations. Mostly because it was a monolith without any documentation or intent for customization.
The underpaid temps doing the rewrite barely knew the language. Their own people were dead keen on avoiding work. Which was good because they had only one guy who could code his way out of a wet paper bag and all he knew was COBOL
4 months before that pesky Jan 1 payroll switchover (no, Feb 1 won’t do at all on these) these clowns still hadn’t converted the source system’s tables to the new database after months of “ready next week”.
The (much larger) overall outsourcing contract was re-put out to tender 2 yrs later and Accenture was publicly asked not to respond to its RFP.
Here’s a bit of Wipro love, courtesy of their breach:
Seriously, who’d be clueless enough to hire one of those big jackass-at-all-trades consultancies for their customer facing site?
Wipro... A few years ago my department (IT in finance/banking) was outsourced to Wipro. I waited 3 months in hope that their employee payroll website will be working at least 8 out of 10 times when accessed. 3 months in hope that they make it look like a professional intranet page and not an IT assignment for secondary school students. Big words, promises of great future, thousands of managers from Bangalore delivering beautiful speeches. Bored and disgusted with all this crap I abandoned ship, never looked back and lived happily ever after.
So, standard practice? Every "web design" company I've dealt with did shit like that. Ignored specs then charged more to fix their mistakes. Since they usually dealt with technically illiterate salespeople-managers on the other side, this worked when wrapped in enough baffle-gab. And when I said "dealt with", I mean "worked for, then quit in disgust at their unethical practices". Anyone in the companies involved who tried to push for a more ethical, competent practice was either shoved out the door or told in no uncertain terms they weren't to interfere with a profitable business model.
I hope Accenture is sued into the fucking ground on this one, and that other clients of theirs get the courage to stand up and do the same.
Never say you cannot do what the Customer asks , especially if you cannot understand what is being asked for !!!
Once the contract is signed, hire the cheapest 'bodies' that can do the job and charge them at the highest rate as if they were the ultimate skilled people on the planet.
Everyone then goes into 'wing it' mode, the Senior Consultants use FUD to hide the lack of knowledge and progress. The junior consultants do most of the day to day leg work and lie if it is needed.
All the 'grief' is passed to the cheap bodies they have hired and they are run ragged trying to do the work, timescales and quality are the casualties.
Eventually when the 'problems' become 'issues' the Senior Consultants go back in to try to convince the Customer that it is really under control and unforeseen 'issues' that the Customer has are their fault and for a 'small' fee the project can be rescued *but* there will be compromises to the original Spec/functionality to meet then 'New' timescales for delivery.
Sound at all familiar !!!
None of this story as reported is new, I have seen *all* the same issues on projects that I have been indirectly involved with. (I.e. providing a small related project to a 'BIG' firm managed project and seen the games that they went through. Yet they always got more work even though they never delivered to the original contract.)
Popular trick they tried was to request (via side channels) you changed your deliverables to a very much later time due to them being 'not ready', then trying to blame their issues on your now late delivery.
Did not get caught by that one by making sure every action was confirmed 'Directly' with the client and covered by hardcopy confirmation of the customers understanding of the request being from 'them' and the full understanding of the consequences to the small related project. Usually this meant that the Customer would have 'words' with the Consultants to clarify the reasoning !!! Must often this was the last you heard of the request :)
I once had a ran a web design project where we ended up having to use an agency we had used previously (who had mucked us about plenty). However due to the preferred agency not passing due diligence it became our only option.
However, I insisted on a 'watertight' spec for the site. It was only a smallish site, a total cost of less than £100k. The agency however started to be crap like usual and delays started, they didn't fix stuff until threatened with non-payment. The most comical bit though - part of our spec described a reporting function that was required. It had full details of the reports it had to produce. However 50% of the time it would fail as the report was too big and the code too slow that it timed out. They claimed, however, that the function had been built and produced reports as per the spec, however the spec never said that it had to work all the time, or even that it had to work. To make it work all the time would cost extra money!
In the end I spent a day looking at the code and realised they were just doing a SQL table grab from multiple tables and then joining and manipulating in server code. I managed to rewrite the SQL side to return the already joined data and push it back from a stored procedure - worked in about 100th of the time. I'm no DBA but the fact I could fix it in a short time compared to the months of wrangling over whether something being in the specs means it should also work seemed pointless.
The issues with these web projects is that the agencies work on a project at a specified 'day rate' per staff member. Your company wants a fixed price as you would never trust an open ended agreement to produce a product and then pay the cost at the end. As soon as the day rate total approaches or exceeds the fixed price agreed the account manager gets twitchy and starts cutting corners all over the place to produce the bare minimum rubbish. The project spirals downhill from there.
Sounds like you're presuming that the same clown who wrote the server side code wrote the stored procedures. My money would be on these being the responsibility of separate teams, so rather than use an efficient for-purpose stored procedure, they improvised with what was already there. As far as performance testing, that's another team right? As for showing any form of initiative, if they ever had any then I am convinced that their training involved beating it out of them.
The trouble with "fixes" like this is that they are all special cases. Yes, I could (and have) done that kind of fix for a specific problem in the past, but I'm not generally happy about it.
You need to find a solution that works inside of the framework you're using, so that it can be maintained by the next person without having to read some documentation that says, "go check this code in a foreign system" (i.e. a stored procedure). You need to make sure it works on different databases, including local development and demo systems. You often need to make sure that something *doesn't* use stored procedures because the hosting provider doesn't give you the permissions to create them. You may need to create and test some new schema migrations.
You need to make a design decision:
* do the same for every report
* document that this report is "special" and how it is special
You need to check for any other code that depends on this and make sure it's compatible.
You need to write tests for your fix, and run it through QA. You need to profile it to make sure that it's not having any side-effects like locking tables while the report is processing and causing your payment gateway to mysteriously drop transactions.
Even the golden path through your fix would take a lot more time than you're implying. If the developers got 90% of the way through and discovered that there were knock-on effects that would take longer to fix or cause more problems than the original bug, they'd have to start again and come up with another solution.
I'm not saying agencies don't inflate their charges and rip people off. I've worked for a few, and seen the behaviours of a lot more. I'm relatively lucky right now that I'm working for one that's (usually) managed to give their clients results at or under budget.
But $32 million? I'd do it for $31 and give you 20% commission. And I'd have it done by tea time.
Also: yes, that site does look a bit shonky.
Just about everything you have said isn't correct. I mean even the bit about hosting providers? Who on earth would be dictated by a hosting provider what you can run on a machine. They wouldn't even know what you are running on it. You buy your machine and you load an run whatever you want. We're not talking a mom and pop store or a wordpress site or some kind of shared hosting thing. You also don't need to 'copy' over new features to dev, test, live etc. You use proper version control and deployment systems so everything works and runs the same at the structural level.
You don't break other stuff by creating a stored procedure or view .. etc etc
The company I work for haven't paid anywhere near 32m for a front end website, but we have used a few companies recently.
The jist of it, it that for 6 months they did a p*** poor job and knew they could get it past our sales / marketing team. Problem arose for them when the marketing team was so impressed they showed the dev team... we were less impressed. 12 months, £200,000 and every developer in that room knew they could of scratched it up in a week. It was so unpolished, broken libraries... it was a wreck.
Anyway that company that also wanted £32,000 a month for ongoing costs (website updates and marketing support) was binned a couple of weeks after the show and tell.
They weren't happy that the dev team had looked under the sheets and taddled on them :)
Chris Collins 1,
It is an old problem !!!
To use your own in-house devs you need to go through the correct proceedures and your 'Good idea' may get shot-down in flames because it conflicts with other projects or is part of a 'Bigger' project yet to be initiated etc etc. Also because the company has other depts that need things too (Quelle Surprise !!!) the time to delivery may be 'too slow' as others are in the queue before 'Sales/Marketing' !!!
Basically, Sales/Marketing always want something *now* with no regard for how it may impact others.
They therefore reason that if they get an external set of devs to do what they want *now* they will get exactly what they want without having to navigate the internal procedures etc.
Usually there is a way to get these things done if the costs can be signed off at a high enough level, this 'works around' the internal devs !!! (Or so they think).
As per the article, the punchline is when they show the 'New stuff' to the internal devs/Support staff etc when they need to get it included as part of the 'supported' systems the company uses. (So backups/DR etc are including their 'New stuff'.)
The usual way to 'encourage' this *not* to happen, is to highlight up front & companywide that any systems that are independently 'acquired' are *not* covered by the existing Support staff inc devs.
These systems are treated as you would treat any other external customer and would need a properly costed out support contract that Sales/Marketing would need to pay for. The cost of this when advised after the proper analysis of the requirements usually makes these 'Quick & Dirty' work arounds not economic ..... particularly as Sales/Marketing often don't (want to) recognize the real costs of providing the support they need for their systems.
"Anyway that company that also wanted £32,000 a month for ongoing costs (website updates and marketing support) was binned a couple of weeks after the show and tell."
It took that long?
"They weren't happy that the dev team had looked under the sheets"
The history of website cowboys goes back to the beginning of the web.
I know one company which paid a couple of million for something so bad that their print team knocked up a better site in an afternoon just to get rid of it.
And I'm a lot cheaper than those boyz. And I'll actually get the site out the door, using your specs. I'd need to look at the specs closely before I can give you a go-live date. It's already the end of April, though, so don't count on anything before the end of December unless this is a really simple site, which it doesn't seem to be.
The existing Hertz website had a big flaw. What actually happens in real Hertz life (at least in USA) is that you turn up at the pickup point and they say "see that car park full of cars, well just pick the one you like." I actually like this. The trouble with the website is that it doesn't explain this at all and if you search for a car between this date and that date it off-puttingly gives you what looks like a choice of only one model of car.
Not reading any other docs but it seems that Hertz was blind-sided by Accenture when deliverables weren't and when they were totally deficient.
Doesn't a big corp like Hertz have the wherewithal to allocate a few highly talented contract managers who know the requirements to do some week-to-week monitoring?
Hertz probably has some "contracts" people. But one has consider their competence and how much they can see through the BS and if they have a tech background and not a sales droid background.
Plus I'm sure there were some excellent business trips for meetings on this with many benefits especially if the contracts people have the marketing mentality in that they buy the BS shoveled at them.
You would be surprised at just how small the IT departments are at many, many large companies.
Every damn thing is outsourced. The permanent staff are busy putting out fires or mired in pointless meetings. Or about to get fired and replaced by... outsourcing. Or their workload increases beyond reasonable and safe limits. Generally, all three.
The great collapse is coming. People have no clue how vital IT is to the modern world and are literally cutting their own throats.
To be fair, the typical definition of IT would include a lot of things related to the website, such as where it is being run, where the database is, and what the spec would be, but the writing of the code that runs the site would probably not be a specifically IT thing. It mostly depends whether programmers and administrators are both lumped into IT or not. In most places I have seen, these groups are in different departments and simply connect. On that basis, it's not the IT department's fault if the programmers, in this case outsourced, fail to write the code properly.
I think the issue of outsourcing is rife across the IT and programming departments of large companies. Many, most, of the higher up managers either don't understand what their IT and programming departments do so can't reasonably evaluate the importance of their work or don't care because outsourcing helps move expenses to a different balance sheet which then makes the department they run look better fiscally. Whether anything remains functional doesn't seem to matter. Greed and MBAs will and probably have already been the ruination of many a company. This is becoming so prevalent that it will hurt national economies and all that brought to us by greedy, stupid and corrupt politicians.
Not so much not knowing what IT do, but more of a case of being told by expensive consultancies that "non-core business" shouldn't be part of the competency of the business.
This makes sense.Should Hertz (for want of a better example) have business expertise on how to clean offices?
The problem with this in IT is that many businesses (and Hertz are very close to this) are increasingly IT businesses that specialise in something.
Eg. "IT Business that specialises in finance" is more commonly known as a bank.
But much of the (senior) management and the most visible functions (especially sales and accounts) have yet to realise this.
This makes sense.Should Hertz (for want of a better example) have business expertise on how to clean offices?
This highlights the falsity of this argument. The client company needs staff who can actually monitor the bought in work.
A company's staff can mostly identify whether the offices are being cleaned according to contract ( as long as the contract makes sense). As an educational manager I checked the premises and held the contractors to account. But even there I met limits. I couldn't check everything always and some contracts were negotiated well above my pay grade, usually by people who had no idea of how we used the spaces. (As in a padded secure room for kids with serious behaviour issues tends to need a bit more time for cleaning than an ordinary classroom- and yes we did have that argument with the bean counters!). Even with cleaning there are technicalities that need expertise from the client to ensure the job is done correctly.
Outsourcing the capacity to monitor and understand the service that's been outsourced means that you are leaving yourself open to be ripped apart. A big IT project needs to be monitored by staff who understand what's going on. Sadly the accountants that promote outsourcing seem to base their calculations on the entire role and skill set being taken off the payroll when in reality you need to retain a percentage simply to write the initial contract and then keep control of the process.
Outsourcing the capacity to monitor and understand the service that's been outsourced means that you are leaving yourself open to be ripped apart. A big IT project needs to be monitored by staff who understand what's going on.
That can also be part of the problem. So big IT project, hire project managers and delivery consultants to manage the outsourcer.. Who ideally have lots of experience managing that outsourcer, and know the tricks they play. Like knowing that those contractors will be looking for the next big project, once this one is delivered. So sometimes those watchmen aren't necessarily best aligned with the customer's interests.
But that's outsourcing for you. IT's still too often seen as a cost, rather than the glue that holds many businesses together.
The throat cutting continues even after outsourcing. I work for an IT firm that has numerous contracts across the world for outsourced support. I've seen some customers for whom we provided a proper managed service cut that so they're just "hiring the engineers" through us, then wonder why support quality drops without anyone to manage ticket queues, SLAs, schedule work, etc.
Now I've seen that customer look at pulling on-site support completely from certain locations. imagine a site with 250 users and no local IT support at all (only remote L1 support, and an asset swap service for broken laptops). They save 1 engineer of costs, wonder how long before employee downtime wipes out that saving.
AC for obvious reasons...
"They save 1 engineer of costs, wonder how long before employee downtime wipes out that saving."
Target on headcount but no obvious way to apportion the employee downtime attributable to poor local support so gets lost in general noise (sickness, turnover &c)?
"The great collapse is coming. People have no clue how vital IT is to the modern world and are literally cutting their own throats."
The Great Collapse. The incessant drive to reduce cost and cut corners. The constant struggle of technical workers against tech-illiterate managers/bean counters. Not to mention the sheer lack of skilled workers (how many of them are working on "useless" software like Snapchat, TikTok, etc?). It's gonna eventually hit a wall but not until vital aspects of life are "computerized", e.g. monies become totally virtual, connected vehicles, planes, homes, etc.
Companies that outsource development to monster consultancies are both lazy and deluded.
Deluded, because they imagine that the consultancy has special access to a massive pool of talent that the company doesn't have. The opposite is true: read the comments here to find out whether people like working for consultancies.
Lazy, because they can't be bothered to run their own project.
"Doesn't a big corp like Hertz have the wherewithal to allocate a few highly talented contract managers who know the requirements to do some week-to-week monitoring?"
On big IT contracts your average Contracts Manager does not have the skills, particularly when you are required to 'Translate' the information you will get off the Consultants into English and then decide if you are getting what you paid for, on time and to spec ???
Contracts Managers responsible for these projects need IT 'Inplants' in the Contracts Dept to help them understand what they are getting 'Day to Day'. Not only does it help the management of the Contract BUT it also enables the companies own IT Dept to get early 'Heads up' on problems that are coming down the line from these external projects.
Note: I am not saying the Contracts Managers are dumb, just that there is a limit to how detailed their knowledge can be and they should use the in house 'Experts' to ensure all is as advised by the Consultants.
This is another example of companies being hurt by the various functions in the company being reluctant to let 'outsiders' into their domain. A little bit of cross-fertilisation between depts would be good for everyone and many problems could be caught earlier.
This shit has been going on for at least 20 years at Accenture. I had friend who left there in the late 90s for this very thing - low-bidding jobs they knew they could never complete for the quoted price, pitching the job with senior partners, and then immediately replacing them with no-nothing college grads billed at $200/hr as soon as the project started. And they knew full well that years into the project they had the client by the balls and could start adding costs at will. Which they did. Want to make partner? Put away your conscience and morals and play along, and soon you will be the guy making the pitch and leaving a steaming pile of expensive poo at the end, all while collecting massive bonuses for a job well done. Their business model is by design, and they get away with it far more often than they should.
One of my ex-colleagues claims Accenture’s only core competency is the contract renegotiation specialists they bring in to smooth feathers.
“They’ll leave you thinking a 6 months deadline extension ‘to ensure quality first’ was actually your idea” is how he put it.
But these are not the day-to-day PMs you get from them. No sirree.
I think most large consulting or software dev houses do something similar - it's just the classic bait & switch. Put your most competent people front of house for sales pitches & initial bit of work, then move them on to the next pitch asap and replace with your cheapest employees (not that you change what you charge). I've seen it so many times, and most times the relationship between the companies is so weakly worded that the consultants can get away with it. It looks like in this case that Hertz at least has some good documentation showing that Accenture fubar'ed the work, so I hope they win their day in court and take 'em to the cleaners.
>It looks like in this case that Hertz at least has some good documentation showing that Accenture fubar'ed the work
I would hope that someone at Hertz kept a journal of all the Accenture powerpoints promising how wonderful the website would be and progress against milestones...
> I would hope that someone at Hertz kept a journal of all the Accenture powerpoints promising how wonderful the website would be and progress against milestones...
That would be "sales collateral", and probably not admissable in court as all sales material is generally labelled as a work of fiction. Dated, reviewed and signed product specs such as those listed in the article (responsive design, non-PDF documentation, etc...), plus emails/server logs showing transfer of those materials to Accenture, are much better
The law of averages suggests that they should have successfully hit at least one of the requirements or standards. But you have to hand to Accenture they didn't let the usual standards of incompetence stop them. Nope it sounds like they managed to dodge every single requirement or standard.
Congratulations Accenture! I'm impressed
Why you'd choose a bunch of accountants to build a rental website is beyond me, however it must have made sense to the long line of Noah's as they boarded in pairs. I can only imagine the animated conversation sinking in, regarding, the ever increasing cost of business liability insurance. #aBpleaseBob
You're right 32M doesn't sound bad. It might sound to good to be true, but enough not to be an extreme low ball. I think it'd be great if Hertz drags this through a court that won't allow publication bans, so that we might learn more of the massive incompetence involved.
And that's where it gets complex..
Client "Once you've collected the information we need you to load it into our mainframe back end. It's an MQ/CICS/DB2 stack"
Con-sultant "What's a mainframe?"
Can't really see what went on but a typical scenario like this is that you can guess that they have a 60-year-old sales rep that can't even use an iPad and goes to the client and says yes then comes back to the developers and tells them what they have promised and then the clients never stop asking for more features before it goes live.
We have all seen it: The type of company that puts marketing staff in the IT department because they use a computer for something other than Word.
Should have hired me 20+ years web developer + App developer, experience matters
"Oh hell yeah. Fuck the marketing department sideways. Worthless arrogant assholes."
And therein lies the problem. Every department thinks it's the most important one in the company, without whom it would crash and burn. The reality is they all depend on each other. But I agree, marketing (and sales!) are usually populated by arrogant arseholes.
I'm not so sure this is the usual "client continualing asking for more" failure. Just the things mentioned in this article were in the contract. I mean if they're in the contract that means they were there from day 1.
So No, this looks completely like an Accenture Bonfire. Hertz just got to feel the hurts....
>Should have hired me 20+ years web developer + App developer, experience matters
So starting to get closer to that "60-year-old" than the wet-behind-the-ears college graduate...
> the clients never stop asking for more features before it goes live.
I thought that's one of the reasons we did 'Agile'...
Scrapping the bottom of the barrel: IBM, Accenture fight for massive Australian Defence ERP deal
I think you're neglecting the important part of this potential contract and that is that a few politicians and higher level bureaucrats will be able to stuff their nests for retirement. Functionality, who cares?
To be honest, it shocks me that these behemoths, with these terrible track records are allowed to bid for government or large corporate contracts. You'd think they'd have burned those bridges by now. I think it shows the lack of understanding of the technologies that now run the businesses of the world.
"To be honest, it shocks me that these behemoths, with these terrible track records are allowed to bid for government or large corporate contracts."
Because they are the only ones big enough to do it. Governments don't have competent people to mange large projects with many small contractors. The only option is to go to a one-stop-shop. And the big boys don't want to use open standards because they want vendor lock-in. They will always have plausible reasons for doing so, even if the Government usually mandate the use of open standards and interoperability.
Don't get me wrong, there are some very talented people working in the field of web design, people who do know their stuff and do it very well but they're as rare a a sunny bank holiday Monday.
You've got it backwards. Ever since website building was considered software dev, we've got programmers who add so much cruft everything turns into a resource hog.
Low load, high performance website building was mastered in the last decade. All you had to do was throw faster servers and switches at your traffic growth. You could scale up fast. Now? Broken, non-compliant, non- standardized code and scripts everywhere.
Do I blame programmers specifically? Not really, just cheap overseas ones (or inexperienced college interns) that everyone now uses.
Oh, and the fucking marketing department as always.
I'm involved in some of that. 30+ years in Client Server or Server sided software , db, etc development, including mainframe stuff (ok, don't say that TOO loud), I find myself having to do FULL stack development.
I'ts shite, I'll say. TOO many stacks to thru to do simple things.
HOLY BatSHit batman !!!! Now something is borken, where the FSCK is it?
My point - ME with over 30 (ok 40) years experience, who can explain all sorts of THEORY, write my own database engine, have read Knuth's books, can barley keep things together, HOW IN THE HELL can some place like Accenture with JUNIOR programmers even make an attempt at things of this size and scope?
For reference just look at what happened here in the USA and the first round of shit programming on the "version 1" of our own ACA website ....
If you're referring to the Enron scandal that enveloped then destroyed their accountants Arthur Anderson- which Anderson Consulting/Accenture was a spinoff from- that had nothing to do with it.
Quite the opposite, Accenture is generally considered to have been *very* lucky to have avoided guilt by association by (coincidentally) changing their name at the start of 2001, just months before the problems at Enron became public and rapidly snowballed.
Wikipedia claims that this due to a court-ordered name change as part of an arbitration case that resulted in the final severing of ties between Arthur Anderson and AC/Accenture less than six months prior.
So, yeah, they were very lucky to have severed ties *just* before the scandal hit, but Enron wasn't the reason AC/Accenture changed their name.
Accenture couldn't deliver a web site and mobile app with 12-18 months? Who would have thought it...
As for the spec, if you deliver what they asked for (which Accenture didn't even get that far) it's usually not what they wanted anyway.
In-house developer(s) would know this and create a product that is "wanted". Plus the business would have line-of-sight of the development progress.
It's definatey be cheaper than spending $32m...
I'm reminded of back in the '90s the institute wanted a computer mouse ordering system to replace the paper one then in operation. The spec was that it be a web page on the intranet so it was platform independent. This was good news for us as we had just got rid of our last windows machine having gone full Mac.
The head of animal services, a vet, got his nephew to do it (keeping the funding in the family) and what was developed was a windows only program. We had to buy a windows box whose only function was to order animals (mice).
How you get from 'a web portal' to a windows program is beyond me. I expect the nephew told his uncle 'I can't do that' and thus the spec was consigned to the round filing cabinet.
Having worked for Avanade, which is half Microsoft, half Accenture and is meant to be the MS technical sister company (we're apparently Microsoft's largest single customer and supplier or so I've been told) I have seen both sides.
I hate these large corporate consultancies for the very good reasons mentioned here. My current project is horrendous. I'm doing the infrastructure for an Azure application development project. (Web, naturally). Everything is PaaS, and of course it's all (fr)agile. There literally isn't a spec or anything, they are building as they go, and the client is changing their mind every few weeks. Literally a new Azure technology is being trialled and then included on a weekly basis. We're due to go live in August, and I have no LLD for the Infrastructure and no solution design for the application at all. It's a fucking disaster.
On the flip side, at my last client, we were looking at being 3 months late for delivery. Was a migration for a well known brand from 7 different mail platforms (Exch 2010, 2013, EXO and Notes) to a single O365 platform. During pilot we told the client we've hit a few snags, and the 22,000 mailboxes won't be done by the end of 2018 and will be Feb instead. This was after 2 years of working on the project. Client pulled us off the project and literally a company formed 3 months earlier who's directors are best friends with the client CTO took ownership. They guaranteed they could provide the same business benefits as our migration by Dec '18. We said we cannot use their method as it's batshit crazy and MS won't support it. A few of us (5 out of circa 50) stuck around for support and niche bits, whilst the 3rd party implemented a mesh (federation) between all 7 organisations. (So co-existence, not a migration).
I rolled off end of Feb 19 and it still wasn't in place, and the migration would still need to have been done afterwards, and by Feb '19 we would have completed the migration too for the same fixed price.
Things aren't always as black and white as they seem in a single news article. Accenture are a bunch of money grabbing wankers, but then again have you seen their clients?
Anon for obvious reasons
>Was a migration for a well known brand from 7 different mail platforms (Exch 2010, 2013, EXO and Notes) to a single O365 platform. ... This was after 2 years of working on the project. ...
Back in the late 1990's, I did a mail integration and migration project for a newly merged global company: 27 different email systems, 20,000 accounts completed within 18 months of the merger. Yes (fyi) the new mail platform was MS Exchange. Combined client and integrator team was 10 people - however, no one from either MS or Accenture were involved.
>Back in the late 1990's, I did a mail integration and migration project for a newly merged global company: 27 different email systems, 20,000 accounts completed within 18 months of the merger. Yes (fyi) the new mail platform was MS Exchange. Combined client and integrator team was 10 people - however, no one from either MS or Accenture were involved.
The problem was probably not the migrations - those would have taken a few weeks, maybe a few months at most (depending on size, number and risk appetite). The problem is usually tidying up all the legacy crap from previous migrations.
Currently 18 months into a similar project and still having to clear up Ex2003 objects left behind from that migration, CCMail stuff left behind from that migration, Notes stuff left behind from that migration and lots of non-standard usage around everything. And we've still got the legacy apps that send email to migrate...
Apropos of the general theme, my current employer and the one before last are fucking Vikings when it comes to fixing the spec on signing the contract and charging a fortune to change it. It works out better for everyone; people actually think about what they want before they agree to buy it, and it doesn't change unless it really, really needs to.
Agile, whatever it once was, has just become a way everyone involved avoids having to think about what it's actually meant to do. While it may well have started life as a defence against incompetent customers who don't know what they want and the incompetent sales people who sell them something without actually identifying what it is they want, it seems to have become a way to actually embrace that incompetency and let it flourish.
Noticing the countless millions involved, it will probably be more like a global car reservation system, usable through the browser, including API's for third parties to reserve cars as well.
Since it is global, not only coding issues arise, but also taking into account all the differences in laws, insurances and other rules that exist between countries and continents for renting cars.
With such large projects, prototyping on a smaller scale, hiring a third party to evaluate the proposed technologies and the project approach are critical.
Too bad details concerning these events are always kept away from the public, it could be an useful learning experience.
This story can't be true. This is a private sector firm contracting another private sector firm and it's gone pear shaped? The private sector is a perfect exemplar of market efficiency and gimlet eyed competence so it can't have happened. I know this for sure because every time something like this happens in the public sector the wise pundits of the El Reg comment pages start muttering 'wouldn't happen in the real world'.....
I honestly never thought I'd hear myself saying this, but it sounds like hertz could have avoid this by getting Accenture to deliver "agile stylie" [sic].
This would have given hertz much earlier visibility of the f***wittery of accenture given agile constant delivery of value. Agile has it's faults, I'll grant you, but Hertz could have also adopted an Agile payment framework, making payments only when fully functional code and capability was delivered, addressing defects as they were found.
However, I am aware I'm a "glass is half full" kinda guy and also wish for world peace...
It sounds like they developed the system, then tried to make it responsive. As if they started from scratch with no code reuse.
Wouldn't it be just a trifle better to find or develop the best responsive framework that they can and use it across most of their projects, fitting each one into the framework with whatever adjustments are needed?
That would also make maintenance easier as all their staff would know how that framework worked.
The same goes for the back end stuff - a company that big ought to have lots of existing general purpose code which can easily be tweaked to fit.
It's almost as if they handed the entire analysis and development process to completely inexperienced staff who had never worked on any other project for them and had no access to the code for their other projects or any advice from the experience gained in building them.
Since the entire point of using one of these firms rather than developing in-house is to take advantage of their existing experience in doing this kind of development this seems, shall we say, a trifle odd.
... but having been involved in one of these big tendering wheezes, I've seen a mindset that can only be described as conspiracy thinking between the client's managers and the consultants. It goes something like this: "You guys have done loads of these jobs, right? You're men of the world-- if we tell the sodding beancounters how much this thing is really going to cost, they're going to walk away right now. So you scratch our backs and we'll scratch yours: you go low on your estimate and take an, ahem, flexible stance on any change requests we lob your way, and we'll be likewise flexible on signing off on any overtime requests you need to get the thing over the line. How does that sound?" Except not said in so many words, of course. Certainly not in any way that could be taken down and used against any of the parties involved, nudge-nudge, wink-wink.
Of course, doing business this way basically means the company's IT department is in a conspiracy with an outside consultancy against its own management and finance department, and all those estimates, requirements documents, lists of "deliverables", timelines, and other precious artefacts of the tendering process are so much junk from the get-go. The project is off the rails from the start because the project scope is a fiction concocted to get sign-off and nothing more. And so it ploughs on, burning money, until the political embarrassment becomes too much and the managers call a halt going, "it was the filthy consultants! They made us do it!" It's a win-win for the IT managers, see: if the gamble pays off, you collect the ass-pats due to one who successfully delivered a critical infrastructure project; if it doesn't, well, it was only someone else's money anyway and you have a handy scapegoat to blame. "It was the consultants!" Yeah, that wouldn't ring so hollow if the exact same consultants weren't back six months later working on the next white elephant...
"So you scratch our backs and we'll scratch yours"
Ok, so, whichever side you're on, in the end, who wins ? Nobody! Management will think internal IT are a bunch of morons, which they are, and the consultancy company will get a bad rep because they failed to deliver.
Motto in software projects: always start small, add features later.
Rules in software projects:
1. once the speccs are done, and contract signed, NOTHING gets added
2. nothing goes live before all boxes are ticked.
3. changes are to implemented with agile-like methodology.
3.1 AGAIN, nothing goes live before all boxes are ticked, including any new features recently added.
3.2 goto 3.
Why Hertz decided going live with a website from hell is anyone's guess and they hired Accenture, which is enough evidence for me. Accenture is #SureToFail - it's the Craptita for the rest of us.
> once the speccs are done, and contract signed, NOTHING gets added
In the real world things do change. All that change control is all about patching that gap (badly).
Agile takes the opposite approach (assumes things will change) and so does not lock things down. But it also means there is no fixed timeline (because there is no fixed scope).
(Of course most agile projects being "we call it agile but really we're trying to do waterfall but not being honest" fail by being neither willing to change nor willing to not change.)
Well, you have speccs, you design to them - the base MUST be clear. Once you have the base, agile is best to implement any enhancements, changes etc.
That is my personal $0.02, YMMV, my projects DON"T fail, though ... agile is like everything, too much of the good stuff kills it, just like waterfall, in the long run, unsustainable.
This stuff is true at all the large 'consulting' outfits and not just for IT.
When I was at a large aerospace company, we had a couple of Wipro contractors who were made responsible for designing a processing card to go into a mission computer. The reason was 'offsets' which are rife in that industry.
Fortunately, this had to get reviewed - by us.
The first design review comment I put in the book was 'Will never power up'; I won't go into the details here, but it was an error that no engineer with a few (perhaps 3) years of electronics should ever make.
They were heavily supervised after that (and it took me weeks to make them fix the circuit). All that extra supervision probably cost more than doing it all ourselves.
The amount of times I have come across Accenture (and the others) projects that are an utter pile of shit is shocking. I dont know why anyone would even consider using them - its not like they have a good reputation for delivering technical projects! There is a whole industry polishing the tech turds Accenture, Wipro etc drop on their customers.
Accenture deserve to be sued, but I blame Hertz for this, companies need to own their solutions if its core to their business and maintain technical oversight and day-to-day governance of whats being delivered and built. Also, fixed-price/fixed-scope contracts never end well if its complex with multiple technologies etc.. Hertz allowed Accenture to fuck them over - they probably thought they were getting a good deal and I dont know why they are surprised with what they got in the end.
yes, actually, many times. the single requirement to get Accenture to do a great job is if the deal is gainshare. If the solution works as advertised and delivers business value/savings, Accenture gets paid. If it doesn't, they don't.
anonymous because I worked for them and in this industry, you don't want to be blacklisted by them
They are doing something right though as a wider organisation? Accenture share price has just hit a new record $180+ USD per share with a market capital of over $115bn, so unfortunately and in conclusion:
a) $32m isn't really going to make any difference (if they lose)
b) they can afford the best defence
c) this will carry on and their growth will continue
d) what they hell were Hertz doing in the past 2 years that allowed this to get to this stage? Was it a T&M or fixed fee contract? If it was fixed fee then clearly the change request route was used - all suppliers do this and in my experience Accenture aren't the worst at this - the pureplays are masters at it
e) The client complained, accenture rotated the team and the client was surprised that the incoming team didn't have the same level of client experience? Duh?
Out of curiosity I had a quick look at their UK web site and pretended to make a booking.
I literally could not find any cars. I was stuck in an endless loop of "select a location near you" hell.
Surely, being able to book a car is a basic requirement for a car rental web site?
I was looking for a car to rent last year, I used the obvious search engines as well as looking at the offers on the sites and, Hertz site experience was abysmal, well beyond the worst experience .... "useless" is the adjective that comes to mind. I booked elsewhere, though I really wanted to book one of their southern German brand cars ... I felt they did not want my business, fine by me, saved quite a bit but was not impressed by the car I ended up driving ...
My question is, have we gone full circle? Will business that should be able to afford to start to bring this type of work back in house? It seems to me that with the importance of a web presence and the retention of the data from the booking of services it would seem to me that having even a small, well paid team that can maintain your website(s), back end, and build in changes as required would be well worth the money spent.
This is a larger problem with the "Big" Advertising, Marketing and Consulting way of doing business.
Hertz purchased services through Accenture, but it wasn't them doing the work. They perhaps managed it, but outsourced to external parties. The ones performing the work were 5-7 levels down the chain and perhaps multiple language and cultural barriers. Hertz > Accenture > Accenture Web > External web developers > Manager > Developer Lead > People doing the "grunt work".
So symptomatic of enterprise dinosaurs like Accenture and the problem also stems from well established businesses assuming these overgrown organisations can deliver something tangible.
The problem is that these enterprises get successful over time but lose the ability to be dynamic and responsive assuming their patterns and practices will cater for all - on that level they simply cannot compete with the real agile organisations who tend to approach any project with a dynamic customer first attitude. Case in point is how Accenture saw the car hire business as a source of money and made absurd claims that XYZ would cost millions to implement (I am willing to bet the same work could have been delivered by a much smaller consultancy).
I hope this is a lesson for other businesses and that they should consider and hire smaller independent consultancies who usually are lighter on their feet and that much more capable of responding to the customers needs.
Someone mentioned Capita in the comments...haha I have seen Capita burn things as well they are not a good example!
I have worked for some of these companies here in Costa Rica, and the mess is the same.
My path with these big IT companies started when I was a Junior Developer up until I became Senior Developer and then I even tried other positions such as PM, SM, QA and other things, just for a change, but all the time it was the same. Badly developed systems, no practices, no methodologies, very few deliveries, excuses, excuses, and more excuses, team changes, and so on. It was terrible to have meetings with unsatisfied clients, always having to listen to the Account Manager giving excuses and promising things we knew would never get done.
One day I said it was enough, I was there just to get a salary, not to create something that I was proud of, so I thought I could work on my own and finally give real solutions to my clients. What was difficult was to find someone interested, but once I did my reputation skyrocketed and now I have several clients, I started my own team and I assure you I'll never look back.
It's not easy, every time I chat with my clients they give me suggestions, ideas, they have concerns and so on, but it's definitely not the same as before when my face was red all the time because I was ashamed of my own work, not because I was mediocre, but because the entire organization was.
Small teams work much better than big ones for most cases.
A long time ago, while I was still politically naive, I was on a medium sized project as a programmer. While waiting for the project leader to be available to answer some questions, I overheard him and the project manager discussing a minor issue. I took some notes, did some thinking, followed with some typing and by the end of the meeting between the PL and the PM, I had a nice, working solution. Unknown to me at that time, but told a year later, was that they had sent out an RFP to one of the major consulting firms, which came back as a proposal for three man-months work. In hindsight I suddenly understood why I was getting such nice, interesting little side projects to work on next to the main project.
No need to be anonymous as the only two persons to know this story and be able to add necessary details were firmly on my side in this.
I teach a senior-level university course in 'real-world software engineering', which focuses heavily on why so many large-scale software projects are seriously late and over-budget, or fail altogether.
It's been a while since I've seen so many classic failure patterns and factors in a single article. I'm linking to this from the class website and may use it as a case study for the Fall 2019 semester (just finished the Winter 2019 semester, but sent the link out via Slack to all my students).
I agree, it boils down to a project management and risk issue. Either there was a big amount of risk taken because the proposed plan wasn't vetted by someone with experience, or the communication breakdown resulted in a loss of project controls. Either way, great lesson to be learned vicariously.
It seems a thing that was wanted was defined. It was not delivered. In projects, I have spent much more time in defining what was wanted, than in anything else . Then pay stages are only by proven results. Defining exact payment milestones and getting agreement is 90% of all negotiation, plus your sub-contractor must have a bank to give bank guarantees drawable by client for any fault, if a) delivery of proven workable item is delayed or b) no workable solution is created.
I am not surprised at all, as a former Accenture employee. On my first project, the CEO came in the office, and kicked us out. He told us to hand in our badges and leave the building immediately. They sued Accenture and of course, Accenture tried to shift the blame on the customer. At the next customer, a senior manager was so angry he stood yelling in the office. Needless to say, I quit. They are so incapable it is unbelievable. Funnily enough, I talked about it later with other former employees, some went on to Deloitte or any of the other clowns, they told me, Accenture was much better, probably the best of the clowns! Their business model seems to be to undercut the market, underdeliver, then sue the customer and cash the bonus for the effort.
Beyond the shear mismanagement of this project which most likely lied on both sides... the irony is that Hertz paid a MASSIVE premium to hire Accenture because the thought “they were buying IBM”. I have been handling enterprise Websites for years at Jacob Tyler and we have under 20 people and while their may have been a hiccup here and there (I’m assuming this isn’t a 100% one-sided story), we would have 100% delivered for most likely 1/10th of the price... and would still be super profitable. It would be great if some of these fortune 1000’s would sit back and look at firms beyond the behemoths that sell you on the project with seniors and then bring on juniors to complete it. Good learning lesson for them.
I used to work for accenture. their performance and promotions system rewards narcissistic c*nts rather than talent, who invariably surround themselves with sycophants and low-cost dredge-level staff.
on my last project they hired a bunch of university leavers to write the core systems code for a webservices application. these guys made every mistake you would expect from naive, inexperienced university grads. meanwhile there were a bunch of highly paid contractors (friend of the partner, of course), who f*cked around all day deciding which application to draw the f*cking db schema diagrams in (the the schema that myself and a colleague had already implemented as soon as we'd got dumped on this nightmare of a project)
i left as soon as i could. i felt really bad for those university grads. they had no software architect to watch and guide them, or take ownership of the solution.
yes, they got blamed for the inveitable failure.
lesson 1 on big consulting contracts - never be the last person to leave the project.
Biting the hand that feeds IT © 1998–2019