There has been a lot of hype around cloud and perhaps justifiably so: it is exciting and, given the take-up, it is here to stay.
Companies that deliver services in the cloud are attracting phenomenal amounts of investment from venture capitalists and the number of vendors is growing exponentially.
At almost 50 Petabytes, Autonomy managed probably the biggest private cloud. What made it valuable is not its size, but the information it contained.
In this case, practically every piece of communication in most financial institutions and many FTSE100 corporations. And then, what we could do with that information provided extra value to our customers.
So now that I am unencumbered by any corporate ties and now that cloud computing has entered the mainstream, I wanted to give you my objective list of ten ideas in considering the cloud.
Some you may agree with these, some you may disagree but in any case, let’s start the debate:
1 Existing systems will not move
This means, if the legacy systems are meeting a company's needs, there is no reason to buy the application in the cloud.
So I am unconvinced of all the VC money going into cloud versions of standard applications. We may recall an old IT saying: If it ain't broke, don't fix it.
The cloud will triumph for new apps.
2 Security will become a non-issue
There is a belief that services in the cloud are somehow less secure than those on-premise.
In fact, at Autonomy we handled the most sensitive non-military information, as well as all the email messages of Wall Street with the highest levels of security.
This didn't just mean best-in-class encryption and encoding. It also meant very exacting codes of conduct demanded by customers at our data centers, where workers were routinely drug-tested.
Whilst there will be scandals and errors, I’m not sure it’s much different to on-premise. Those organisations that had no choice but to move to the cloud overcame the security concerns a while back.
3 Bag-of-bolts services will lose out
Beware of being sold what I call a bag of bolts. Some of the more out-of-touch large providers are creating a cloud strategy of bits from their catalogue.
The beauty of the cloud is a working service. A car, not a kit of parts. If you can't be provided with the fully-deployed service or software then think twice.
4 The buyer will shift
In other words, procurement of IT services is increasingly shifting from IT departments to the business units.
SaaS applications are user-friendly, easy to buy and much faster to install than obeying the corporate code of conduct, and allow business units to avoid high internal recharge and the delays of internal IT.
The CMO may soon be spending more on tech than the CIO. The cloud buyer is moving to the business units.
5 Unofficial cloud services will blossom
Like vertical companies that sell to direct business units through the cloud, large consumer companies, such as Amazon and Google and smaller outfits like Dropbox or the Box, are becoming parallel repositories of corporate information as executives use them to send each other large files.
Regulating this will be a challenge for IT departments in months to come.
The popularity of the cloud and the increasing savvy of the workforce mean that today, some services are being purchased directly from the web, purchased with a company credit card used entirely unbeknown to the IT department.
Bring your own device (BYOD) is another offshoot from the consumerisation of IT that is challenging CIOs.
Unlike the unofficial use of cloud that is taking place across desktops throughout the company, this one is driven by the top, as board members demand to use tablets and smartphones for work purposes, even though they may not be standard issue.
Now that the information is centrally in the cloud, we can see an explosion of devices accessing it.
6 Services may suffer
Whilst initially the cloud is a boost for services, it will become a major threat.
When customers can get standard cloud products that work instantly, rather than complex expensive customized solutions that take ages and cost a fortune, they will reject the need for a few extra customised functions.
In a world without change requests, the services model looks challenged.
7 The hardware market will be hit
Power density in the data center is all. Many organisations have vast server centers that are a challenge to keep cool.
New cool chips will be key. Another major move is from expensive powerful servers to arrays of cheap servers. This move hits hardware margins.
Also, the large cloud players buy in such quantity that they have all the buyer power. The cloud is not good for server margins.
8 Pace of software innovation will quicken
Change a license product and you need to test it for months before it ships to the most obscure corners of the world.
Find a fault in your cloud system, and you can fix it for all customers in hours.
So software will move to a more agile model, leading to faster innovation.
9 Cloud margins will be a battle ground
The world and his dog are rushing to be in cloud. Perhaps they have no choice.
Vendors are desperate to get customers and so are quoting low-margin prices as they race to the bottom with an argument that somehow we will make money from customers later if we only win them now.
Thus for many players, there will be no margin in the cloud. Think the disastrous effect of the rush to sign outsourcing contracts on many vendors.
When there is no financial model, eventually the customer also suffers.
10 Data ownership may become the key to market dominance
Perhaps the winners in the cloud will be the vendors that end up owning the customer data.
In the consumer world, think of all that Facebook or iCloud information. Once it is there, it doesn’t move. It ties the customer in and all actions on the information must go via the platform.
In the enterprise, are the candidates Autonomy and Salesforce or Amazon and IBM?
We are at the very beginning of the cloud computing revolution. Perhaps these observations will serve as some basis to think of its future progression.
Manganese Bronze Holdings underestimated losses due to accounting error