Web Applications: How will they affect the server market?

There’s an ever increasing drive towards the use of web applications, as opposed to running various services locally and in-house.  But what does that mean for the corporate world?

Not that long ago, for corporate e-mail as an example, the only feasible option was to host your own local mail services, on single or multiple servers depending on scale.  This is where services such as Microsoft Exchange became popular – business e-mail, contacts, calendars and collaboration within the organisation for all employees.
Nowadays, there are multiple online services which offer similar hosted services, but for a monthly cost – no local servers, no initial outlay, no costly upgrades, and no costly administration time required.

Google Apps is one such example of how things are changing not only for the individual, but also for businesses around the world – the case is simple – why spend time implementing a costly system, on your own hardware, when you can use a service, often better (although that is subjective), for an ongoing monthly or yearly cost.

With the announcement of Google Chrome OS, the case to use web applications is greater than ever – Chrome OS’ primary focus is on web applications, so it is clear on which direction Google would like to see people going.

But what about manufacturers such as Dell, HP, IBM, and the many other server suppliers?  Undoubtably, this trend will effect their corporate sales, but to what extent?  It seems to me that these companies may have to reassess their markets at some point soon for certain sectors.

Companies in the small/medium business area are morely likely to be looking at the new web applications, and certainly be interested in whether or not it would work for their business.  The companies who would be purchasing a single server to host their mail – would they do that, and incur the costs of the setup, maintenance, and upgrade costs if they could use a perfectly functioning, fully featured web application? Web applications are always getting better, and there will come a point when the benefits are too difficult to ignore.

The majority of articles around the internet are focusing on how Chrome OS and the trend to use web applications will affect Microsoft and Apple – both who have their own, successful, server operating systems and applications.  In the case of Microsoft, their Office package is used worldwide, by millions of people, and it’s a similar story with Microsoft Exchange for e-mail.  Both, however, are replaceable, and as mentioned above, Google Apps has made big steps in doing that.

Very few articles are focusing on the impact of the manufacturers of the hardware they run on and there seems to be some big questions which we will see work out over the coming months and years.  But how can they minimise the impact of web applications to their business?

  • They may need to start focussing their core sales at larger corporations, who may not be able to benefit straight away from these new web applications, and still require in house services, maybe for regulatory purposes.  I’m not sure migrating 50,000 employees to a hosted e-mail service is a great proposition right now.
  • The providers of the applications themselves – end user companies may not be buying servers, but the companies developing these web applications most definetely will require hardware to serve their new application to the world, and for the successful ones, serious infrastructure will be required – large storage arrays for the data, many servers to ensure redundancy and high performance, and a solid network infrastructure.

However, there does seem to be an increase of  development towards higher end solutions – new virtualisation techniques aimed at data centre usage, new storage hardware including more affordable SANs, and new ways to save energy usage for large server farms (again, very useful in a data centre environment where power is a significant cost to consider), so maybe they’ve started to anticipate this trend already.

What do you think?  Maybe you work in a company as a system administrator – would you outsource some core services and demote those local servers? I’d really like to hear your thoughts, so please leave comments on this post or tweet about it on Twitter @dazuk.