The reported ‘battle’ between Microsoft and Google for the delivery of on-line services saw Microsoft chalk up another win this week, as the university of Arizona chose the Microsoft On-Line services solution for it’s 18,000 seats.

This is a significant ‘win’ and shows that when the decisions are made it’s not only about pricing – Microsoft’s BPOS stack is typically around the $120 per user per year mark, compared to Google’s $50 – but also about integration, migration, security and risk mitigation. I’m sure that there was also a decent dollop of “I’d just feel better with Microsoft” in there too.

I’m involved with a couple of incipient projects which are also of the magnitude of the University and the decision processes are protracted as the risks of moving core services on-line are identified, discussed, addressed and re-addressed. The costs are very much secondary in these discussions as long as the mitigation is not overly significant.

The Microsoft announcement of the deal said “In addition to updating the university’s numerous and antiquated e-mail and calendaring programs to one centralized system, employees will get 10 GB mailboxes and new features such as instant messaging, presence and online meetings with the ability to share their desktop, audio or video with other users on or off campus,”. And this shows where the adoption of on-line services is really going to take root.

Many, many organisations – not just educational ones – have either old, legacy systems or a whole clutch of systems inherited though mergers and acquisition activity. The capital costs of replacing these systems in-house is usually significant and the benefits are often difficult to put a value on when building a business case. Moving the services on-line becomes THE viable solution in these circumstances as there’s little capital costs (the engagement of an SI for migration activity being the only significant, but potentially vital, one) and the ongoing costs are directly related to usage. The on-line services will usually provide a greater SLA than an in-house solution and will probably also deliver more storage and other functionality that would be implemented in-house too.

Microsoft, Google, or someone else. The decision soon will be not IF they move is to on-line, but WHO will be chosen to deliver the service.

The momentum is growing and the technologies are becoming more and more abstracted. I love both of these things and I’m going to tell you why.

The number of Service Providers offering ‘proper’ services is growing, by proper services I mean those which offer a valuable use for the technologies underneath without selling those technologies directly. I’m proud to have been, peripherally, part of the creation of Rise, the new B-2-B IaaS solution from Fasthosts which set’s it’s stall out very clearly in terms of the services it offers but makes little or no mention of the underlying technologies. Of course these can be gleaned by a visit to the main Fasthosts site where the DataCenter on Demand solution is pretty easy to find. The point is though that for Rise, it’s market simply aren’t interested in the “how” it works, simply in the that it works and provides what they need.

But Rise an DCoD aren’t the only solutions rolling in, look at nGenX. Their new GeoCoud offering takes the whole this a step further and abstracts away the datacenter itself, you pick the services you need and they’re delivered from ‘out there’ with geographical redundancy – obviously organisations with data-boundary limitations need not apply, but for the rest, it’s a nice way to be “global”.

The game continues to change. Service Providers with a quick eye are already part of that change, the slow-pokes may get left behind!

Last week saw Microsoft’s annual get-together of hosters and partners from around the world in Bellevue, Washington. A move of venue up the road to the Hyatt Regency was, in my opinion, a good move as everything seemed a little more accessible and open than previously.

I had been invited to take part in a pre-event workshop for solutions architects which was slated to provide some more technical depth that the other presentations which would take place over the following 2 days. And some of what was shown was actually pretty good. The Orchard project shows that Microsoft are starting to understand the demands of the consumers of services and aims to deliver a content management system based on MS technologies. In particular the use of SQLite as the database store is interesting as this in-process data-store means no more mucking around with database creation and management (probably the only sticky bit of using WordPress for example). The ‘consumer’ focus is also backed up with an entry level site-creation toolset which I would describe as Visual Studio with all of the complicated bits stripped out. This sounds like it’d be a bad thing, but the demonstration I saw was both accessible and seemed very intuitive.

This latter aspect has been achieved by bringing through some of the new Intellisense features which have been brought into both this product and the new VS 2010 product suite. I’m no dev, but this seemed to make the drudge elements – constructing function structures etc – and automate them, allowing the dev to simply plug in the relevant pieces. The final  and most interesting piece – for service providers – is the MS-Deploy endpoint. This IIS extension allows for devs using VS (and the new lighter product) to publish packaged applications and databases up to the server environment. If you are pitching at web developers then this is something your systems MUST be providing.

Video was a big topic too. We are apparently seeing a huge growth in video delivery and consumption over the ‘net – 180% over the last two years with 80% of all ‘net users watching video – and  this is predicted to continue. Service providers should certainly consider video publishing as a potential product, Expression Encoder is free to download and links directly to IIS for simple publication. The guidance here is pretty clear, provide a publication service with a link to some free software for encoding and publishing and get on board sooner rather than later.

Storage has historically been a sticking point when it comes to video delivery but the new paradigms enabled by the XXX-as-a-Service solutions means that everyone – including Service Providers – can take advantage of on-line, “Cloud” storage. After all this is the perfect example of utilisation, if I need the storage I pay for it and it’s provision becomes a direct, measurable, scalable cost for the the extended service I’m providing out to my customers. It starts getting very easy to bring products to scale when you are simply using a massive platform in an on-demand basis.

The rest of the workshop was less engaging however. A technical dive into SharePoint 2010 revealed that installing and provisioning a multi-tenant environment is still not as slick and smooth as it should be and some of the really nice feature of SharePoint (file-system storage of large binaries for example) are dependent upon features available only in the Enterprise edition of SQL server – this is a significant cost per month on SPLA and isn’t something which a lot of SP’s will stump up. This position runs contrary to Microsoft’s general position which is that SharePoint is a solid platform for website creation – cost is ALWAYS an issue when we talk about web-sites.  The final session covered Windows Server Futures. This actually covered Windows Server Pasts and was principally all about what was done in Server 2008 / SP2. The NDA elements regarding Windows Next were so generic that they could have been applied to almost any product produced, ever. Sorry this was rubbish!

Overall it’d have been nice to get a little bit more architectural information throughout the day. We’re about building systems and solutions and there was nothing there to support any thought processes in that direction. I did mention this and there was an acknowledgment that this was a missing element in the day. WIN!

The Summit proper was principally a set of keynotes and, for me, fluff. Yes, we know that on-line services are a big thing, yes we know that Microsoft are “All in” with the Cloud. Of course Microsoft still see hosting partners as strategically important, as a revenue stream SPLA is pretty significant! What we are seeing now though is that, rather than cookie-cutter technical solutions the guidance is now around value differentiation. I’m very happy about this as it vindicates what I’ve been belting on about for a couple of years now, if not longer.

The bottom line?

Microsoft ARE committed to on-line delivery of services. See BPOS, see Office Live, see Azure. They are also committed to spreading the use of the Microsoft stack to deliver services through their hosting partners, what’s changed over the last 18 months though is the method through which this commitment is being delivered. Gone are the end-to-end hosting solutions of times past and in their place are products and services which are designed with ‘Net delivery in mind and open-sourced examples.

The result is that many of the smaller hosters who relied on the plug-and-go solutions are now faced with a significant choice, if they want to continue and advance their solutions they need to start investing in the development and differentiation of their services – otherwise they may well wither on the vine.