Sunday 30 November 2008

BLOG TERMINATED (please go to

Hi Everyone

I have made the decision that maintaining two blogs was bit ambitious, so have decided to close this one down and focus on the Open Reasoning instead, so please go there for ongoing posts.

You may also be interested in my more formal output in the form of reports and media articles, which can be found here.


Wednesday 20 August 2008

iPhone: First impressions of a Blackberry user

For a while now, I have maintained two mobile phones – one for business use and one for personal use - and in terms of requirements, I need different things in each context.

There are four ‘must haves’ for my business device – good battery life, quick and convenient calling from a large directory of contacts, a solid, immediate and user friendly email capability, and acceptable security.

As a long time Blackberry user, three of these have been ‘givens’ for the past five years, the only compromise in the early days being relatively clunky functionality for making and receiving calls. The Blackberry Curve I am using at the moment, though, delivers well on this front too, so all of my needs for business use are catered for effectively.

On the personal side of the equation, my requirements are a bit different. From a calling perspective, I tend to be dialling from a much shorter list of contacts – tens rather than hundreds – and telephony use in general is much lighter. As a small business owner, there is still a requirement to access business email (as you never know what might need your direct attention), so connectivity to Microsoft Exchange and security are still important. Immediacy and user friendliness of email functionality is less so, however – these just need to be good enough allow periodic inbox browsing and very occasional replies.

Battery life is an interesting one. When using a device off duty, if I see the juice is getting a little low, I can curtail my usage and prolong the life left in the device. This is generally not an option for business use given the communication intensive nature of the job I do.

Beyond the communication stuff, there is also the recreational side of things – music, games, photography and perhaps a little web browsing. This brings me to the iPhone, and when I was looking for an upgrade to my personal device a few weeks ago before going on holiday, I felt obliged to check out this option.

Like most people who pick up an iPhone for the first time, it immediately felt quite natural, and it is the first device I have used that appeared to deliver a genuinely usable full web browsing experience – at least when connected to WiFi in the O2 store. When digging a little deeper, the Microsoft Exchange access seemed pretty well covered, the device was pin-securable with remote wipe capability, and the embedded iTunes, GPS enabled mapping, etc looked great. The only thing that seemed a little naff was the camera spec, though I figured it was probably good enough for snapshots of the kids, dog, etc.

So, I succumbed, and signed up for an iPhone on the basis that it seemed to do most of the things I wanted. But has it lived up to expectations?

Well three weeks on, I have to say that I still really like the iPhone and am pleased that I went for it. It has stood the test of real life use and quite a bit of experimentation over my recent 2-week holiday. It functions OK as a phone and call quality seems pretty good. The music playback quality is also good, especially when compared to my iPod Nano. Beyond this, there’s a reasonable number of games available to keep me amused, and, as suspected, the camera is actually OK for family snapshots, though, unsurprisingly, no good for ‘proper’ creative photography.

However, the iPhone is far from a perfect device. The most immediate problem I ran into was battery life. Perhaps optimistically, I started out running with all of the defaults – relatively bright screen setting, 3G enabled, GPS switched on, email delivered from our Exchange server through the ‘push’ mechanism (similar to Blackberry), etc. After returning home a couple of times at the end of a day out with the battery almost exhausted (with relatively light use), I suspected a little tuning was in order.

Fortunately, the iPhone allows you to switch off 3G access with the flip of a soft-switch, leaving the device running purely on the GSM network with data access over GPRS or EDGE. This improves battery life considerably, and as I don’t do much browsing when out and about, I have left it this way, figuring I can always enable 3G again for short periods when I really need to. The other adjustment that seemed to extend battery life significantly (apart from the obvious move of winding down the screen brightness) was disabling the push email mechanism and setting the device to poll the Exchange server every hour instead. Again, this adjustment can be made through soft-switch flicking, allowing the polling frequency to be set to every 15 minutes, 30 minutes or whatever, with more frequent polling clearly consuming more power.

Interestingly, enabling and disabling WiFi doesn’t appear to make a huge difference, so because of the convenience of the iPhone automatically hooking onto my home network when I arrive at the house, and discovering hotspots when out and about, I have tended to leave WiFi switched on.

As a disclaimer, I have to say that my tests have not been that scientific in that I have just been making adjustments in an attempt to get a configuration that works for me while I get on with my life. Unlike a lab test, no two days usage have been exactly the same, so what I am picking up here are gross differences in performance. That said, the one conclusion I have come to is that the combination of battery life limitations coupled with the inability to swap batteries when the power runs out makes the iPhone far from ideal for heavy business use on the basis of the power issue alone.

And while I wasn’t explicitly evaluating the device for business use, there are some other things that would cause me concern in this context. Apart from the widely reported lack of cut-and-paste capability, I noticed some quirks associated with the email client, for example, which make it difficult to do some things offline and require zooming and horizontal panning to read some email messages that refuse to word-wrap. Then, while I was pleasantly surprised at how usable the soft touch-screen keyboard was for casual text entry, I cannot imagine ever getting to the level of speed and unconscious use that comes naturally with a device that has a decent physical QWERTY keyboard. This may not be a concern for many, but it is major consideration for me, as I tend to use mobile email very interactively for business purposes.

In terms of issues from a corporate adoption perspective, others may also be concerned about lack of data encryption on the device itself, but with a sealed unit, pin access and remote wipe capability, if you take a common sense approach to assessing risk, there is probably not a huge security exposure for most business users.

When all things are considered, I would say the iPhone comes nowhere near devices such as the Blackberry Curve or 8800 series in terms of business fitness for purpose, particularly for heavy mobile data users. As a predominantly personal device, however, it is a great example of where mobile technology is going, and as I said, I am very pleased with the overall package.

As an industry analyst, I should probably grumble at the closed business practices of Apple itself in terms of controlling the distribution of content for the iPhone, but when I then think about the convenience and ease of use for a non-technical user, I can see that there is a also an upside to controlling things end to end for mass market consumer adoption.

So, the bottom line is that based on my initial impressions, I would not discourage anyone from buying an iPhone for personal use, but I would urge them to think about their requirements and do the appropriate due diligence before investing in the device for business use. As for large-scale deployment in a business environment for hard-core mobile requirements, I am not sure the device is yet ready in its current form, though if anyone has any experience to the contrary, I would love to hear from them. How do you rate the iPhone from a policy management, software distribution, maintenance and end-user support perspective for example?

Whatever the current situation, the end-user appeal of the iPhone will ensure that it makes its way into many businesses one way or another, and with Nokia, Microsoft, Palm and others already challenging RIM on fitness for purpose, we can look forward to an interesting couple of years as it all shakes out.

Tuesday 15 July 2008

Desktop power management

It’s encouraging that many of the conversations we are having at the moment in relation to IT and sustainability are moving beyond power management in the data centre. It is not that optimising the use of central IT isn’t important, but it really is only one way to drive an organisation’s environmental agenda. And even before we get to main question of how technology can enable more eco-friendly working practices, there is another place we can look to for operational IT power savings – the desktop.

When looking in this direction, though, I have noticed that there is a tendency to apply the same kind of thinking that is used on the server side of the equation. Fair enough, accelerating hardware refresh to introduce more power efficient kit into the equation reflects a similar game to that being played in the data centre, but with the carbon cost of manufacture/disposal taken into account, the net gains are hard to establish. In the data centre of course, hardware modernisation is augmented by consolidation and virtualisation to drive up average server utilisation and thus improve energy efficiency.

Virtualisation is a different game on the desktop, however. Sure, some will go down the route of running virtual PCs on the server and accessing them through thin client configurations, but it will be a long time before this is the norm. The reality is that most organisations will remain wed to their fat clients for the foreseeable future, so we need to think of the energy question a bit differently. Essentially, the challenge boils down to optimising the power consumption of desktop machines that typically idle for the majority of time they are switched on.

In order to deal with this problem, we need to think less about utilisation and inherent power efficiency of hardware and software, and more about controlling the state of machines in terms of their sleep/wake cycle. In practice, a configuration exhibiting a high degree of runtime energy efficiency, but has no active policy to transition to a low power state when idle will consume considerably more power than a less efficient machine whose state is properly managed.

This something that Microsoft makes a big point of when talking about Vista in the green context, and indeed early adopters with large Vista estates corroborate Microsoft’s claims that Vista’s enhanced manageability translates directly to power savings. The problem is, however, that Windows XP isn’t going away in a hurry, so what about all of those organisations who are interested in desktop power management but will be maintaining older versions of the operating system for some time to come?

Well the one approach that is generally acknowledged not to work that well is to educate, encourage or threaten users in an attempt to get them to keep their power configuration set in accordance with environmental policy, and/or to manually shut down their PCs or put them to sleep when they are not in use. IT managers relying on this kind of user discipline are probably not going to see the results they were hoping for unless they’re working for a totally green-tinted organisation.

Fortunately, third party solutions exist that can help to enable/enforce centralised power management – a couple of examples being Verdiem and 1E. Using such technology, you can not only cure PC insomnia from a policy enforcement perspective, but also allow real-time remote control of power state so machines can be woken up for backup or software distribution purposes then put to sleep again afterwards. So, if you are serious about saving energy across a large XP estate, the options are there.

Something I haven’t had time to look into is whether similar solutions exist for alternative desktops – namely Mac OS X and Linux. Apple kit is certainly not renowned for its enterprise management friendliness, but perhaps ‘right on’ Mac users aren’t so much of a problem as they are of course more environmentally aware. As for Linux, I would be interested in any views, recommendations or experiences.

Meanwhile, it would be great to see a bit more awareness raising from Microsoft on the availability of solutions to centrally manage power consumption by Windows XP, rather than automatically seguéing from this discussion into a Vista upgrade pitch.

Friday 27 June 2008

Justifying a large scale Vista migration

Over the past couple of months, I have had in-depth conversations with five CIOs that have made a significant commitment to Windows Vista.

One of the main issues I explored with each of them was the foundation upon which the business case for migration was made. The responses I received were remarkably consistent, and not completely in tune with the way Microsoft articulates the Vista proposition.

What all these guys said was that their business case for Vista, i.e. the one put before the board, CFO and/or other significant stakeholders, was founded on benefits in two key areas - security risk management and operational cost control.

From a security perspective, the focus tended to be on three specific attributes of Vista - better run-time security in the operating system itself, more effective policy enforcement, and the ability to encrypt data on notebook PCs through BitLocker.

What I found interesting was the view that while all three of these security related benefits were considered to be significant, it was the last one in particular that was most frequently highlighted as resonating directly with business stakeholders. Recent high profile press coverage about notebooks storing sensitive data being lost or stolen was seen to have an influence here in terms of awareness. Against this background, Vista’s ability to deal with an acknowledged business risk straight out-of-the-box was perceived to be of significant value.

Beyond security, double-digit reductions in operational cost generally formed the substance of the business case in financial terms. The general streamlining of the management and maintenance process was highlighted as part of this, and the dramatic simplification of image management in particular was seen as a significant contributor to the savings in the large multi-national environment.

Something I was personally very sceptical about, but which three of the five CIOs defended very strongly, were the savings in relation to desktop power consumption. Numbers from 50 Euros per year per desktop upwards were cited as savings, though to be absolutely clear, the benefit comes from better centralised control and enforcement of power management policies rather than efficiencies in the way Vista uses hardware resources.

When asked about the element that was clearly missing from these business cases, namely improved user productivity, the general consensus was that this was a red herring. The most positive view was that there is likely to be some impact in this area, but it is impossible to measure in any tangible way, so why would you dilute an otherwise solid business case with something that could easily discredit it? Best to stick the list of intangibles in your bottom drawer and run with what you can defend with confidence.

And it is on this point that the CIOs I have been speaking with diverge from the view articulated by Microsoft. In fact one said the obsessive reference to the great user interface, user facing productivity features, etc caused a lot of distraction and confusion when he invited a Microsoft executive to meet some of his business sponsors. When a stakeholder says, “I don’t understand, I thought we were doing this to save money”, it doesn’t actually help to get the investment case signed off.

There are a couple of lessons that fall out of this. Firstly, if you are going through the process of evaluating the business case for Vista yourself, the abovementioned criteria will hopefully provide some thoughts based on where at least a few others have put the emphasis – particularly in a large corporate or public sector environment.

Secondly, the feedback suggests that you should be prepared for business sponsors to get confused about the rationale for migrating based on the messages broadcast by Microsoft both directly and indirectly through advertising, the media, marketing collateral, etc. The trick here is agreeing that it will be a great spin-off benefit if all of the claimed or suspected end user productivity gains are realised, but keep the investment case itself focused on the more solid stuff that can be defended under cross-examination.

Finally, there is a message in here for any Microsoft executives reading this. If you can curb your enthusiasm for obsessing about the Wow! and focus on the things that drive decisions, you might see more movement in the market.

Thursday 12 June 2008

Business Intelligence and the bolting horse

There appears to be a revival of interest in Business Intelligence (BI) among IT vendors at the moment. Some pretty big guns, the likes of Oracle, IBM, SAP and Microsoft, are trying to position themselves more aggressively in this space following the spate of acquisitions.

So is this renewed vigour justified?

Well from a customer perspective it undoubtedly is. It is pretty clear when you research BI that the gap between business need and IT capability is as great as ever. When we interviewed a bunch of senior business managers from City of London financial institutions last year, for example, they were very clear about this gap:

And if you look at this chart closely, you will notice something quite interesting. While business information availability isn't that bad at an overall financial and arguably operational performance level, it is not very good when you look at more detailed measures and indicators.

Why is this interesting?

Well because it tells us that by the time those managing the business find out about something important, it is often too late to do anything about it. Stories of product, client or partner related issues only coming to light when someone starts investigating why a higher level number has been missed are quite common.

To put it another way, business managers usually have what they need to monitor the ‘effects’ of doing business, but are typically underserved when it comes to the information required to manage the underlying ‘causes’ of those effects. We discuss this more in the research report from the study if you are interested, but it does bring home the importance of incorporating continuous analytics capability into the business process itself, as well as having traditional retrospective BI operating off to one side.

The aforementioned vendors are therefore spot-on when it comes to making a big noise about the principle of integrating BI capability into applications in a more embedded fashion. Now, whether they have done a good of integrating their recent acquisitions into their broader solution set in practice is another question, but it is at least worth hearing them out.

Sunday 1 June 2008

Talking at cross purposes, or being deliberately misled?

Ever had one of those conversations where you debate something for a while then it dawns on you that each party has been talking about something different? It has happened to me quite a few times recently.

One example was in relation to Business Process Modelling (BPM), which is something I grew up with and in my mind is about, well, modelling business processes. It’s a discipline that business analysts have been involved with for a years, and while the technology to support it has moved on, and arguably some of the methodologies too, the fundamental principles haven’t changed that much for a long time now. Then someone asked Freeform Dynamics to design a research study to figure out the level to which organisations had adopted BPM. When I argued during an internal project start-up meeting that you couldn’t really ask someone about when and how they were taking something on board that they had been doing for a decade or two, it turned out that the ‘BPM’ we were being asked to investigate was actually 'Business Process Management' and was based on a definition which included the technical side of things – workflow rules engines, SOA orchestration, and so on. Not quite the technology-independent business view of BPM that I was taught earlier in my career, but as soon as the misunderstanding was cleared up, we could design the research accordingly.

Another example was prompted by a report I read the other day claiming that Software as a Service (SaaS) is now a mature and pervasive model. This was reminiscent of claims made during a number of other conversations I have had recently with SaaS advocates, that I have been struggling to reconcile with the findings of our own research. The latter has shown quite conclusively that while larger organisations are starting to make selective use of SaaS for delivering business application functionality, 'pervasive' is certainly not a word that applies in this area. Then I realised that some of the advocates were throwing a whole bunch of stuff into their definition of SaaS (or the related S+S model) that I would never dream of including when discussing the delivery of business application functionality. Internet search, traditional ISP services, and even things like consumer content services, online help and automatic updates associated with desktop software can sometimes be lumped together when referring the 'SaaS market'. Again, once the ambiguity is cleared up, you can see where people are coming from, and make a judgement on the usefulness (or otherwise) of what they are saying.

I guess we at Freeform are particularly sensitive to precision when it comes to discussing market activity, as primary research designed to figure out what’s really going on behind the buzzwords and the hype is so central to what we do. The experiences I have outlined, however, highlight how easily people can be misled by imprecise or ambiguous definitions if they are not on their guard. And with so much vested interest and evangelism driving the market, the temptation for some to spin and exploit our ever changing vocabulary is significant, so we all need to careful about what is behind those stats and definitions.

Sunday 27 April 2008

Cloud Computing and Web 2.0

Don’t you just hate it when another woolly ambiguous term is forced upon us? When I was approached by yet another journalist the other day asking me my thoughts on the impact of cloud computing, I simply sighed and told them it is a bit like Web 2.0. In itself, it is difficult to pin down exactly what is meant by it. The best you can do is say that both of these terms refer to a general direction in which the industry appears to be moving.

In the case of Web 2.0, it is about the Web becoming a generally more interactive medium. This can manifest itself at a technology level through everything from Ajax through mash-ups to SOA, and at a behavioural level through social media and the simple fact that websites are generally now more geared up to a two-way dialogue than they used to be.

In the case of cloud computing, it is about the evolution of dynamic virtualised infrastructure that allows us to think more in terms of resource pools than individual IT components. This in turn opens the door to delivering computing resource on a utility basis, which is equally applicable both internally (i.e. with regard to the way you use your data centre) and externally – which takes you into the realm of utility computing and software as a service.

The point about both Web 2.0 and cloud computing is that they both sprung up arbitrarily on the evolutionary timeline, and seeming embraced anything and everything that could be thrown into the mix. While the very specific phenomenon of social networking is certainly noteworthy, this bears little relationship to evolution of rich user interfaces and composite applications, in fact many social networking sites have appalling UIs by traditional standards. Yet Web 2.0 can mean either of these things, and, confusingly, lots of other concepts too.

Similarly, we have been talking about virtualisation ultimately leading to computing grids and utility computing for years, and giving it a new name doesn’t actually change anything in terms of the underlying trend. In fact, you knew where you stood much better when you could talk about virtualisation and grid technology as the enabling stuff, and utility computing and application services as what it enables. As everyone jumps onto the cloud computing bandwagon, it all gets mixed up and confused, just like Web 2.0.

So, if you are one of those people wondering what cloud computing is really all about after listening the IBM explanation, the Microsoft one, and the evangelical rhetoric we have heard recently from the Google and camp, don’t worry, you are not alone. The trick is to think of it as a label for a trend at one level, and an industry bandwagon at another, and keep your expectations pretty low in terms of clarity and consistency for the time being. Don’t however, dismiss the underlying trend it itself. While we are not looking at a revolution here, some of the developments in this general area are really quite interesting and valuable – though, you probably knew that already, even before the marketing hype was thrust upon us.