Lufthansa tosses IT biz to Big Blue, inks outsource deal with IBM Lufthansa has announced that it’s about to sell off its IT infrastructure division to IBM and rent IT services from Big Blue in an outsourcing partnership. The German airline has been restructuring itself in an effort to cut costs and said earlier this year that it was looking for a buyer for its IT biz so it could better …
A pity that Lufthansa didn't do a bit more research before believing that an outsourcer is going to do the job better or cheaper than they will themselves. WIth around 115,000 employees I don't buy the argument that Lufthansa lack the scale.
A more pressing concern is that IBM made (4Q13) made gross margins of 40% on technology outsourcing. Let's assume Lufthansa have been made promises about a 20% cost saving (that's the magic number most outsourcers seem to fixate on). If Lufthansa are going to save that money, and IBM are to avoid diluting GTS division results, then IBM need to be able to operate what will be fairly heavily dedicated IT infrastructure for an operating cost fully 60% lower than Lufthansa are currently incurring. Can't see that myself.
Mind you, my employer's board swallowed similar promises, and it did result in IT costs going down. But only because all routine IT costs went up after we outsourced (along with a reciprocal decline in service standards), and desperate IT managers reacted by slashing operationally-needed projects and new systems from the IT investment budget, and called the resultant net effect a "saving". That's not the view of the operating units, but who asked them?
Oh, and another word of advice for the directors of Lufthansa: You know that long, long, detailed SLA you signed, that commits the vendor to deliver all the promises they made? You might as well wipe your German backsides on that. Outsourcers have large, heavily resourced teams who work all day, every day on writing these agreements, and have big bonuses tied to a successful outcome (which includes the vendor not having to pay up when said vendor messes up). You, on the other hand, had a small and under-pressure team of procurement generalists who write agreements like these once every few years at most, accompanied by a team of IT people already under the cosh for being "too expensive", and most of whom actually had an operational role.
Maybe I'm being unfair. But it seems that either Lufthansa's people didn't do their research properly in the matter of IBM versus Cable & Wireless, or alternatively they chose to believe the salesmen's assurance that the leopard had changed its spots. And salesmen, they always tell the truth.
It's even worse than that. A contract* can only ever be effective if you assume that you know precisely what you will need in 2,3,5 years time. In which time the world will have changed, the business will have changed and the technology will have changed at least twice. If you have your own staff, you can simply say: hey guys, stop doing X and start doing Y. An outsourcer will say: unfortunately the changes you request are not covered by our agreement and will therefore be chargeable at our outrageous daily rate.
Even if you believe you can nail this down, I've seen a large global outsourcer threaten a large global company with simply walking away because they weren't making any (enough) money. Maybe after a few years of expensive lawyers you could win significant damages, but (starting tomorrow) you have no computer hardware, software or staff - how long are you going to stay in business? Top management caved in and signed a new contract at substantially higher rates.
* Contract: a legal document enforceable by the stronger party upon the weaker party.
I don't think you are being unfair, though I doubt the margin numbers are right upfront (I suspect outsourcers make most of their money in the later years of contracts when IT costs have gone down but clients are too tied in to move. A year one saving of anything big would be a bit scary).
Even assuming you're halfway right though, there are only a couple of angles I can think of that make this work:
- Their IT staff are mostly German and impossible to directly get rid of, whereas IBM are planning to move the work elsewhere. IBM may be better placed to do that than Lufthansa, but it certainly won't be popular and carries some risk of service disruption (depending on your opinion of offshoring or whatever the latest name for it is).
- They are really planning to get rid of a bunch of projects as you hypothesise they'll need to do, but German law won't allow them to get rid of the people, so they're outsourcing the downsizing to IBM (a minor variation on point 1 above, and I've no idea if it's legal, but it probably is).
- IBM have enough reusable systems and processes that they can genuinely do things sufficiently more efficiently to make the desired profit (which seems unlikely at the scale Lufthansa must operate).
I suspect one or both of the first two apply given that German law is pretty rigid and IBM have huge operations in India and elsewhere that can be leveraged.
though I doubt the margin numbers are right upfront (I suspect outsourcers make most of their money in the later years of contracts when IT costs have gone down but clients are too tied in to move. A year one saving of anything big would be a bit scary).
Agreed that the normal BPO/ITO model is based on backloading the returns to the outsourcer, but the way this translates is that the vendor usually promises year 1 savings, and takes a loss in years one and two of the contract. It's a rotten business model, but it works a treat because corporate managers fall for it time after time.
Their IT staff are mostly German and impossible to directly get rid of, whereas IBM are planning to move the work elsewhere.
This is usually a big part of the deal, with the salesmen promising vast, risk free savings from outsourcing and offshoring. The redundancy payments are I guess the big cost Lufthansa quote for the setup costs. Unfortunately, when you take into account the lower productivity of most cheap labour, high turnover in key outsource destinations, and the big fat vendor margin, those savings evaporate.
They are really planning to get rid of a bunch of projects as you hypothesise they'll need to do, but German law won't allow them to get rid of the people,
This is a common fallacy outside Germany, that you can't get rid of people. But its not true. A German company I worked for recently got rid of around 10,000 German employees, and the UK based company I worked for before that rationalised its German business from 3,000+ down to 1,600. It does cost a bit more, and it takes a bit longer than in the UK, but German companies can certainly get rid of people if they want to. The German psyche that says that companies should look after employees is rather stronger protection than German law these days.
IBM have enough reusable systems and processes that they can genuinely do things sufficiently more efficiently to make the desired profit (which seems unlikely at the scale Lufthansa must operate).
That's a favourite from the management consultants who write gushing papers about the benefits of outsourcing. But it makes a typical consultant error, in that it assumes that cost savings for the vendor (from re-use of existing systems or code) translate to savings for the buyer. IT departments of all people should understand that the whole point of code, IP and processes is to reduce the cost for the owner, whilst still charging the user as much as they can afford to pay.
In my experience of looking at such matters, the outsourcers allude to such efficiencies and "best practice" all through the pitch, but don't actually have any such systems or code up their sleeve. The business model works best when the customer has broken or inefficient processes, because the game is a margin one. When you make 40% GM, why on earth would you want to make things run at lower cost for your customer by offering them a brilliant, lean system that reduces cost by (say) 35%, and thus reduces vendor profits by a similar figure? In reality the outourcers want to take on whatever shitty processes the customer has, and then charge the customer through the nose to make a few basic changes that don't materially reduce the operating costs.
I'm not sure this is quite the same thing. Amadeus and Sabre are (at heart) multi-airline reservation systems (I think Sabre started life as the American Airlines booking system). It makes sense for airlines to use them for this and related purposes rather than develop their own versions in house and then have to interface with them. While Amadeus and Sabre themselves are fundamentally software houses, and could reasonably choose to outsource the running of their 24x7 systems to a larger global player like HP.
Outsourcing large parts of your global IT systems, as Lufthansa would appear to be doing, is a much bigger step (braver, in the 'Yes Minister' sense).
Quite. I wouldn't be at all surprised if within 3 months servers start to appear in the office as a 'temporary' workaround to the costs and timelines IBM quote for performing even basic new tasks.
Outsourcing specific functions is sensible and normal. For a company to shift its entire IT outfit which is purely a cost base to a firm which considers it a substantial revenue source is self-evidently going to end badly.
Sabre has been outsourced to HP for 24x7 running for some years now.
I note that Lufthansa are keeping airline hosting to themselves? Lufthansa are a system user on Amadeus, but have their own reservations system behind this that also hosts a number of airlines.
I'm unsure what LH are actually outsourcing? The split into three with two parts retained implies it is 24x7 and desktop that is being outsourced?
That would be right and shift all the pain of dealing with the German works councils and the hoops to IBM, they are in for a world of pain there. When they discover how hard it is to make them (the Germans) redundant I'm sure the Brits and the US will suffer culls instead as they are both easier to fire.
A previous poster mentioned that most airlines already outsource, and that's true. The problem is that, on paper, the costs are lower. The reality is that the vendor will proceed to nickel-and-dime the airline on a time and materials basis for everything that isn't literally written into the contract. The service quality drops because the vendor replaces the knowledgeable in-house talent with the cheapest they can find. I worked for an airline in the US that outsourced its data center and network to a large provider, and was retained on the in-house IT team. Everything routine that we used to do became an absolute nightmare because of finger pointing and process/paperwork. We got to the point where we were figuring out how _not_ to trigger changes in the environment to avoid the massive headaches and late night conference calls. And since this is IBM...they have a lock on the patent for how to provide the least amount of service for the most amount of money.
I seriously wonder if someday, far in the future, MBA school case studies will look back on this era (80s to now basically) as the outsourcing era and teach it as a "how not to do things" case. It really doesn't make any sense -- how could anyone possibly do a task cheaper than your in-house, "free with purchase" resources?
And I hope you're not. Lufthansa texted me at 1.30am on Monday to tell me my flight was cancelled (due to the strike, which they didn't say), but MAYBE they could get me on an earlier flight. To even possibly get home I had to forgo all sleep, book a replacement train ticket at my own expense and then hope.
Fuck Lufthansa. Let them burn.
The company I used to work for collected data, sliced and diced it, and sold it to multiple customers. The primary source of data was a government agency, and one of the major customers was the same agency. But we could collect the data, process it, clean it up and provide the structured and useful data much cheaper than the agency could have done themselves. The major reason was we had 1/2 dozen other customers for the same data, so we could spread the cost among all of them.
Similarly, perhaps IBM will be able to use and adapt the existing Lufthansa infrastructure, and sell the equivalent service to other airlines. This synergy will reduce the delivery cost per customer. The more airlines that sign up, the better for both IBM and those airlines - assuming they run it properly!
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
