Tuesday, February 15, 2011

SaaS suppliers are no black box, evaluation is key

Don't short-cut your due diligence

An aspect of cloud computing that is often overlooked is supplier evaluation and management. We spend time considering the pros and cons of new delivery models, but just like any other area of IT, the vendors and service providers you choose to work with have a huge bearing on the outcome of any investment or contract commitment.

This is particularly relevant in the area of Software as a Service (SaaS), the flavour of cloud computing that from an enterprise perspective is typically aimed at delivering business application functionality as a hosted service. Offerings here range from full blown ERP and CRM, through comms and collaboration solutions, to more tactical services aimed at workgroups or even individual users. Some would characterise online backup, security and other systems level capability as SaaS too, but for the purposes of our discussion, let's stick to the business functionality layer.

The first point to make is that it is still relevant to consider pretty much all of the criteria we would normally apply when selecting a piece of traditional on-premise application software. The solution must meet immediate needs in terms of business functionality and allow configuration around policies and processes as necessary. It will also ideally be flexible and expandable enough to deal with evolving requirements over time.

Stating such things may seem obvious, but all too often the seductive immediacy of SaaS, with many providers promoting a 'subscribe and go' approach to adoption, leads to short-term thinking and short-cutting of the important needs assessment and gap analysis process. If anything, the requirement to ensure a good functional fit with room to grow is even more important with SaaS.

While most organisations nowadays try to minimise the degree to which they customise core parts of application packages, at least it's an option with on-premise solutions should it be necessary. With SaaS, the software installation is not yours to tamper with, beyond standard configuration facilities, so there is a limit to how much you can work around supplier imposed constraints.

This brings us onto another area of common misconception. In the context of larger scale enterprise computing environments, the chances are that SaaS solutions are likely to have to integrate with other systems and technology.

The technical architecture therefore really does matter, at least insofar as it manifests itself in terms of application programming interfaces (APIs), support for industry standards, and interoperability with existing management tools, security systems, desktop software, mobile devices, and anything else that is important to you.

Thinking of SaaS offerings as black boxes is generally a mistake, especially when we consider that so much cost, risk and general frustration and inefficiency in IT is as a result of fragmented and disjointed systems.

Turning to providers, we clearly need to consider the basics. It's important to consider how stable the organisation is financially, how well it is tuned in to different customer requirements in terms of industry and size, and how capable it is of providing the kind of support needed in the places it's required.

On this last point, there is a specific consideration with SaaS around the support process. Some providers are comfortable with and capable of supporting end users directly, while others require the customer's IT department to deal with front line interaction.

There is no absolute right or wrong model here, but mismatches between requirements, expectations, capability and processes can cause significant issues if they arise.

It doesn't help, for example, if the provider is offering end-user support but works on the basis of strict demarcation of responsibility, resulting in users being handed off too frequently to other parties because the problem is not proven to be associated with the service.

Neither is it helpful if in-house support staff are taking calls about the service, but don't have the visibility or tooling to troubleshoot problems effectively. In today's IT service delivery environment, the need for end-to-end service management, and therefore end-to-end visibility, diagnostics and troubleshooting capability, is becoming ever more critical to meet business expectations.

Another practical area to consider is the supplier's policy on software release and implementation. Some providers trickle out a constant stream of updates in the spirit of continuous improvement. Others have one or two major releases per year, with very little happening in between.

There is then the question of whether you will have a choice of if or when you take a new release on board, with some providers allowing a degree of customer control, and others forcing new updates onto all subscribers unilaterally. Again, there is no right or wrong, but again, mismatches can lead to issues.

The areas we have discussed are just some examples of what matters when it comes to SaaS supplier selection. We could extend the discussion to include pricing models, contract terms and a number of other areas. The bottom line though is that due diligence is just as important with SaaS as it always has been with any key IT related decisions.

Originally published on cio.co.uk.

Tuesday, February 08, 2011

Next generation systems management

Strategic enabler or survival tactic?

The IT environment is complex, and undergoing continual upheaval. For many businesses, although the IT infrastructure is felt to be modern and generally fit for purpose, many challenges remain, especially around operational issues. One of the most common headaches we hear about in our surveys is just how difficult it is to keep up with the day-to-day operations, let alone optimise IT to face new challenges.

A big part of this is down to the lack of investment in management tools that are able to deliver joined up IT services to the business. Our research shows that most of you are probably sitting on management environment that is not really performing as well as you would like.

Historically, IT has evolved in a fragmented manner, resulting in an equally fragmented management environment. As you look to the future at where the business is going and how IT needs to support this change, you will need to take a more joined up view of systems management.

This doesn't mean that you need to go out and spend money straight away — if at all. Instead, the place to start is to have a good look at what you do, and how you do it. You may well find inefficiencies, overlaps and conflicts, as well as gaps and hidden costs.

Organisations of all types can benefit — from those where IT is a core part of the company and funded to enable innovation — to those where the least amount possible is spent to keep things ticking over.

Let's consider a couple of scenarios.

If you are working in one of those companies that invest heavily in IT, the challenge is not getting systems to work together. Management processes tend to be well defined with many formal processes and policies, and integrated tool sets enable joined-up IT services to be delivered to the business.

Instead, the issue is likely to be scalability and responsiveness. The big problem is that much of the management is still reliant on people to make changes. Unfortunately, repetitive tasks are limited by the productivity of individual staff, adding massive overhead. People also take time to prepare and travel, and to make and test changes. The end result is that change is slow and costly, although very well managed.

The reason this issue is being felt more keenly is that IT is becoming more dynamic. Virtualisation is taking hold, and dynamic IT is starting to have an impact. Doing things the manual way is not really viable in this new virtual world, and change needs to happen. This means that it is time to look at automation.

You may have looked at this in the past and felt it too risky to be practical. Time has moved on, and you may well find that to ignore automation is the risky option. Modern tools are able to integrate many more applications and automate much of the routine management activities while still allowing manual control where required for dealing with exceptions.

Implemented properly, it can also greatly aid the implementation side of change management once decisions have been made – although arguably the whole approach to change management will require significant changes too to deal with a more dynamic business.

Moving to automation will be a long arduous process. Politically, there may be fears of job losses, but the most likely outcome is that operations staff are freed from mundane, low impact tasks to dealing with much higher value and more interesting work.

If you are having to manage an IT organisation that is not valued by the business, investment in management tools may not seem at first glance to be an obvious way to save money. Indeed, just making the case for investment in modern management tools will be a challenge all in its own right as we are told repeatedly that senior management do not recognise the link between better management and the services delivered to the business.

But consider the alternative — a patchwork of tools that have evolved as various systems and services have been put in place that add overhead and complexity to the on-going operations. Ask yourself how many different tools you use for troubleshooting problems or for day-to-day systems management.

Each additional tool adds time and cost, while also requiring investment in training and support.

It also makes it more difficult to implement any form of structure and process around systems management, as difficulties in working with different management systems is often cited as a problem. This is important because end-user satisfaction then also tends to be low, resulting in a chicken and egg situation where expectations of IT are low, and therefore the willingness to spend more on IT remains low.

In this situation, the most pressing problem is to consolidate and simplify the management environment. The feedback we've had is that moving to lightly structured management processes works well in combination with a single main management suite. The reality is that the management processes are likely to be attempted initially, with difficulties in implementation prompting investment in the tools a little later on.

You don't have to go all out for completely integrated IT systems management. Starting small and implementing changes a step at a time can go a long way towards increasing the value of IT to the business.

Identify which issues are having the greatest impact on management, and tackle these with investment in updated or new tools where possible, rather than slapping structure on areas that have a minor role to play but that are easy boxes to tick.

Whichever situation you find yourself in, investment in improving your systems management capabilities is likely to generate big results. The problem is that it is often realised too late in the cycle to have the most impact.

If we can leave you with an opening thought for the New Year, it's to put management at the heart of all new projects.

Originally published on CIO - Registration required