Watch Google!

Google is pursuing a very powerful strategy. One way of looking at it, odd though it seems, is to consider the original browser wars between Netscape and Microsoft. Netscape had the opportunity to undermine Microsoft, but it did exactly the wrong thing. It took the battle to the PC, delivering browser functionality on the PC that Microsoft could match quite easily. The simple bundling of IE with Windows destroyed Netscape's challenge in very little time. Netscape should have chosen to fight on the server.

As it happened, the company that won the browser wars was Yahoo. It provided much of the server functionality that Netscape should have been focused on. Unfortunately, for Yahoo, it never saw Google coming. Google came right out of left field. It established its search engine business in a crowded market when it should have been all but impossible. Its killer capability was to provide a well designed interface. (Damn it, its interface is still streets ahead of the competition.) Then it rose to challenge Yahoo, and AOL, and Microsoft, and it gaily marched past them all.

This week Google announced a chat (instant messaging) capability that also happens to be a VoIP capability. You can only sign up if you are also a Gmail (Google email) user – but now it has opened Gmail up to all comers to make the adoption process easy.

This is what I think is going on: Google is going to deliver unified messaging as a service. Right now it's not sophisticated, but right now nobody else is doing that, so it doesn't matter too much. Google's chat capability is inferior to those of Microsoft, AOL and Yahoo and its VoIP is inferior to Skype. Never mind, it won't remain so for long. And as for VoIP, I had been wondering who would be able to challenge Skype and now we have the answer: Google.

How Google will evolve its service is not so easy to predict, but it will – it will bring VoIP, chat and email together by virtue of a single directory. If Google gets it right quickly then it will gradually attract both consumers and small businesses. (Why not have Google look after all of your messaging for free?).

Of course, this is not Google's only initiative – it is competing furiously to maintain its grip on the search market – but this is the initiative that will give Microsoft pause for thought. Maybe Microsoft will buy Skype? (No, I'm just kidding).

The Appliance Trend

I had a meeting on Monday with Patrick Peterson, the CTO of IronPort. Among other things, we discussed the appliance market. Appliances are, as you probably know, servers in a box, which have a well defined role. IronPort appliance products, for example, are email gateways that stop spam and email viruses and allow for the definition of email policy.

We discussed the appliance market and the clear trend to delivering applications ready loaded on appliances.The attraction of an appliance is that it can be implemented without disruption and will deliver value quickly. Also, it is simpler from an administrative perspective to source a solution to a problem from a single vendor.

It took the market a while to accept the appliance idea. Three years ago there weren't many appliance products, so there was naturally some distrust of the idea. Some early appliance products caused problems by not being as transparent and isolated as they should have been, perhaps causing related applications to fail – teething troubles. However, there was already a tradition of buying 'network boxes', routers and hubs, from the likes of Cisco, which were usually rock solid.

In any event, nowadays when I'm being briefed on appliances, I always ask questions about:

1. Time to configure: usually it's hours or at most a day or two.
2. Resilience: usually there will be built-in redundancy and a pass-through capability so that if the appliance fails it doesn't affect any other part of the network.
3. Time to value: this is application dependent, but often vendors can legitimately claim to deliver value in the first few weeks after implementation and will deliver full value very quickly.
4. Integration: this naturally varies according to what an appliance does and thus what it may need to integrate with, but appliances that relate to system management in any way, normally provide integration capability.
5. Management Overhead: this too depends upon what the appliance does, but normally it is low – and it needs to be.
6. Cost: what is the lowest cost entry point?

In the last year or so I've looked at a wide variety of appliances (password management, IDS, email appliances, encryption/VPN appliances, and a variety of system management appliances). What appears to be the case is that IT departments are now very partial to buying appliance solutions if these can demonstrate the right characteristics and can be proven to deliver value.

I have no statistical evidence, but I'm getting the definite impression that appliance companies seem to grow faster than software-only companies. This is so much the case that I will often ask software vendors why they do not implement their capability on an appliance. Interestingly, Dell now has a division devoted specifically to building appliances – and from what I hear, it is doing excellent business. The appliance is here to stay, for a while at least.

Linux PC – Slow Progress

I didn't expect the Linux PC to gain much traction this year. It doesn't play particularly strongly in the home market. Admittedly, it can be found in some impressive outlets (Wal-Mart, Micro Center and Fry's in the US) and certainly, significant volumes of units are shipping, (tens of thousands per month). However, it is far from being a mainstream home PC product. It has been eclipsed to some degree in this market by the sudden rise of Apple.

What is clear though is that it is making strong progress in the business market. HP cites the thin client as offering the most promising area of growth for desktop Linux and there is a wide variety of devices from many vendors (Wyse, HP, Symbio, Fujitsu-Siemens, NIC, Neoware and many others). Devices tend to be in the $200 to $300 range and are usually diskless, which translates into a very low management overhead. There are a variety of application areas and contexts, such as retail kiosks, where a browser-based device is all that is needed.

What is less clear is what is happening in the developing world – which is where I expect the Linux PC to proliferate first. I have heard references to Linux being used extensively in Africa and in the less developed parts of Asia. In South America, there is the high-profile switch to Open Source and Linux on PCs that is underway in Brazil, at the instigation of the Brazilian government. The Brazilian government intends to push the advantages of Open Source at a UN IT summit in Tunisia in November, and has the backing of India for this.

Open Source clearly has mindshare in developing countries, but this may not translate into market share particularly quickly – by which I mean that a fast growing market for Linux PCs isn't going to suddenly bloom. A large commercial market will take time to form. Nevertheless, I can think of nothing that will stop it from happening.