6 ms·

Ex Deloitte here: In several years I never once saw the firm not deliver what they agreed to in writing. There’s one way to deal with a consulting firm: a roc

Ex Deloitte here: In several years I never once saw the firm not deliver what they agreed to in writing. There’s one way to deal with a consulting firm: a rock solid statement of work and people on the client side to verify you got exactly what you paid for. If the system didn’t work as agreed in the statement of work and that was clear, Deloitte would be footing the bill. If you gave out a time and materials contract without clear acceptance criteria, testing requirements, and verification / support / warranty detailed, you probably shouldn’t be handling 100 million dollar budgets.

h0l0cube 6y ago
> acceptance criteria, testing requirements, and verification / support / warranty detailed Some orgs just don't have the in-house capability to do this, so they seek a consulting firm... oh wait. Edit: I guess the strategy is to hire technical staff, keep them around a few years until you can be sure they can deliver on projects and understand your existing systems and processes, and then embark on the larger project to transform those systems and processes. Alternatively, shop around to find a consulting firm that is motivated to have good long-term outcomes. Probably a smaller operation, that has good word-of-mouth reputation, but doesn't have an established brand to fall back on.
- aplummer 6y ago
  Yep on a 100 million dollar budget you need to hire people to manage software delivery on your side and can never absolve end responsibility is definitely my point. I’m not describing how the world should work, just how it does when outsourcing IT to consultancies (which can work anywhere from successfully to very unsuccessfully)
- aplummer 6y ago
  To your edit: A good strategy is to hire a team, get a few consultants who have literally built the same system down the road, and use that blended team until 2.X release. That way you a team that has made a thing already (and often consultants are just as passionate about their work too!) ready to build a better 2.0, plus you keep responsibility and control.
  - h0l0cube 6y ago
    > A good strategy is to hire a team Some orgs may not know how to hire. I think here, competent (and well connected) consultants can help, and maybe source from their own networks, but we're back to the dilemma of which consultants to trust. And let's not speak of recruitment agencies :)
- harperlee 6y ago
  Or hire 2: one for all those things, another one for actually delivering technical work. Hint: the rivalry will make it also expensive (extra meetings, defensive bullshit). But you will end up with what you wanted, at a price.
throwaway_kufu 6y ago
> If you gave out a time and materials contract without clear acceptance criteria, testing requirements, and verification / support / warranty detailed, you probably shouldn’t be handling 100 million dollar budgets. That is oh so very Deloitte...look if we took $100M from you and you didn’t get what you wanted, that’s Clearly not our fault, YOU shouldn’t have been handling that type of budget in the first place. After all there are people like us out there that will hornswoggle you, luckily you can retain us Again to deliver the original contract and you should consider a separate retainer so we can protect you from people like us in the future. Id love to see a single instance where Deloitte admitted to fault and footed the bill for a $100M contract, care to link proof for such a claim?
- aplummer 6y ago
  I’m not saying how the world should work - I’m just saying how it does work. Why would they admit fault? The entire point is to take the hit and preserve reputation - you being able to find a link would defeat the entire point.
  - throwaway_kufu 6y ago
    I agree it’s the way Deloitte works and not how the world should work. > Why would they admit fault? That’s just it they never would. But reading your comment: > If the system didn’t work as agreed in the statement of work and that was clear, Deloitte would be footing the bill. You certainly make it sound Deloitte is a fair company that would admit when a system doesn’t work as agreed and then foot the bill, but that’s not the case. Whether they admit fault For delivering a failed system or not, do you have an actual example where Deloitte was magnanimous and returned $100M to a client after delivering a system?
    - sdoering 6y ago
      you state a lot in a very aggressive tone of voice. maybe that is because non verbal context is missing. I am not the OP, but working for an agency shop (way way smaller than Deloitte) I know we footed some bills. But we don't publicly talk about that. wouldn't be good for business. it's a double edged sword. I know of projects that were sub par. I have seen systems build by others (Deloitte included) that were nothing but a shit show. and I have seen systems from the same brands that were doing way more than originally expected. sometimes it is the client, sometimes it is the project management sometimes it is the consultants. and sometimes it is the circumstances. I wouldn't expect a system designed for normal load to hold in times of covid-19 and work flawlessly. some examples like the call center being overwhelmed is to be expected when designed for normal loads. I strongly believe, that there are real faulty systems to be found. and that for some of these the contractor is to be blamed. but without knowing the specifics I wouldn't feel qualified to judge from the sidelines. disclaimer: I work for an agency that got bought by a Deloitte competitor some years back. all opinions are my own.
dwd 6y ago
You also need a competent project management team (backed by contract lawyers) that can push back. The QLD Health payroll system debacle that ended up costing 1.2b is a good case of how not to negotiate and deal with one of the big 4 contractors. https://www.smh.com.au/technology/worst-failure-of-public-administration-in-this-nation-payroll-system-20130807-hv1cw.html https://www.smh.com.au/technology/worst-failure-of-public-ad... I believe Canada had a similar project fail in the same way - probably even the same software.
- aplummer 6y ago
  > You also need a competent project management team (backed by contract lawyers) that can push back Yep I agree with you totally.
- recursivecaveat 6y ago
  Yes, here north of the border it was Phoenix Pay. It seems to me anytime you pay 1.2B for a software project it is destined to fail from the start. The consulting company will overcomplicate things and cram so many cooks in the kitchen to justify the price tag that it's doomed. https://en.m.wikipedia.org/wiki/Phoenix_pay_system https://en.m.wikipedia.org/wiki/Phoenix_pay_system
loopz 6y ago
Yea, so fixed price and t&m while being agile and lean, am I right? Just because something got delivered, doesn't absolve the trainwrecks..
- aplummer 6y ago
  T&M only works for contracts for a few resources in a blended team in my view, with the only risk being a 2 or 4 week price on getting rid of them. Then you’re basically hiring overflow individuals you can get rid of easily who may have particular skills. Accountability only works with fixed price IMO
  - loopz 6y ago
    I've actually got good experiences with t&m, but that'd be anecdotal. The problems grow from when incentives become misaligned, and external work/consultancy is being used against coworkers.
    - aplummer 6y ago
      Yep it can totally work, but I think it’s a bit of a coin toss and I wouldn’t risk it with these sized contracts.
  - Spooky23 6y ago
    Only for trivial projects with tight or well defined scopes. It makes sense to pay $5/mailbox. $500k to deliver a system that does <x> usually not. Fixed price deliverables are usually more expensive due to bidders pricing risk. Or you see big delays due to PCR requirements. Or the project owner don’t have the political power to not accept a deliverable and tell the vendor to go pound it.
    - aplummer 6y ago
      > Fixed price deliverables are usually more expensive due to bidders pricing risk. Correct, that’s the only way the vendor takes on the risk really. That was my point. > Or the project owner don’t have the political power to not accept a deliverable and tell the vendor to go pound it. Yeah I agree... but if you have 100m to hand out and the in org politics where you can’t tell a vendor to go pound it, there are real structural issues in your co preventing you delivering systems.
Spooky23 6y ago
You are correct. I hate to defend these companies, but they deliver exactly what is asked for, no more. If you provide a vague spec or don’t have good governance of a more dynamic/agile project, the consultants aren’t the root cause of failure.
- moron4hire 6y ago
  That doesn't absolve them of their ethical duty to be good stewards of their client relationship.
- thephyber 6y ago
  There is a massive asymmetry to the two parties here. I'm not sure the consultants are malicious, but they enable the moral hazard of the worst of these types of contracts.
  - Spooky23 6y ago
    Oh no doubt. I’m not shedding any tears on these consultants behalf believe me. The folks in these companies are masters at separating customers from money, for sure. But, these guys aren’t ripping off clueless dentists either. The typical F100 customer isn’t a babe in the woods. When the parties are Accenture vs. Exxon or IBM vs. the IRS, who is the good guy? :)
mattmanser 6y ago
So if you're onboard for signing up for a waterfall process, you're going to be alright? You know, the process that's been shown again and again to be pretty useless in delivering a working software product? Great way to highlight how incompetent you all are! Honestly can't tell if you're trolling or being serious.
- myalphabet 6y ago
  Nobody said anything about waterfall. A good agile process can and should have all of the things the parent commenter listed. If you eschew those things by using the "well we're agile, that means we don't plan ahead!" excuse, that's a sign of incompetence.
  - dragonwriter 6y ago
    > A good agile process can and should have all of the things the parent commenter listed. A good agile process will not have an ironclad upfront definition of what is to be delivered in a large project. A process with that may not be waterfall, but it's definitely non-Agile Big Upfront Design. OTOH, Agile is largely about managing risk and incrementally delivering value without overcommitting forward, so the contracting process with Agile should reflect that and shouldn't commit to a giant project in a lump contract.
    - myalphabet 6y ago
      >A good agile process will not have an ironclad upfront definition of what is to be delivered in a large project. "what is to be delivered" does not have to be a rigid definition of "5,000 lines of code program with features 1, 2, and 3". A good agile contract will allow the possibility for changing project requirements but also still define how those changing requirements affect expectations and responsibilities of the consulting company. The agile process itself can actually be a deliverable that you define in a contract, and it should spell out, for example, how new requirements are prioritized and the criteria you use for determining if it's something that the consulting company is responsible for. In my experience if you can't write out your agile requirements in a contract, then you don't have a good enough internal understanding of your own agile processes and should probably work on that, too. Far too many people use "we're Agile" as an excuse for "we don't plan or document anything" and extend that to contracts, but that's absolutely not what good Agile actually is.
mlthoughts2018 6y ago
It’s right in the article naming Deloitte specifically: > “ "Deloitte also faced blowback in 2013 for an unemployment benefits system it built in Massachusetts. There, the benefits system arrived in 2013, two years late and $6 million over budget—only to be filled with glitches that erroneously cut workers’ benefits, according to the Boston Globe." Deloitte delivered a major project way behind schedule, over budget and full of defects.
moltar 6y ago
Ok but the problem is that customers don’t always know exactly what they want. Software isn’t like architecture. You can’t always predict every possible issue or corner case. And that’s where they get ya! They’ll win a lowest bid contract for the exact word for word spec. But then will bill wayyyy more for any change request. Which will be numerous obviously. That’s how actual project costs get inflated.
- ilikeerp 6y ago
  Which is totally reasonable. I can do X work for 1 million but if you throw Y into the mix then I obviously need to charge more. Want your oil changing? 100$ no worries - oh you didn't realise you also needed new tyres and we only found out when looking at them? That's extra.
lcrz 6y ago
> In several years I never once saw the firm not deliver what they agreed to in writing. I know people working with big consultancy companies and the recurring thing they complain about is a difference of opinion between client and provider about what exactly the terms of signed agreements mean.
harperlee 6y ago
> If you gave out a time and materials contract without clear acceptance criteria, testing requirements, and verification / support / warranty detailed, you probably shouldn’t be handling 100 million dollar budgets. This is the money quote. Big 4 companies (Deloitte, PwC, EY, KPMG) are historically accounting auditing companies. And their culture is heavily skewed towards billing their employees to clients per hour/day, or per unit of service, as auditors and lawyers have been doing for more than a century. It is not in their DNA to "create products" but to "render services"; that is, to sell employees availability. Boundaries and scope of those traditional services are typically clearer than building a new unknown program from scratch. If you are hiring these firms for exploration of how do you work, how you should work, and a complete functioning system that sustains a big number of new processes that are not yet defined, you are in for a big bill; you better be sure that you have put on the time to think about what you are buying in written form. When you see big failed projects like those the problem lies in that no one stopped to research the complexity of the ask before entering into a contract. Even then, these consultancies normally are very flexible when a contract is actively harming the client, and the client wishes to change it / stop it (it would not be good for the business to do otherwise, because you want to see more after this contract is finished). When these costs go out of control is because internal politics in the client impede the rational decisions of killing projects, changing command, etc., and the sponsor keeps feeding money into the project to save their ass. Source: worked in big 4 for years.
- aplummer 6y ago
  You couldn’t have written a more insightful comment.

h0l0cube 6y ago

> acceptance criteria, testing requirements, and verification / support / warranty detailed Some orgs just don't have the in-house capability to do this, so they seek a consulting firm... oh wait. Edit: I guess the strategy is to hire technical staff, keep them around a few years until you can be sure they can deliver on projects and understand your existing systems and processes, and then embark on the larger project to transform those systems and processes. Alternatively, shop around to find a consulting firm that is motivated to have good long-term outcomes. Probably a smaller operation, that has good word-of-mouth reputation, but doesn't have an established brand to fall back on.

aplummer 6y ago
Yep on a 100 million dollar budget you need to hire people to manage software delivery on your side and can never absolve end responsibility is definitely my point. I’m not describing how the world should work, just how it does when outsourcing IT to consultancies (which can work anywhere from successfully to very unsuccessfully)
aplummer 6y ago
To your edit: A good strategy is to hire a team, get a few consultants who have literally built the same system down the road, and use that blended team until 2.X release. That way you a team that has made a thing already (and often consultants are just as passionate about their work too!) ready to build a better 2.0, plus you keep responsibility and control.
- h0l0cube 6y ago
  > A good strategy is to hire a team Some orgs may not know how to hire. I think here, competent (and well connected) consultants can help, and maybe source from their own networks, but we're back to the dilemma of which consultants to trust. And let's not speak of recruitment agencies :)
harperlee 6y ago
Or hire 2: one for all those things, another one for actually delivering technical work. Hint: the rivalry will make it also expensive (extra meetings, defensive bullshit). But you will end up with what you wanted, at a price.

throwaway_kufu 6y ago

> If you gave out a time and materials contract without clear acceptance criteria, testing requirements, and verification / support / warranty detailed, you probably shouldn’t be handling 100 million dollar budgets. That is oh so very Deloitte...look if we took $100M from you and you didn’t get what you wanted, that’s Clearly not our fault, YOU shouldn’t have been handling that type of budget in the first place. After all there are people like us out there that will hornswoggle you, luckily you can retain us Again to deliver the original contract and you should consider a separate retainer so we can protect you from people like us in the future. Id love to see a single instance where Deloitte admitted to fault and footed the bill for a $100M contract, care to link proof for such a claim?

aplummer 6y ago
I’m not saying how the world should work - I’m just saying how it does work. Why would they admit fault? The entire point is to take the hit and preserve reputation - you being able to find a link would defeat the entire point.
- throwaway_kufu 6y ago
  I agree it’s the way Deloitte works and not how the world should work. > Why would they admit fault? That’s just it they never would. But reading your comment: > If the system didn’t work as agreed in the statement of work and that was clear, Deloitte would be footing the bill. You certainly make it sound Deloitte is a fair company that would admit when a system doesn’t work as agreed and then foot the bill, but that’s not the case. Whether they admit fault For delivering a failed system or not, do you have an actual example where Deloitte was magnanimous and returned $100M to a client after delivering a system?
  - sdoering 6y ago
    you state a lot in a very aggressive tone of voice. maybe that is because non verbal context is missing. I am not the OP, but working for an agency shop (way way smaller than Deloitte) I know we footed some bills. But we don't publicly talk about that. wouldn't be good for business. it's a double edged sword. I know of projects that were sub par. I have seen systems build by others (Deloitte included) that were nothing but a shit show. and I have seen systems from the same brands that were doing way more than originally expected. sometimes it is the client, sometimes it is the project management sometimes it is the consultants. and sometimes it is the circumstances. I wouldn't expect a system designed for normal load to hold in times of covid-19 and work flawlessly. some examples like the call center being overwhelmed is to be expected when designed for normal loads. I strongly believe, that there are real faulty systems to be found. and that for some of these the contractor is to be blamed. but without knowing the specifics I wouldn't feel qualified to judge from the sidelines. disclaimer: I work for an agency that got bought by a Deloitte competitor some years back. all opinions are my own.

dwd 6y ago

You also need a competent project management team (backed by contract lawyers) that can push back. The QLD Health payroll system debacle that ended up costing 1.2b is a good case of how not to negotiate and deal with one of the big 4 contractors. https://www.smh.com.au/technology/worst-failure-of-public-administration-in-this-nation-payroll-system-20130807-hv1cw.html https://www.smh.com.au/technology/worst-failure-of-public-ad... I believe Canada had a similar project fail in the same way - probably even the same software.

aplummer 6y ago
> You also need a competent project management team (backed by contract lawyers) that can push back Yep I agree with you totally.
recursivecaveat 6y ago
Yes, here north of the border it was Phoenix Pay. It seems to me anytime you pay 1.2B for a software project it is destined to fail from the start. The consulting company will overcomplicate things and cram so many cooks in the kitchen to justify the price tag that it's doomed. https://en.m.wikipedia.org/wiki/Phoenix_pay_system https://en.m.wikipedia.org/wiki/Phoenix_pay_system

loopz 6y ago

Yea, so fixed price and t&m while being agile and lean, am I right? Just because something got delivered, doesn't absolve the trainwrecks..

aplummer 6y ago
T&M only works for contracts for a few resources in a blended team in my view, with the only risk being a 2 or 4 week price on getting rid of them. Then you’re basically hiring overflow individuals you can get rid of easily who may have particular skills. Accountability only works with fixed price IMO
- loopz 6y ago
  I've actually got good experiences with t&m, but that'd be anecdotal. The problems grow from when incentives become misaligned, and external work/consultancy is being used against coworkers.
  - aplummer 6y ago
    Yep it can totally work, but I think it’s a bit of a coin toss and I wouldn’t risk it with these sized contracts.
- Spooky23 6y ago
  Only for trivial projects with tight or well defined scopes. It makes sense to pay $5/mailbox. $500k to deliver a system that does <x> usually not. Fixed price deliverables are usually more expensive due to bidders pricing risk. Or you see big delays due to PCR requirements. Or the project owner don’t have the political power to not accept a deliverable and tell the vendor to go pound it.
  - aplummer 6y ago
    > Fixed price deliverables are usually more expensive due to bidders pricing risk. Correct, that’s the only way the vendor takes on the risk really. That was my point. > Or the project owner don’t have the political power to not accept a deliverable and tell the vendor to go pound it. Yeah I agree... but if you have 100m to hand out and the in org politics where you can’t tell a vendor to go pound it, there are real structural issues in your co preventing you delivering systems.

Spooky23 6y ago

You are correct. I hate to defend these companies, but they deliver exactly what is asked for, no more. If you provide a vague spec or don’t have good governance of a more dynamic/agile project, the consultants aren’t the root cause of failure.

moron4hire 6y ago
That doesn't absolve them of their ethical duty to be good stewards of their client relationship.
thephyber 6y ago
There is a massive asymmetry to the two parties here. I'm not sure the consultants are malicious, but they enable the moral hazard of the worst of these types of contracts.
- Spooky23 6y ago
  Oh no doubt. I’m not shedding any tears on these consultants behalf believe me. The folks in these companies are masters at separating customers from money, for sure. But, these guys aren’t ripping off clueless dentists either. The typical F100 customer isn’t a babe in the woods. When the parties are Accenture vs. Exxon or IBM vs. the IRS, who is the good guy? :)

mattmanser 6y ago

So if you're onboard for signing up for a waterfall process, you're going to be alright? You know, the process that's been shown again and again to be pretty useless in delivering a working software product? Great way to highlight how incompetent you all are! Honestly can't tell if you're trolling or being serious.

myalphabet 6y ago
Nobody said anything about waterfall. A good agile process can and should have all of the things the parent commenter listed. If you eschew those things by using the "well we're agile, that means we don't plan ahead!" excuse, that's a sign of incompetence.
- dragonwriter 6y ago
  > A good agile process can and should have all of the things the parent commenter listed. A good agile process will not have an ironclad upfront definition of what is to be delivered in a large project. A process with that may not be waterfall, but it's definitely non-Agile Big Upfront Design. OTOH, Agile is largely about managing risk and incrementally delivering value without overcommitting forward, so the contracting process with Agile should reflect that and shouldn't commit to a giant project in a lump contract.
  - myalphabet 6y ago
    >A good agile process will not have an ironclad upfront definition of what is to be delivered in a large project. "what is to be delivered" does not have to be a rigid definition of "5,000 lines of code program with features 1, 2, and 3". A good agile contract will allow the possibility for changing project requirements but also still define how those changing requirements affect expectations and responsibilities of the consulting company. The agile process itself can actually be a deliverable that you define in a contract, and it should spell out, for example, how new requirements are prioritized and the criteria you use for determining if it's something that the consulting company is responsible for. In my experience if you can't write out your agile requirements in a contract, then you don't have a good enough internal understanding of your own agile processes and should probably work on that, too. Far too many people use "we're Agile" as an excuse for "we don't plan or document anything" and extend that to contracts, but that's absolutely not what good Agile actually is.

mlthoughts2018 6y ago

It’s right in the article naming Deloitte specifically: > “ "Deloitte also faced blowback in 2013 for an unemployment benefits system it built in Massachusetts. There, the benefits system arrived in 2013, two years late and $6 million over budget—only to be filled with glitches that erroneously cut workers’ benefits, according to the Boston Globe." Deloitte delivered a major project way behind schedule, over budget and full of defects.

moltar 6y ago

Ok but the problem is that customers don’t always know exactly what they want. Software isn’t like architecture. You can’t always predict every possible issue or corner case. And that’s where they get ya! They’ll win a lowest bid contract for the exact word for word spec. But then will bill wayyyy more for any change request. Which will be numerous obviously. That’s how actual project costs get inflated.

ilikeerp 6y ago
Which is totally reasonable. I can do X work for 1 million but if you throw Y into the mix then I obviously need to charge more. Want your oil changing? 100$ no worries - oh you didn't realise you also needed new tyres and we only found out when looking at them? That's extra.

lcrz 6y ago

> In several years I never once saw the firm not deliver what they agreed to in writing. I know people working with big consultancy companies and the recurring thing they complain about is a difference of opinion between client and provider about what exactly the terms of signed agreements mean.

harperlee 6y ago

> If you gave out a time and materials contract without clear acceptance criteria, testing requirements, and verification / support / warranty detailed, you probably shouldn’t be handling 100 million dollar budgets. This is the money quote. Big 4 companies (Deloitte, PwC, EY, KPMG) are historically accounting auditing companies. And their culture is heavily skewed towards billing their employees to clients per hour/day, or per unit of service, as auditors and lawyers have been doing for more than a century. It is not in their DNA to "create products" but to "render services"; that is, to sell employees availability. Boundaries and scope of those traditional services are typically clearer than building a new unknown program from scratch. If you are hiring these firms for exploration of how do you work, how you should work, and a complete functioning system that sustains a big number of new processes that are not yet defined, you are in for a big bill; you better be sure that you have put on the time to think about what you are buying in written form. When you see big failed projects like those the problem lies in that no one stopped to research the complexity of the ask before entering into a contract. Even then, these consultancies normally are very flexible when a contract is actively harming the client, and the client wishes to change it / stop it (it would not be good for the business to do otherwise, because you want to see more after this contract is finished). When these costs go out of control is because internal politics in the client impede the rational decisions of killing projects, changing command, etc., and the sponsor keeps feeding money into the project to save their ass. Source: worked in big 4 for years.

aplummer 6y ago
You couldn’t have written a more insightful comment.