Technologists will always seek to improve infrastructure. This is a good thing, but is often separate from the program you are trying to deliver.

Push your due diligence to uncover where things will go wrong. Because they will, and people won’t really come out and share details unless you show you really want them.

Devoting what feels like too much time to release management processes is a good thing. Decreased risk will pay you back five-fold, and will keep old problems from becoming new again.

Your ability to link accounts across businesses is only as good as the underlying client data; much of that was input by hand 25 years ago, by the 10 different banks you were back then.

Business case development should always be thorough. But it also needs to be fast. The time you take at the beginning of the project delays delivery just as much as events at the other end.

Expert teams with great relationships still misinterpret critical communications. Re-clarifying important concepts is critical at every stage.