10 key takeaways from a successful public cloud implementation

During the past months I had the great pleasure to work with Mr. Ettore Spigno in my role of BMC executive sponsor for Telecom Italia. Mr. Spigno is  Senior Vice President for the Mercato division of Telecom Italia,  providing information and communication services to small, medium and large enterprises. The main focus of Mr. Spigno and his team is the “Nuvola Italiana” (Italian Cloud) project, the most relevant, business-oriented, public cloud project currently under way in Italy. The goal of this project is to transform the way enterprises consume services, leveraging the cloud computing paradigm. According to Mr. Spigno, this project represented a huge transformational initiative for Telecom Italia, radically changing the way services are delivered and supported from beginning to end. Achieving this transformation required innovations spanning not just network redesign but also organisational and process aspects.

In various occasions, during the cloud design and implementation, Mr. Spigno shared with me some of the strategic guidelines and key takeaways from the “Nuvola Italiana” strategic cloud project. I strongly believe that his experience should be shared with the rest of the industry.

Here is my summary:

1. To ensure success in cloud, it is preferable to start from a green-field situation. The reason is that this helps you focus on important things instead of spending precious time and money to repurpose a ton of legacy infrastructure and processes. 

2. Innovation cannot be only at the technology level; you will also need to allow for organisational changes. At Telecom Italia, they decided to create entirely new roles so as to run the new Cloud Org as a separate business unit that is not burdened by legacy issues.

3. Decide carefully what to put in the cloud. As a business they serve multiple market segments. Each segment has specific needs, spending capabilities and service levels, so the offering must also be carefully segmented to ensure both profitability and efficiency.

4. Focus on automation at all levels. Only high degrees of automation can ensure success in the cloud, while also ensuring that cloud project will be profitable. Not automating is not an option.

5. Don’t discount the network. The network is crucial in the cloud. Re-architect your network to make it reliable and flexible for automation, workload mobility, risk mitigation, security and compliance. This is not an easy job, but given their telco background, they were able to put in place a very satisfactory design that opened the door to efficient and effective cloud-based delivery.

6. Make clever use of your internal resources (humans in primis). Companies like Telecom Italia have a tremendous workforce that may have been under-utilised in the past because of time and budget pressures. In turn, these pressures forced them to leverage armies of external system integrators and consultants. For this project they decided to go for a smart sourcing approach, combining the best vendors and system integrators on the market with their internal people. In mature organisations like Telecom Italia, if you rely too much on external resources, you will end up paying your workforce anyway, and then paying for consultants on top of that. The use of system integrators will create a dependency relationship that is very difficult to break. Instead, involving internal people in high-profile projects like this will have a huge impact on the morale of people, contributing to the success of the overall cloud initiative.

7. People need to break their mental silos. Cloud delivery is “end to end”. You need people who think “services” rather than the discrete task in front of them. This might mean something as seemingly trivial as datacenter physical equipment operators, who need to think twice before unplugging even the most irrelevant cable, all the way to the people in the control room. All these people need to have a clear understanding  of the kind of services they are delivering. It can¹t be like in the past any more, with everything in separate buckets. In the cloud, application, virtualisation, storage, network, risk and compliance are all blended together. Pointing fingers will not help in this new situation.

8. Establish an architecture (engineering group) that “thinks” in terms of both technology and business. Choosing the right architectural solution for the segments you are serving is crucial, and it is a decision that will impact much more than just technology.

9. Make capacity management a high priority for the organisation. Idle inventory can kill profitability for a service provider. On the other hand, the risk of under-provisioning must also be avoided.

10. Choose the right technology partner(s) and be very demanding with them. Try to customise solutions as little as possible. You will always need an easy upgrade path because system-wide downtime is never an option. Savvy platform providers will work with you to provide more out of the box. Help them understand what you need next and work with them on your joint roadmap from the early stages of the project, sharing feedback and hints

These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.

Share This Post



CTO Performance and Availability at BMC Software. In my previous life I was CEO and co-founder of Neptuny, a provider of capacity management solutions acquired by BMC.