Step, Path, Shoes, Sole by schroederhund at

Every journey starts with the first step

There’s been a couple of interesting stories in the tech press this week…

Firstly around the latest profit announcements from both AWS and Microsoft. Amazon as a whole reported a “very strong quarter” with AWS driving a significant proportion of that:

AWS revenue for the three months to 30 September 2017 hit $4.58bn, resulting in operating income of $1.71bn — up 8% from the previous quarter.

Meanwhile, Microsoft is reporting that it has “just hit its $20 billion cloud goal, almost a year ahead of schedule”:

The software giant’s commercial cloud business is anchored by Azure, its cloud computing service, and by the Office 365 productivity suite, both of which are growing like crazy. Azure revenue jumped 90% in the company’s fiscal first quarter, compared with the same period a year earlier. And revenue from the business-targeted version of Office 365 grew 42% over the same period.

Secondly, the news that HMRC’s switch to AWS killed a small UK cloud business, DataCentred.

The business was started by TeleCity founder Dr Mike Kelly with £9m funding from venture capitalist John Moulton and local authorities. The firm turned over £1.2m in 2016, but was plunged into the red after losing the HMRC contract and subsequently went into administration in August.
Reg readers will no doubt debate the inherent weaknesses of being overly reliant on one customer, but also the irony of HMRC paying for the services of AWS, a firm sometimes criticised for its tax efficiency.

Commentary on the two stories is somewhat to be expected, generally along the lines of “the UK government is breaking its own commitment to support SMEs by taking business to hyperscale public cloud providers” and “why is the UK government supporting a public cloud provider like AWS who doesn’t pay its correct share of tax”.

I don’t particularly want to get embroiled in the debate around the second of these — even if true, it strikes me as highly likely that AWS are playing within the rules set out by Government. If the rules aren’t working for tax payers (and they may well not be) the fault lies at least as much with Government as it does with anyone else.

As a tax payer, I want the big players (all big players, not just the big cloud players) to do the right thing (i.e. pay their fair share of tax) but that requires that both they and the Government work together to make that happen. Actually… scratch that — I just want all tax payers to do the right thing, irrespective of size!

On the SME issue, I think it is worth remembering that all the major global public cloud providers (I’m thinking particularly of AWS, Amazon and Google here) actively encourage an ecosystem of smaller partners and other players to grow up around them. So whilst some of the SMEs who used to thrive in the old world might suffer, I guarantee you that there will be others that are emerging profitably in the new world. What is the balance between SME winners and SME losers? I don’t know but G-Cloud sales numbers might at least provide some indicative numbers.

This is a particularly pertinent story for Eduserv, as we transition ourselves from the old world to the new world — not an easy journey I can tell you.

We’re doing a lot of talking to our public- and third-sector customers right now — both existing and prospective. On public cloud adoption, we’re seeing some challenges in the public sector which are worth noting. We find that, often, one or other of the following ‘cloud adoption’ anti-patterns comes into play.

Colocation is good enough to take the cloud foot off the gas. Where closing a data centre is one of the key drivers for moving to the cloud, we often see public sector organisations using colocation as a staging post in that journey. To be honest, we often actively encourage this approach. The story goes that we can free up on-premise real-estate faster by lifting-and-shifting infrastructure to a colocation staging area than we can by moving direct to the cloud. And the move carries the promise of onward migration to the cloud. The anti-pattern is that as soon as the lift-and-shift has happened, the pressure to move is removed and the drive to cloud significantly reduces.

In a time of financial pressure, this lack of drive can be so significant that the cloud move never happens.

Short-term financial constraints trump long-term strategy. The other major driver for public cloud adoption, wrongly in my opinion, is cost-savings. In this case, what often happens is that an organisation that has notionally taken a strategic decision to move to the cloud gets stuck because each individual migration that they consider (particularly when taken in isolation) may not stack up financially against the historical investment that they have already made in their own data centre facilities. The anti-pattern here is that organisations never take the first step to the cloud, because the short-term financial constraints take precedence over any long-term strategic gains (many of which will not be directly financial).

In both cases, I would encourage public sector (and other) organisations to ask themselves the following two questions:

  1. Do we still want to be running our own data centre in 5 or 10 years time?
  2. How can we attract the best and brightest people to come and work for us to drive significant change in our ability to deliver the best digital services to our citizens?

If the answer to the first is “yes”, I would say, “Good luck with that — let us know how you get on (but remember that if you change your mind in 2 or 5 years time, you’ll be 2 or 5 years further behind the cloud curve than you otherwise would have been!)”.

On the second, I would argue that to attract the brightest and best, public sector organisations have to position themselves firmly within the mainstream public cloud arena. That is where all the action is. We are no longer in 2003 — VMware really isn’t sexy any more. Get real!

I don’t know how we make that happen — particularly in local government where the financial constraints are so pressing right now. But we have to do something because otherwise we are storing up a reservoir of poor digital service delivery in the local government sector for years to come.

We need to see a change of mentality. We need some imagination. We need to stop seeing ‘cost’ as our only driver. As a citizen, I want to see the same kind of drive for better digital services that I see in GDS and central government to be also coming from my local council.

My gut feeling is that the change of outlook probably has to come from central government in some way because it’s not clear to me that local government, as a whole, can make that change on its own?

In the meantime, if you work in a public- or third-sector organisation that recognises it needs to get to the cloud somehow, pick something. Pick anything! And take your first step. You’ll learn a lot and when you get to the second thing, it’ll be easier, faster and probably cheaper.

But take that first step because until you do, your journey to public cloud hasn’t really started.