In the first two blog articles of our series from Konrad Petrusewicz, we started by covering the fundamental challenge of building a business case for migrating from traditional data centres to the Cloud and how this could be resolved by getting a clear understanding of an enterprise with all its complexity. We then addressed how to plan and build a unified Cloud Migration Pipeline and tooling strategy.
Even once you’ve dealt with the challenges covered in the first two parts of this article, there’s still a large part of the brick wall remaining, and it’s cultural. It’s easy to talk about creating Cloud teams, but hard to do in a flexible, consumable way; skills are short, and the centralised vision and political influence is often lacking in large enterprises. Organisations can be very resistant to the kind of structural change that will help to break those barriers, as they are formed of large, resilient silos.
These silos were sometimes created with good reason; in an IT Infrastructure Library (ITIL) world, it's best to ensure you have all the skills to solve a particular set of technical challenges on a given technology in one place. Most non-critical problems have a single root cause, so teams don't have to spend a huge amount of time and effort liaising with one another through anything more than Change Advisory Board meetings and ServiceNow tickets.
Cloud (both on-premise and Public) development differs by requiring a great deal of technical co-operation across multiple disciplines. These communication challenges are often heightened by the outsourcing of IT departments over the last 15 years. A central Cloud team can often help to fix this, but their priority to bring as many applications onto the Cloud as possible doesn’t always align with broader organisational priorities. More bricks, a bigger wall.
Cloud offers the opportunity for teams to break out of these historical silos and start to be fully cross-disciplinary, bringing greater efficiency and understanding. “Two-speed” IT, whereby some parts of the estate continue to operate as before and others transform, should be carefully managed to reduce friction between teams and dependency-management complexity.
Additionally, there’s a cultural shift in how teams run their systems. Historically it was fine to run a server 24/7 – after all, individual teams never had to pay the electricity bill. With public cloud, those costs should be made visible to all, and with the correct cultural and technical changes, a technical team should be able to gracefully switch off services when not needed, even if they are legacy and commercial off-the-shelf (COTS) applications, and especially for developing and testing work. Don’t forget that running a server for more than 70% of the time on Amazon Web Services (AWS) can be more expensive than running it on-premise, destroying your business case. Amazon doesn’t make its money selling books, as November’s Keynote at AWS Re:Invent made clear.
How can you break through the oncoming brick wall? Each of the problems in our 3-part series has a solution, but whilst they are easy to write they are much harder to implement. Having a clear, unbiased understanding of your estate, with its huge complexity, is vital. Having a clearer understanding of the organisation as it currently stands, warts and all, with a realistic vision of how it can be transformed is even more vital. An achievable strategy, with realism and pragmatism at its heart, will go a long way to move a firm's culture to a more cloud-friendly operating model. These are all hard for organisations to do, as they generally don’t spend their time constantly transforming – but it’s what we do. We spend our days solving these challenges for leading companies and public sector organisations.
By optimising existing Cloud deployments, we typically save our clients 30% on their cloud spend. We have years of experience helping our clients go on the cultural transformation journey to the Cloud and a great track record in Agile transformation. We work together with our clients to make huge changes to their operating model, defining and transitioning to a successful target state.