BankerVision

The economic implications of power

I was in a meeting with Cisco yesterday where they talked of the work they are doing with respect to greening the network in the data centre. Basically, they said, you can increase the efficiency of a data centre by decreasing the number of ports that each server uses. With virtualisation on the rise, that is apparently a big concern. The newest switches apparently take 10 watts of power on each side of the cable to run. Add that up over thousands of ports and you can see the point.

Actually, this is an ever-so-common story. The availability of power to data centres is swiftly becoming the limiting factor in much of what we do these days. Remember those old glory days where we had to optimise our systems to use as little compute and storage as possible because there just wasn't much to go around?

We are now in the same boat again - but now there isn't enough power to go around.

Gazing into the future a little, what is the effect of not having enough power to go around? Yes, the price will go up. Getting more requires large scale and long term investments in infrastructure, so to all intents and purposes the supply is fixed.

Is it so hard to envisage a situation where large corporate users of power are expected to kick funds into the consumer market to stabilise prices due to their disproportionate use? Or get charged a special carbon tax to offset their power use? Or actually wind up having to build their own power generation? I think it isn't.

Now, if that or any scenario similar occurs, we are faced with a very dramatic increase in our IT expenses- without any increase in capability. Can you imagine the impact on our cost to income ratios, which have been falling steadily over the last decade? They'll suffer a swift reversal.

20 years ago, the biggest cost when you wanted to start an airline was buying the planes. What is it now? The fuel. The entire dynamic of the airline industry was changed by it.

How likely is it that the largest costs of a particular IT capability - over the life of the system - will be the power that drives the system? I think very. Such a development will have architectural and economic consequences for everyone that needs a lot of compute.

Banks? We're practically all computer.

In deregulated markets, it is already the case that you can buy power in advance. For a fixed price. Having a futures exchange for power is a natural consequence of scarcity, so I suppose those will become more common in the near future.

But power is an input in a long value chain that results in data centre compute. There are so many other things that have to be taken into account as well: the cost and availability of network links, the skills of people in the area, and the economic environment of the local region. These all result in various - and fluctuating - prices per MIP in the data centre.

Wouldn't it be natural, then, that we might see the emergence of MIPS markets? A futures exchange where one buys cycles in the cloud? If we take the airline analogy again, that is certainly what happens when carriers need to guarantee the price of their fuel for the future.

I think it won't be that long until we will need the ability to run our workloads on any cloud, anywhere. That will be because we'll get our MIPS for the best price, in advance, on a futures exchange. Of course there are technical and cultural issues with this, and I can already imagine the security folk meltdown when anyone dares to suggest putting customer data not just on someone else's data centre, but on a random facility dictated entirely by price.

But economics are a powerful motivator.

Where is that uniqueness in services?

The acquisition of EDS by HP is all over the news, and apparently will result in the second largest tech services firm in the world after IBM. Now, that's all very well, and we are a very big buyer of technology services, but size does not a compelling value proposition make.

I am challenged, in fact, to find anything about services companies that would make me want to choose one over the other.

Yesterday, for example, Christophe and I met with such a company. They had some interesting things to say, but ideas are rarely enough to get us going. We need execution.

Our key execution question was around their distinctive uniqueness: let's face it, if you are going to charge those big rates, there better be something different to justify it.

Technology is a commodity, not a uniqueness. And if we want business expertise, we go to a business consulting house, not a technology vendor.

Anyway, as far as I can tell, there wasn't anything very unique about this particular vendor, and neither will there be that much uniqueness in the new services behemoth that is HP and EDS. In fact, they will likely be distracted for at least two years while they sort out the integration.

The fact is, there are only so many ways to package up a bunch of people and try to sell them for a decent margin. Especially when what you are doing is helping us do business-as-usual.

Or is there?

There is a monumental difference between the outputs of a star performer and an average one. I think the number usually quoted is ten times more effective, and that's a huge upside when you're under pressure to deliver. The thing is, star performers are pretty hard to find. Consequently, they aren't likely to come cheap and are almost certainly not going to get commoditised. It would be amazing to find a services company that had a disproportionate share of star performers, when by an large, everyone is after the same people.

Everyone has access to the same labour market that services companies have, and so we know that arguments around "having the best people" are probably spurious. And if those people have chosen to work in a services company, they usually show up for the pitch, but are nowhere to be seen at execution time. Ten times more effective people are a precious commodity not to be wasted on anything so mundane as servicing customers.

So how about this for a unique proposition: have only star performers, and put up your rate card ten times to cover the fact that you have only one tenth of the headcount. Be so confident that your people can do miracles that you don't bother with risk-based pricing. Charge us nothing if you fail to meet any commitment at all, even if it isn't all your fault because your superstars can fix any problem. Make it a no-brainer to hire you because that is the only way that we can guarantee to our management that without fail we can deliver something.

I think we'd be glad to pay for uniqueness like that.

HP and EDS: where are the star performers in your now 200,000 plus average services employees? Do you even know? And how will you use them to give us uniqueness that will make us want to hire you?

How many innovators does it take to change a light bulb?

Answer 1: Three- one to ideate, one to innovate, and one to determine if we actually do have light.

Answer 2: None- the major advance in light was the discovery of fire and nothing new is going to happen in bulbs so we're moving onto transcendental luminescence.

Answer 3: One, but only once. This new bulb doesn't blow.

I've learned that if you ask the light bulb question and don't get all three answers, an innovation team doesn't have a sufficiently broad set of views.

Inevitably, that means you leave money on the table.

Doing consumer electronics

There is a revolution going on that - it seems to me -  banks haven't much noticed.  The revolution is in how low the barriers have become to create desirable consumer electronics. So low, that even if you don't know anything about what you're doing, you can do it.

I first stumbled on this last year when I was out at the Microsoft Research labs in Cambridge. They were doing lots of work that involved wearable sensors, and these things looked like you could just walk into a shop and buy them. I wanted to know where they got all their great looking hardware: "Oh, we just use an engineering firm down the road, it takes them a week to make anything we need". You would not believe the off-hand tone this researcher used, as if what I was asking was of so little consequence that the question was a waste of oxygen.

I was amazed, having imagined teams of people, just like the ones we have to have when we do software in the bank. So I subsequently investigated this claim and found it to be true. There are engineering companies that can turn around hardware for you in a week.

Actually, doing hardware is way less expensive than building software. It is in an interesting reversal of the otherwise universal maxim that bits are cheaper than atoms. Apparently there are all these standard "reference platforms" and they just throw them together in the form factor you need.

So we investigated some more, and discovered the whole industry has moved to an ODM - Original Device Manufacturer - model. You hand over the specification and design (that was likely done by some local engineering firm), and these contract production lines turn out the volume of product you need.

We wanted to know what it would take to miniaturise to an unprecedented level. Apparently you do something called custom silicon: they take your reference design and make dedicated chips or something. The price is higher than off the shelf, but nothing as high as you'd expect.

In other words, you can outsource everything about hardware. You can make sexy consumer electronics. And you can do it very, very cheaply.

Now, doesn't that pose some interesting options for banks, who are normally limited to utilising the devices that consumers have bought elsewhere?

The most famous example I can think of a of a bank doing hardware was 1950 when Bank of America collaborated with Stanford University to create the first business computer. That was a multiyear effort without any certainty of success. When you do hardware today, you don't have any of those costs or risks.

If I needed any reinforcement of this, it came when one of our favourite vendors responded to a query about biometrics with a custom device. I couldn't believe it. They didn't have exactly what we needed, so they made it for us - in about a week - and then furnished us with about a dozen of them for an internal conference we are doing today.

The fact that hardware can now be a sales tool implies that it has a very, very low price of entry. Low enough that banks could design their own consumer electronics to support the customer experience, if they wanted. You can imagine the situation: you choose your bank because its gadgets are the most desirable. Isn't that what happened with iPhone? People changed their phone company just to get one.

Innovation is about building new levers

Last night, I was at dinner with an innovator from another large company involved in financial services. We were discussing the optimum mix of incremental, disruptive and breakthrough innovation. In common with so many other people I've spoken to, it became apparent quickly that the feeling was that incremental was not necessarily the sort of things an innovation team should be working on. Big, bold disruption, that's what innovation teams should be doing!

But the real question comes down to one of risk: with a dollar to invest, do you take a gamble that might pay off hugely, or a more cautious approach where returns will be way more modest, but much more certain? My view is the younger the innovation programme, the less risky you can afford to be with the money you have. Get some runs on the board before you try for the big stuff. And, as I've said here for ages, make sure you can pay the bills. Incremental does that.

Anyway, the discussion swiftly moved to the regularly recurring old chestnut: is  incremental innovation or optimisation? In this case, my companion's thought was that incremental was very much business as usual, and therefore ought to be part of the work we all do every day. Optimisation, in other words. I disagree with this.

Optimisation is the process of moving various levers in a business process to get a better result. You decide where to set the levers by watching whatever measures you have on the process.

Innovation is quite different though, because the goal is almost never to move levers. The goal is to create new levers altogether, or to extend the range of movement of existing ones. The actual moving is best done by the business owners of whatever-it-is.

Conceptualised in this way, one no longer thinks about the mix of incremental, disruptive and breakthrough. The key strategic question is about the size and range of levers you decide to build.

A look back in time: prospects for home banking

You'll love this. I was going through past issues of the ABA Journal recently, and came across the October 1981 issue. It had an article entitled "Home banking prospects: A status report on explosive growth" (you can only get it now through various pay wall databases, I believe). Bare in mind the time scales. Web based online banking didn't really emerge until 1995 when Presidential Savings Bank was the first in the world to launch it. Videotex based systems had never really caught on, though they were available from 1979. And phone banking had been in limited pilots since then too, with broader adoption not occurring until 1982.

Given that, its surprising just how accurate some of these predictions are. Take this for example:

"Banking services are only part of a cluster of home information offerings that will set the pace for growth (of self service). Theoretically, an unlimited number of information providers could offer data bases of various kinds. Banks will be one class of information provider."

The article goes on to talk of a "communications switch" which makes the necessary connections between the user and database vendor. A single vendor switch was not what happened of course, as the internet evolved into a universal switch not owned by anyone in particular. But the concept, then, was clearly sound.

The slow follower syndrome is something we've been aware of in banking for ages now. Institutions don't like to be the first to try the new stuff. Neither do we always understand up front why the new stuff is important. At this moment, we're struggling with personal finance sites and peer to peer lending, but in 1981, bankers were having trouble getting the importance of self-service as a delivery channel:

"Many bankers not involved in any of the current tests (of home banking) may find it hard to get excited about providing new banking services that seem to have marginal value for customers just getting used to automated teller machines."

The biggest problems were technical in 1981. And the question was whether households would prefer telephone based terminals (i.e., a dedicated device embedded in the telephone) or screen based home banking using interactive cable. Screen based self-service were seen as preferable, but already institutions were leaning towards the phone, because of a projection that only 9% of households would have interactive cable by 1990. Of course, Internet was in its early stages for commerce then, and was about to zoom upwards in adoption. Nonetheless, everyone was pretty much set on the idea of terminal based access in the long term. Bank One, the first institution to pioneer phone based self service, stated that it expected that 60% of its customers would have a stand alone videotex terminal costing about $250, 20% would use decoders attached to their TV sets, and the remainder would use home computers.

But one of the most interesting things about this historical backtrack is that most of the nine institutions listed expected to be able to charge for their home banking services. Prices were based on a subscription model, and were up to $30 a month. It is obvious in retrospect, I suppose, that convenience doesn't pay. But you can see the evidence of dollar signs in the eyes of everyone talking about the offering then. Chase, for example, wanted to offer its correspondent banks a franchise opportunity. They would buy devices for about $200 each, and then sell them on to customers, with Chase getting about $6 per month from the deal for each.

And the article is insightful in one more respect: in 1981 it correctly foretells the features arms race that we've all been engaging in since self service became the business:

"any (home banking) convenience, such as home shopping or a sophisticated cash management service, could be the one that pulls an account away from your bank, if you don't offer it, to mine if I do"

So here is my key takeaway from this brief look back. We - banks - aren't as bad as it might seem at predicting the future. From a market and environment perspective, most of this was spot on. Admittedly the technology selections were going down a dead end, but they were overtaken by developments that weren't on the radar then.

It gives me a level of comfort that the course we are presently charting is likely to work out well in the end.

The new little server

I am presently in Canada, attending the latest TTI Vanguard meeting on technology trends. We're exploring how pervasive communications technologies will change art, influence science, and innovate business.

Yesterday, there was a particularly epochal presentation. We're running under Chatham House Rules here, so I won't disclose who was presenting, but the basic point was that the mobile phone will become a personal server in the immediate future.

We are all familiar with the phone as communications device, and are becoming more familiar with it as file storage and music repository. But at the moment, the phone is subordinate to our other devices, such as laptops. The short step in front of us is the reversal of that arrangement.

That's going to change the value of my phone compared to my wallet. let's face it, apart from ID and cash, the wallet is really just a place for a bank to live. IDs are just data anyway, and cash is in decline. NFC is just around the corner.

It is likely that when I have a little personal server in my pocket, I am going to ditch my wallet.

What evidence is there that personal servers are about to happen?

Firstly, while the phone itself doesn't give you an especially good experience (mainly because it has a small screen and microscopic keyboard), the environments around us are increasingly populated with screens and keyboards. They are pervasive.  All that's necessary is that the phone can reach out and use these devices to provide a decent user experience. Conveniently, most phones have Bluetooth now, so the hardware exists. Singapore airlines now provide screens and keyboards in seat backs on some flights.

Secondly, it is possible to run server processes on phones without actually degrading the performance of the voice function. During the presentation, a linux handset -bought over the counter - was used live. It was running a media server, a file server, a web server,  and a frame buffer server all on unmodified hardware.  The latter means that the phone can write to a separate screen over a bluetooth connection.

Thirdly, both the processor and the battery life of current generation devices are capable of doing interesting things.  Here's an interesting statistic: a typical phone can stream 4 hours of media to an associated screen at high frame rates. Intel is planning to put ultra low power PC class processors in mobile phone devices any time now. So you can then run Office and all your other apps that you presently rely on.

Finally, device storage is almost at a magic number: 10gb per square inch. That's the number, apparently, at which the amount of storage becomes interesting for most human applications (media, for example). Incidentally, recording audio for the entire life of one individual would take 3 terabytes of storage. Technology is not so far from that density in a mobile phone form factor.

So the world view being advanced here is that your digital life is your handset, which will make use of other devices it finds in its environment to provide a desktop class experience. Everything subordinate to the handset.

I think that most people would agree with me when I say that - today - if I had to choose between losing my wallet and my phone, I will choose to lose the phone every time. The level of inconvenience and risk involved in losing my wallet is considerable. I will be without my cards, without cash, and without ID.  My life, to a large degree, stops.

But when my phone becomes a server with all my data, that immediately reverses. I would much rather lose my wallet in that case. The wallet, despite the inconvenience, is replaceable. But the data on the phone doesn't represent just the financial slice of my life, it is everything about my life. With a built in data manipulation engine just waiting to be used.

Commentators, such as the prolific David Birch, have been discussing the migration of the wallet to the handset for ages. The fact of the matter, however, is that a personal server offers synergies rather more significant than the move of payment functionality from plastic. It co-locates bank functionality with the rest of my digital life.

Now I won't pretend to have thought through all the implications of that, but it is a very, very significant shift. It will change the way that banks activate their customers. And it will certainly change the way that customers choose to interact with banks.

Innovating work life balance

As at most companies, we have a focus on work-life balance. The other day, for example, I got to watch an interview with one of our most senior people at which the balance question was asked. The answer, and in fact, it seems to be a universal one whenever you get this question, is that work and life are kept separate. The decision of importance is the one which apportions time between work and life correctly given the personal situation at hand.

This answer to what is a very tricky HR question actually concerns itself with management of the balance between home and work which is a pretty critical part of stability in life. But it doesn't concern itself at all with investment, which is the other side of the coin here.

What do I mean by investment in this context? Investment is the set of compromises you have to make in order to expand your responsibilities successfully. That set of things you hate doing when you'd rather be at home. Or the onerous tasks you've left till the weekend because there simply wasn't enough time in the week. These are investments you make in your work.

You make them expecting there to be a payoff.

It is unreasonable for those who choose not to make those investments to get the same payoff.

The fact of the matter is that no matter what corporate policy you have surrounding work/life balance, there will inevitably be many people who choose to invest more substantially than you. Implication: they produce more, and therefore get greater opportunities. It is pointless to rail against this.

When you are faced with an unequal playing field - perhaps because you have children at home or a new relationship to support - trying to match work product for work product is not a winning strategy.

Some people can make few investments and still reap substantial rewards. What is their secret? I was watching such a person the other day to try and see if I could crack the code, and I think I have.

They are innovators to the core. Their first response to anything that comes up is to evaluate how they can spend their (limited) time doing whatever-it-is differently to everyone else. They don't waste time with a tried and true response - a brute force approach requiring a full working week and then some - they seek to optimise.

The new thing they invented is not always better than the old thing they could have done. That is irrelevant - the change gets noticed. And on those occasions where there is an improvement, the innovator tends to get the credit immediately.

That's in contrast to the standardised approach, where you are competing on a playing field that is not level. It's a strategy which rewards those who have the most time to spend on work.

When you treat every situation as a new one, thereby requiring an innovative response, you have a system that self-optimises. People who practice that habit just seem to be more successful, no matter how hard they work. And, of course, it makes their teams more successful too.

The democratisation of complexity

Everywhere you look, very complicated things are being done by people without very much training in how to do them. Have you noticed? Finance people are building super-complicated computer models without any formal training in computer science. They can do that because the tools of production of complicated models have been democratised in the form of Excel.

Elsewhere in the business, work-flows with complicated approvals and integration with other systems are being built by workgroups. Perhaps they don't call what they're doing workflow, but they are doing it nonetheless. Here too, the tools of production of complicated business processes have been democratised.

Actually, it is happening everywhere, and not just in technology. The other day, I cam across a document by an intern that featured super complicated monte-carlo analysis as part of a forecast. Not that this intern was especially statistically trained. It was just that the tool was easily accessible, so it was used. Complicated statistics democratised.

Somehow, all this democratisation of complexity has happened under our noses without us really noticing. But there is a significant take-away I get from this. What people want is not a specifically engineered function that solves a specific problem.

They want generalised tools they can use to solve problems for themselves.

They're bored of doing requirements and testing, and the cost load-up that comes from formal projects. They can solve an increasing percentage of the problems they have themselves. And the pace of things has moved along so much that now, there are sometimes few choices but to go it alone to meet market demands.

I think that we're getting close to the time when what we build to run our financial institutions will not be monolithic feature sets, but platforms that our people combine together to do new and interesting things. We will have little understanding in advance of what those things might be.

If anyone doubts that end-users are capable of this, one has to look only at the rise of FaceBook in this area. When they launched their platform a year ago, there were hardly any applications. But now there are thousands, and all built by users. Same story with SalesForce.com, but in a corporate context.

I'd imagine there are two steps in the ultimate journey for institutions who want to take advantage of all this democratisation of complexity.

The first is to realise that in the relatively short term, systems that can be customised by their owners have significant competitive advantages compared to those that can't. Parameterised core systems are a key example. When you can roll out a new product on the say-so of a product manager without a huge IT fuss, you are going to be in a much better position than otherwise. Parameterised systems are good, of course, but really amount to handing over the levers of a system to its owner. The real key is to let system owners themselves make new levers. Doing so makes it possible to harness the democratisation of complexity and drives an innovation agenda.

But the second step is more radical by far. Let customers make levers too. Let them mash their banking services into new and unexpected things. Do it from within a walled garden so that the key service provided by the bank is safety and trust.

Then, what you have, is an institution that is successful not only because it has the best people, but because it has the most engaged customers. Customers that choose to co-create the bank. The sort of customer you'd definitely want to have working with you, and not against you.

A new kind of event please

I have an appeal for vendors everywhere: create events that are unique and interesting.

In our line of work, we get many invitations, and we go to some of them. But there's a definite trend, I think, to homogeneity of all vendor events around several issues only.

For example, not so long ago, practically every event that hit my in-box was about multichannel integration. "That old chestnut", I used to think every time someone told me about an event concerning on this topic.

Before that, it was the Internet channel and how to make billions with cross-selling.

Presently, the hot events we're getting are all about Green IT, New Payments, and (not surprisingly) innovation in all its forms. Now really, there are only so many ways we can be told to save power, or consider contactless, or reconsider our innovation process.

And there are only so many formats these events can take.

Firstly, there is the discussion dinner (not usually a lunch in the UK). Depending on the vendor, in some case a rather lovely restaurant (or for more exclusive clientele, a private dining room), a moderator is brought in, and the people round the table talk about the issue in question.

There is the presentation plus drinks format, usually with a much bigger crowd (invariably clutching their "VIP" invitations). At which there will be one or two speakers, (one will be a crowd-drawing speaker with interesting content, the other will be a vendor with a sales pitch).

And there is the all day, possibly including lunch, event, which features many presenters, possibly some stands with interesting gadgets, and a drinks option at the end.

I've pretty much stopped going now. We are practically never told anything new, and there are very few vendors that can make events work without the obvious sales pitch. So here is my plea: please tell us something that can justify our investment in time and that will help us do our jobs. Challenge us and change our thinking.

For example, I'd like an event that would tell me how to run a bank in the cloud. Now I realise that its not something that can reasonably be done right now, but it is vendors who will work out the major problems. And, for extra marks, tell us how to establish a MIPS market so that we can buy our compute in advance like airlines do with fuel. Share your thinking with us!

I'd like a new format too: what about a debate between you and your top competitor (with, of course, your best thought leaders) on the topic? Let us have some crowd-pleasing disagreement with the cleverly clothed sales pitch.

With so many invitations, I'd expect that the vendor that cracks the code of this will be pretty much inundated with acceptances. In the end, the difference is the same as that between a great television show and the ads that pay for it. We'll watch the show, but timeshift the ads.