Until jobs are well scoped, defined, and allocated to available resources, and patterns are established for how work will flow through work groups, estimates of how long it will take to complete that work are often quite unreliable. Once these disciplines have been put in place, teams and customers can begin to plan their work more predictably, but a number of iterations performed by a stable team will still be required before you can realistically forecast dates for future deliveries of complex products.

I personally favor evidence-based schedules. If you can't use this specific tool immediately, you can employ these techniques in Excel. Priorities can also be developed using similarly simple techniques. Don't try to use a more complex tool like Microsoft Project until its strengths (dependency analysis and resource loading) are really required. Using such a tool effectively takes training and experience, and if you try to use it when it's not really necessary, you'll likely to be very frustrated, and may lose an important connection between people's individual commitments and the outcomes that must flow out the other end.