Will SaaS and Cloud Kill the Appliance Market in Australia?

In a period of intense application renewal the traditional suppliers of hardware appliances are being forced to re-evaluate strategies for growth. On the one hand mega technology vendors such as IBM, Oracle, and HP are consolidating solutions across all aspects of information, applications and technology stacks.

This shift is forcing channel partners and value added resellers to become proficient in true solution style engagements where they may have previously only been specialists in one point solution. For example as IBM moves ahead with its Information on Demand agenda incorporating information strategy, business intelligence, analytics, content management, and master data management it is no longer enough for an IBM BI channel partner to simply offer point solutions in Cognos.

At the same time, the as-a-service phenomenon continues to grow at pace, fuelled by end-user demands for everything from system performance improvements through to virtualisation, remote worker, and mobility enablement. As the available as-a-service offerings become more prevalent and wide-ranging to address these needs the fact is that any supplier of technology that is currently standing still with regards to as-a-service offerings will not meet the demands of their traditional client markets over the coming years. Change is required. However in the rush towards the cloudbased services only agenda that is sweeping the industry it is important that organisations not forget the role of infrastructure in the mix, even if it is only a transitory one. In this regards the on premise appliance is proving to be a significant change catalyst and is by no means dead. Instead the onpremise appliance may yet help to redefine the success of both the vendor channel and the end-user market in the current technology cycle.

The on-premise appliance is the physical manifestation of many as-a-service based offerings in the market today. Its role is important both from a raw technology perspective and a cultural adoption perspective. On-premise appliances give those companies that are unsure about first steps into as-aservice business models a sense of the comfort that purchasing traditional tin deployed in data centres and server rooms the world over has always provided. But assuming that on-premise appliances are sold as services, there the similarity ends. There are two clear types of organisations emerging for which the transitionary approach to utility computing through the use of an on premise appliance is becoming highly applicable. And specific solutions are emerging for each.

The first type can be categorised as the single office, the branch office or the SME. The requirements for these types of organisations are either the need for low impost management and maintenance support of technology (i.e. don't have much time and may not have much experience), cost constraints driven either by organisational size or a lack of planning and strategic ICT capability, or speed of deployment, relocation, or decommissioning of portable technologies based on growth or cyclical workloads in multiple locations.

The second type of organisation for which the on-premise solution is emerging can be categorised by the requirements of multi-site operations, remote locations, or the enhancement of mobility enablement within large enterprise. For this category particularly WAN optimisation is emerging as an important driver. It can provide true bandwidth utilisation reductions of between 60-95%and offerings from the likes of Riverbed and Cisco are being met with the same hype and fervour as the buzz created around virtualisation 2-3 years ago. Early indications are that Riverbed is the VMWare of the WAN optimisation space ahead of Cisco (or Microsoft to round out the analogy). For organisations assessing any kind of on-premise appliance based service, the adoption of a WAN optimised product should be a key assessment factor.

As cloud and software based services increase so too does the range of on-premise appliances that provide the grounded point of connectivity within organisations for the serving of those services. Table 1 outlines four (4) offerings that currently epitomise the range emerging in the on premise market with an "as-a-service" orientation or twist. Each involves the "dropping" of a physical appliance within the business to support a variety of business applications and services.


Click image for a bigger version

Of all the business drivers for software-as-a-service (SaaS) and Cloud Computing emphasised over the last few years from reduced capital expenditure, to the current end-of-life negotiations for traditionally licensed software it is impossible to ignore the cost benefits derivable from consuming network latency through WAN optimisation, mobile and remote connectivity through service centralisation, and virtualisation driven performance enhancements. However the question currently left unanswered is whether or not SaaS and Cloud Computing will rejuvenate or kill the appliance market in Australia and other ICT economies around the world. The answer to that may lie in understanding the logical maturity lifecycle of how ICT organisations will approach the adoption of SaaS and cloud.

The on-premise appliance addresses the adoption lifecycle because it represents an easy start for any organisation. This is the case because rather than jump directly into full SaaS offerings, the prevalence and constraints of consumer grade networking within many businesses in many ways necessitates that a first move is to this type of model. Not only do on-premise appliances with components that specifically target process optimisation (such as email document storage duplication and triplication) through WAN optimisation assist in managing poor network latency but it also closely mimics existing business procurement practices that today involve the act of buying or at least acquiring a physical device. The recognisable 15-22% annual maintenance charge on the box, but for which the services offered are far greater, will also be familiar, and ironically, comforting to most organisations. At least initially.

So ultimately the question of whether SaaS and Cloud will kill the appliance market will come down to the transition of "services" delivered through "appliances" and whether what is currently a good place for many businesses to start because of poor networks, retail and device procurement mentalities and tax breaks will evolve to its full potential. Organisations should not lose sight of the fact that only "as a-service" appliances or appliances that directly address the latency issue will be successful. Specialised appliances will die first whereas those that are smart will "virtualise" their appliances as soon as they can. The bottom line is that for now appliances are good but appliances plus services will dominate within a few years.

Finally, not all on-premise solutions on the market include WAN optimisation technology. Yet as general network capabilities improve with the roll-out and take-up of the National Broadband Network (NBN), the obvious end-point for the on-premise lifecycle is that everything will be sucked up into the cloud and centralised vendor data centres. The timing for that is extended and likely to be in the range of the usual 5-8 year hardware refresh. For now that means that the on-premise appliance remains a valid infrastructure consideration for Australian enterprises and those solutions which deliver the physical manifestation of "as-a-service" offerings will see significant uptake in 2010.