A solid analytics foundation is non-negotiable. My background is mainly in Adobe Analytics, but the platform matters far less than the quality of what sits underneath it. If your underlying data is sound and your tracking is up to scratch, you're set up for success. From there, combining quantitative and qualitative insight is key. The numbers tell you what is happening, and the qualitative side (session replays, heatmapping, user testing etc) helps you understand why, so together they give you a rounded view of the full picture the data is presenting.
But the biggest thing for me is having a plan. A solid backlog of tests, underpinned by robust data, that ladders up to the strategic direction of the organisation is what turns a handful of experiments into a successful experimentation programme. It keeps everyone focused on the problems that matter most, and it makes sure every test is contributing to where the business wants to go.