eDiscovery in the Cloud - Watch Out!

It is clear from many surveys and indeed from my previous posts on cloud computing that data security, accessibility and location are key concerns for companies considering cloud computing solutions. However there is yet another area of concern around data in the cloud that requires consideration and discussing with cloud vendors before embarking on a cloud computing project. This area of concern, which isn' t much talked about in cloud computing circles, relates to eDiscovery.
As with any on premises application, courts and other authorities can issue eDiscovery orders at short notice and with tight delivery timelines (in some jurisdictions this can be as little as 24 hours). The question organisations must ask themselves is will their cloud vendor be able to support them in responding to such an eDiscovery request within the required timeframes. Remember responding to such a request will often require trawling through large volumes of data. This is difficult enough when all the resources are in your control. However, in the cloud scenario where infrastructure is shared between customers this becomes more complicated - for example how will the data be searched without impacting your operational performance for you and that of other customers and will the vendor have the resources and tools at hand to do discovery at all?

None of these issues are insurmountable but it does highlight again the absolute need to think through all impacts of the cloud computing model in advance of embarking on the journey and to ensure you have a comprehensive SLA in place with your vendor.

A Brief Introduction To Cloud Computing Part III - IAAS

In this post in my series on an introduction to cloud computing, I am going to look at the categories of Infrastructure As A Service in some more detail. Like SAAS there are a number of sub categories of Infrastructure As A Service offerings. Key amongst these are:

• Storage
• CPU / Server
• Database
• Integration

Storage

Storage As A Service enables organisations and individuals to store their data in the cloud at a much lower cost and with good disaster recovery and availability options. This is growing in popularity as a low cost archiving and backup option. The ability to quickly scale and up down the required storage is a key benefit of storage as a service.

Examples include Amazon Web Services and Right Scale.

CPU/Server

This is probably the most talked about category of Infrastructure As A Service and effectively involves renting processing power. I prefer to think of this as just that - processing power - as opposed to a particular type of server which somehow seems contrary to the general cloud concept. Having said this, in reality, computing power is normally purchased based on server type, operating system type etc. Again key benefits here apart from price are flexibility, ability to easily scale up and down and the outsourcing of administrative operational tasks.

Examples include Amazon, RightScale and GoGrid.

Database As A Service

Database As A Service offerings provide access to a database or database management system in the cloud. Often this is an add on to CPU as a service where an users requests a CPU with a particular database platform pre-loaded.

Examples include Amazon Web Services.

Integration As A Service

Sometimes this is classified as Software As A Service but I prefer to classify it as infrastructure as a service because of its 'plumbing' nature. Integration As A Service offerings are relatively immature but aim to make integrating cloud and on premise applications easier.

Examples include: Pervasive and CastIron.

In my next post on Cloud Computing I will look at the categories of Platform As A Service offerings.

LinkedIn as a business funding tool

I found this article on how an Irish Software company raised funding using just LinkedIn. While the money raised wasn't huge, I thought it was a great example of how Social Media tools are changing how business is and can be done on the Internet. The article is available here: http://tinyurl.com/yecz8lk.

IT in 2010 - The Right Solution At The Right Time

One of Deloitte's 2010 Technology predictions (www.deloitte.com/ie/tmtpredictions) is that Moore's Law is alive and well but in a different form that heretofore. In this era of mobile and handheld devices the predictions argue that rather than the focus being on faster chips it will instead be on chips that are 'good enough' for the job at hand in terms of processing power but which are cheaper and require less power.

This prediction struck a chord with me in terms of some of my recent client experiences and the pressing need to make IT cheaper and more efficient. As a general rule the IT industry (and their business customers) have often defaulted to 'gold plated' technology solutions, be it in the form of big CRM systems, high availability, disaster recovery or a myriad of other areas. In far too few cases have these decisions being challenged. Ask yourself did you really need that tier one CRM solution?, did that service really need a hot standby DR capability? why does that application need 99.9999 availability? Often, an honest assessment of the business needs and the risks can lead to a much cheaper and sometimes more effective solution that is good enough for what the business needs. This is not to say that we shouldn't design and deploy solutions that consider future business needs and which are flexible, but rather that we should deploy the right solution at the right time.

In 2010 lets try to focus on doing just that with renewed vigour.

2010 TMT Predictions - Whats the difference between T, M and T anyway?

As you know I work for Deloitte. Today Deloitte published the 2010 Technology, Media and Telecommunications predictions. Personally I found them fascinating. You can find them at http://short.ie/ie377t.

The Predictions
I particularly like the Moore's Law, Net Tablet and linear TV schedule predictions.

Top Technology Predictions include:
1. IT procurement stands on its head
2. From grey to green: technology reinvents cement
3. Smaller than a netbook, bigger than a smartphone: net tablets arrive
4. Moore's Law is alive and well in 2010
5. Cloud computing: more than hype, but less than hyper
6. Thinking thin is in again: virtual desktop infrastructures challenge the PC
7. CleanTech makes a comeback. But solar stays in the shadows

Top Media Predictions include:
1. Linear's got legs: the television and radio schedule stays supreme
2. Publishing fights back: pay walls and micropayments
3. The shift to online advertising: more selective, but the trend continues
4. eReaders fill a niche, but eBooks fly off the (virtual) shelves
5. Music as a service rises up the charts
6. TV and the Web belong together, but not necessarily on the same screen
7. Video-on-demand takes off - thanks to the vending machine
8. One step back, two steps forward for 3D TV

Top Telecommunications Predictions include:
1. The smartphone becomes a search-phone
2. Mobile VoIP becomes a social network
3. Widening the bottleneck. Telecom technology helps decongest the mobile network
4. Paying for what we eat. Carriers change data pricing and make regulators happy
5. Nixing the nines
6. Contract 2.0: long-term solutions shorten and multiply
7. The line goes leaner. And greener 8. Reliability redefined and reassessed

My Thoughts

I was involved in putting together Deloitte Ireland's point of view on the results and so spent quite some time going through each prediction for each of the TMT areas. What struck me was the blurring of the lines between the predictions. Almost without exception the media and telecommunications predictions were reliant on technology. For example the telecommunications predictions talked about the SmartPhone becoming a search phone and mobile VOIP becoming a social network. While these are certainly telecommunications predictions, I don't think many of us would have been surprised to see them in the technology predictions either. Similarly on the media side there was a debate around eReaders, Music As a Service and TV v xPlayer services. Consequently, when debating which of our clients to send them to I argued that all our CIO and Senior IT contacts would be interested in all three.

I wonder next year will we have just one set of predictions?

A Brief Guide To Cloud Computing Part III - SaaS in more detail

In this third post on Cloud Computing I will look at SaaS in a more detail. Software As A Service can be thought of a web based services that supports one or more business processes. The services are targeted at the business user (in general) and business super users or business oriented techies who configure the applications and how they integrate with other cloud based or in house applications.

There are many SaaS services available in the market and more are being added every day. However, most of these fit into a couple of sub segments -

• Customer Relationship Management.
• Enterprise Resource Planning
• Supply Chain Management
• Digital Content Creation
• Communication Oriented Services supporting collaboration, instant messaging, video conferencing etc.
• Integration services for managing messaging between systems.

CRM
The CRM sub segment consists of those services related to all aspects of customer relationship management and is one of the most mature segments. Examples include SalesForce.com (the market leader), SugarCrm, RightNow and of course Oracle and SAP's offerings. There are also a range of other offerings in particular niches.

Communication Oriented Services
This category covers a wide variety of services including e-learning, collaboration, enterprise content management, instant messaging, email and team collaborations and is probably the most widely adopted SaaS sub segment.

ERP
To date there has been relatively little adoption in this area. The most successful services are those that can be used for business processes that are relatively generic accross all industries e.g. HR, recruitment, expenses etc. Examples are SAP By Desigm, WorkDay and NetSuite.

Supply Chain Management
This segment is focused on services that enable organisation to manage and improve their supply chain and suppliers. This segment is currently growing quickly and new offerings are appearing regularly. Examples include Descartes, Ariba and Ketera.

Digital Content Creation
This SaaS segment is currently fairly underdeveloped despite a huge demand for digital content creation applications. Examples include Adobe and YouTube.

Integration Services
This part of the market is also relatively immature and is unlikely to grow very quickly due to the cost of switching. It does however have great potential. Examples include Pervasive, CastIron and BoomI.

Summary
The list above is far from exhaustive. As can be seen above there are many SaaS offerings in the market and the market is growing and changing rapidly. Predictions as to growth in the sector vary but in general there is general agreement that the CRM and Communication Oriented Services segments will continue to grow strongly over the next 18 months.

In my next blog posting I will look at Infrastructure As A Service in some more detail.

A Brief Guide To Cloud Computing Part II - Drivers & Inhibitors

In my last post on Cloud Computing I provided a brief introduction to what cloud computing was and promised that in my next post I would look at the inhibitors and drivers of cloud computing in an attempt to bring greater clarity to the subject.

As we all know there is no such thing as a free lunch and like most things in life Cloud Computing is a trade off. There are clear advantages to adopting the technology but at the moment there are a number of issues which are inhibiting its adoption. For organisations considering adopting these technologies it is a case of balancing these two sides of the equation.



Drivers

Lets first look at the drivers of cloud computing - the why would you do it question. The primary drivers for cloud computing are:

1. Financial


2. Speed of Deployment


3. Flexibility & Scalability


4. Focus on core competencies


5. Less deployment risk


6. Less IT Maintenance and
   upgrade activities


7. Easier Procurement

Financial

If you remember from previous posts one of the key characteristics of cloud computing was that it was multi-tenant and paid for on a pay as you-go basis. One of the benefits of this is that it leads to lower capital costs related to the cloud infrastructure and/or software. In fact the costs become purely operational and at pretty much any point you can withdraw or up/down scale the size of your cloud solution without any financial penalties or having to write off capital investments. A side affect of this is that the capital that is freed up can be used to accelerate other capital intensive initiatives in the wider organisation that might otherwise be possible. Clearly lower capital costs also lead to improved ROI and a shorter payback period.

Speed of Deployment

Cloud solutions can be deployed much faster than traditional IT solutions because the solutions tend to be standardised, based on pre-defined templates (that limits customisation) and make extensive use of technologies such as virtualisation.

Flexibility & Scalability

Following on from the previous point the speed of deployment benefit does not apply to just the initial deployment but also to scaling the solution up and down. Cloud based solutions are designed from the ground up to be immediately scalable in either direction, thus allowing organisations to rapidly react to daily, seasonal or yearly spikes in demand for computing power.
Focus on Core Competencies

By leveraging Cloud Computing to effectively outsource all or some IT functions organisation can focus on what they regard as the competencies that gives them competitive advantage.

Less Deployment Risk

Deploying cloud solutions is generally automated and largely taken care by the service provider thus lowering the deployment risk for IT organisations.

Less IT Maintenance and upgrade activities

As I've alluded to above by leveraging Cloud Computing organisations lessen the burden on themselves in relation to performing mundane IT maintenance activities. Upgrades are also taken care of by the service provider, for SaaS offerings often up to 4 times a year, thus freeing IT organisations from these sometimes complex tasks and avoiding the situation where an application becomes hopelessly out of date and requires a large and expensive upgrade at some point in the future.

Easier Procurement

Because Cloud Computing solutions do not require major up front capital investments the internal procurement process is made much easier as less approvals and reviews are usually required. This can be especially useful in getting pilot projects off the ground.

Inhibitors

Given the obvious benefits of Cloud Computing described above why isn't everyone vigorously adopting these new technologies. The answer to this question lies in the inhibitors I have listed below. These include:

1. Security & Compliance concerns


2. One size fits all approach


3. Required Mind shift in IT


4. Provider maturity and viability


5. Integration with on-premises applications

Security & Compliance concerns

This is probably the biggest inhibitor to wide scale adoption of cloud computing. Security is a particular concern for many organisations who fear that they will be exposed in a multi-tenant world with data located outside of the corporate firewall. It can be argued that these fears are overblown as the data centers operated by many of the cloud service providers have very robust security features (including SAS70, SOX, HIPPA and PCI compliance) and that the data centers offer security capabilities in excess of those in place in many using organisations.

Regulation presents another problem, particularly in the public sector. In many countries data privacy and other regulations stipulate that certain data (e.g. related to a citizen) may not reside outside the country. This can be a very real inhibitor for many organisations, though vendors are beginning to offer solutions that either guarantee data is located in a particular country or group of countries or which offer a hybrid model which combines the cloud with hosted data solutions.

One Size Fits All Approach

Cloud solutions, particularly SaaS solutions, provided facilities to enable customers to customise their solutions, however these are limited and will usually mean that organisations need to make compromises or change their business processes in order to integrate the solution into their business. While an inhibitor, this does have an upside from a maintainability and cost perspective.

Required Mind shift in IT

This inhibitor is probably the least understood as it is people related. The introduction of cloud computing into an IT organisation introduces change and requires a mind shift from those in the IT organisation. When fully implemented, Cloud Computing moves the focus away from operational activities and more towards visioning, service provisioning and relationship management. This implies changes both for how the organisations is structured and for the individuals within the organisation. Job roles will change, new skills will be required, process will need to change and quite possibly the size of the IT organisation will most likely shrink.

Provider maturity and viability

Cloud Computing is still an emerging and fast moving technology. Cloud Service providers are emerging all the time and even some of the more established vendors are no more than 10 years old. This makes provider viability a concern particularly in a market where consolidation is likely over the coming years. However many established vendors from the traditional IT arena are now entering the market and with careful vendor assessment and due diligence the risk involved can be mitigated.

Integration with on-premises applications

In many cases it will be necessary to integrate on-premises (particularly legacy) applications with applications in the cloud. This involves integrating outside the firewall and over the Internet. There are a number of challenges related to this including interfaces protocols and security particularly where an existing application or system has not been exposed externally before. Ultimately it is unlikely that this will inhibit the adoption of a cloud computing solution, however it does add a degree of complexity.

In Summary

Ultimately each organisation must weigh up the risks versus the benefits of Cloud Computing for their particular organisation. However, it seems to me that Cloud Computing is here to stay and it is more a case of when rather than if organisations adopt these technologies.

In my next blog on Cloud Computing I will look in a bit more depth at Software As A Service and look at some of the sub categories of solution offerings such as CRM, ECM and Digitial Content Creation.

A Brief Guide To Cloud Computing

Cloud computing is the latest technology trend to lack an agreed definition and as usual this is causing much confusion. This is the first in a series of blogs where I explore cloud computing. This first post addresses the question of what it is.

What is the Cloud?

In my experience when people talk about cloud computing they can be talking about anything from hotmail to Microsoft Azure to storage in the cloud. In this post I attempt to bring a degree to clarity to the definition.

I think that the best way to describe a definitionless concept such as the cloud is to think of its characteristics.
With this in mind, the essential characteristics of the cloud are:

1. It is multi-tenant. By this I mean that multiple companies and organisations share the cloud and its capabilities. No one organisation 'owns' a piece of cloud infrastructure or software.




2. It is highly abstracted. Users of the cloud do not know or need to know any of the details of how (or even which and what) piece of the physical cloud (e.g. server) they are using.




3. You pay as you go. Access to cloud services is paid for on a pay as you go basis. Upfront capital costs do not feature.




4. It is flexible and immediately scalable. Cloud services can be scaled up or down at the users discretion and without the need to provide advance notice or incur penalties.

So what is Cloud Computing?

I think the best way to look at computing in cloud is based on a model that divides cloud 'services' into three categories - Infrasturcture As a Service (IAAS), Software As A Service and Platform As a Service (Saas).

Infrastructure As A Service

This category is easily understood as in essence it involves renting and accessing CPU, storage and certain infrastructure related software (e.g. databases) from and through an internet based third party rather than having them located in your data center. The advantage is that IT departments are freed from many of the basic administration functions of these components and costs are contained as in general IaaS costs become operational rather than capital based. Additionally, capacity can be easily scaled up and down as daily, year end or seasonal demands dictate while containing the costs in involved in doing so.

Examples include: RightScale, Amazon, Go Grid, SkyTap.

Software As A Service

Saas is probably the category that is best understood because of the success of salesforce.com. Nevertheless even a cursory google search reveals a broad and every growing set of cloud based offerings in a range of areas including CRM, enterprise content management, billing, business process management and security to name but a few.

One of the great advantages of SaaS is speed to deployment. SaaS solutions can be up and running very quickly at a low cost and at low risk - doing a SaaS based proof of concept is very cost effective as, depending on the project, there are no capital software or hardware costs involved.

Examples include: SalesForce.com, SugarCRM, OpenText, GoogleApps

Platform As A Service

This is probably the hardest category to understand and is probably the least developed. Platform As A Service offerings such as Microsoft Azure and GoogleApps are essentially web application software development platforms in the cloud. Rather than having the development environment and tools in house they are instead hosted in the cloud. Essentially, using these platforms, developers 'program in the cloud'. In order to do this PaaS offerings provide all the features necessary to support the complete lifecycle of building and delivering web applications and services.

I hope this provides a degree of clarity to what cloud computing is. In my next post related to cloud computing I will look at the drivers and inhibitors of cloud computing which should help in bringing more clarity to the topic.

Reflections on Deloitte 2009 CIO Survey

Deloitte Ireland, for whom I work, released the 2009 CIO Survey results late last year. I was responsible for the survey from inception to completion and thought it was worth reflecting on the process.

A couple of things come to mind:

1. It was a really worth while exercise that genuinely provided an insight into the state of IT organisations in Ireland. It wasn't clear at the begining that this would be the case. The results are available here: http://www.deloitte.com/view/en_IE/ie/services/consulting/cio-survey-2009/index.htm



2. It takes a lot of work to run a survey like this, especially in the first year, but I expect that next year will be a lot easier. The biggest effort was required in formulating the survey distribution list and in following up to get responses from those invited. We built a dedicated results website which also took quite a bit of work, though mainly because it was a new concept. Processing the survey results, spotting the trends and creating the summary report was more straightforward than expected.



3. Social media proved a fantastic tool in building momentum around the results and driving traffic to our results website. We used three tools: a. LinkedIn - we used the 'what are you doing' portion of my own and a couple of others in my group profiles' to publicise the fact that the results were coming and when they were released. We also started discussions in relevant LinkedIn discussion groups around the survey results. b. Twitter - we tweeted about the results and their release. c. We submitted a post to the Irish Internet Association's guest blog. This also went out in their weekly email newsletter. These first and last of these items drove large amounts of traffic to our website, with LinkedIn coming first (by a big margin) followed by the IIA blog. Twitter came in 7th and our own direct marketing emails 5th. We also appeared very high up google searches around CIO Surveys. These results created quite a buzz around social media tools in Deloitte Ireland and is a very valuable lesson to all regarding the power of social media tools.



4. We invited people to comment on the results on the website, but few did. Perhaps this is because of the nature of some of the results (budgets etc.) and a certain 'shyness' in the targetted population to publically contribute. I still think that providing this facility was the right thing to do and think we will continue to do so in the years ahead. An interesting thing that I am finding is that people seem more interested in contributing comments to linkedin discussion groups - Deloitte are releasing the 2010 Technology Media and Telecomms predictions soon and I started a discussion inviting people to submit their predictions. This has generated a very good response and we haven't issued the predictions yet!




Though involving hard work, running the survey was a lot of fun and proved rewarding. We will be running the survey again this year. The 2009 results can be found at: http://www.deloitte.com/view/en_IE/ie/services/consulting/cio-survey-2009/index.htm