Our straight-talking view on a problem we see every week.
Cloud projects go wrong not because the cloud is hard, but because the planning is skipped.


A cloud move is a project with a beginning, middle and end. Treat it like one and it works. Treat it like an experiment and it will cost you.
Have a question about your own systems? Get in touch — an honest answer costs nothing.
A cloud move is a project, not an experiment. Scope it, plan the rollback, test the restores — and the migration becomes boring in the best possible way.
Move test and non-critical systems first, fix the licence and backup gaps, then migrate your line-of-business systems in a scheduled window with a rollback path. Never migrate everything on a Friday night, and never assume a restore works — schedule one in the first month and prove it. The cloud rewards this kind of sequencing; it punishes guesswork.