1: Be honest.
If you plan to reduce headcount and/or restructure the way folk work then say so up front. Explain your reasons why as early as possible and define what the ideal end state looks like. Will people be happy? No, but they know they are dealing with an honest person/company and not a snake.
2: Engage with stakeholders right from the get go.
Engaging with stakeholders unfortunately is often taken to mean talking to managers. Though managers have a critical role to play, its not managers that get the nitty gritty work done in an organization. Whomever is doing the nitty gritty work should be very high on your stakeholder priority list. I could fill a book with stories of IT projects that deliberately avoided engaging with the folk doing the actual work they were going to impact, only to later only sincerely regret doing so. Trust me, when a manager tells you the process his or her charges follow to get work done, the 'charges' version of affairs is almost always very different.
3: Do both of these things way before you chose technology options
One of, if not the most common reason for IT projects to fail is to get involved in procuring (buying) technology way too early. Until you have clearly mapped the 'As Is' situation, mapped the desired 'To Be' situation. Broadcast the need to change, and started to engage and listen to stakeholders etc you should stay well away from technology. A project I was involved with many years ago had raised a budget of several million dollars to tackle a particular problem and buy technology to address the problem. When my colleague and I pointed out on day one of our engagement that the problem was far more simply resolved and could be fixed without any technology...... we were sent home and our contracts cancelled. Fair enough. But way too many big IT projects involving new or replacement technology can be resolved without any technology investments at all.
4: Embrace failure
Its a cliche but any successful person (Richard Branson for an ultra cliché example) will tell you that to get to where they are involved a lot of failure. IT projects are started, change is decided upon - often though a sinking feeling emerges that questions whether this is a good idea or not. By planning and inserting multiple go-no-go milestones in your project plan. By honestly assessing where you are at any point, and whether this project still makes sense you will increase your chances of success ten fold. Embrace failure, walking away early on is far better than undertaking a death march to oblivion. Sadly most projects once started are determined to get to the end, regardless of whether they deliver any value or not.
Clearly there are many more factors to ensure success from building a business case to managing time and resources effectively. But the above four I highlight here as these are fundamental and need to be addressed from the start of any initiative.