Hello? Ja, bitte? Si, Pronto? Allo?

A lovely quote from Chuck Prince, CEO at Citigroup, in one of the many FT’s I leafed through on the to Paris this weekend. Talking about squeezing more money out of the bank’s (or should that be “Financial Super Group”?) cost base, he said

We have lots of great technology and none of it talks to each other

I worked for Citi for 7 years from 1992 to 1999, helping formulate the 5 year vision covering 1995 to 2000. One of the things advocated was a consolidation of legal vehicles – first in the FX business, then derivatives and, in parallel, across European Consumer and Corporate banks. The plan was that this would allow Citi to reduce the number of general ledgers from 18 to 1, stop dealers betting against each other in currencies or interest rate products and allow the consumer bank to be serviced from a single platform. When I left, the plan was well underway – certainly the FX piece was done and the consumer bank part was just kicking off.

Chuck is plainly now looking at the next step which, I imagine, is the thankless task of integrating or consolidating all of the other systems – securities settlement, credit cards, dealing accounts, private bank etc. The year 2000 programme cost Citibank something like $600 million, the preparation for the Euro perhaps $400 million. Big numbers for sure – representing the spend to keep a huge technology base in check.

Citi’s latest quarterly revenue (to October 2005) came in at $21.5 billion. Corporates spend about 2-4% of their revenue on IT, so Citi’s annual overall spend on technology could be somewhere between $1.6 billion and $3.5 billion. That’s a lot of spend to harness (HMRC, as a comparison, spends about £300 million to maybe £400 million, based on the figures when the CapGemini contract was let). It needs a pile of investment to integrate though – which means that spend has to be redirected away from, say, replacement/replatforming, new products or maintenance, and towards making things talk to each other. That is both good and bad – it forces a focus on the business case of integration measured against the business case for a silo product launch – but, let’s face it, the integration case has not often worked in the past as it brings complexity, dependency and higher risk of failure.

Of course, we know that most corporates have too much technology … here’s something that I wrote around this time of year 3 years ago. And something a bit similar from mid-2003 on the lack of channel integration.

The parallel with e-government or government generally is that the recent Transformation strategy talked about redirecting some 10% of the annual spend on technology (in the UK’s case around £14 billion, or $22 billion – perhaps 10x Citibank’s) towards shared services and greater integration. It’s an interesting strategy:

– The money has to be freed up ahead of time (you have to know you’re going to have it to spend) so that you can launch procurements and sign contracts

– You have to put joint governance over it (every department involved has a stake and will want a say)

– Someone needs to be in charge (accountability needs to be clear and the head of the department running the specific project will be the one that stands in front of the PAC to defend or praise the project later)

Contrast that with Citi where the CEO can put the right executives in place to make this happen, can direct spend where it needs to go to create most impact and can drive co-operation across the board through rewards, punishments, death and glory.

I spoke at a conference about 3 weeks ago, in San Francisco, to the top 250 managers of a major tech company. They’d asked me to give them a customer view on how they were doing. My major theme was that systems by themselves rust and that too few vendors remember what the customer was actually going to do with the purchase – service card holders, let people make calls, manage leakage in water systems, sell books etc. Instead, for the last 10 years or more, it had been a game of how much hardware, software and maintenance could be sold. Deals got better every quarter or year end, encouraging customers to wait for the right moment (if they could) and to see the products as commodities rather than as fundamental components of the business. Future growth would come less from the commoditisation of products, but from innovation and delivery that helped the real customer get what they wanted, in turn choosing the provider of that product who, in turn, would choose the partners that they needed to deliver.