Many projects where we start introducing Evo are already running, because often I'm first asked for help when a project is in trouble. We organise an Evo-day to turn the project into an Evo project. The project has already more or less insight in what has to be done, so this can be estimated, prioritised and work-tasks can be defined and planned.
In the case of completely new projects we have the experience that the team members usually hardly know what has te be done and how. At the first Evo-day, the team members cannot define tasks at all, so no estimation and planning seems possible. Still, the goal of any Evo day is that by the end of the day the team knows roughly what to do the coming weeks, and exactly what to do the first week. So there is a potential problem.
This problem can be solved by:
If you ask how much time the team members are going to spend on these activities, the answer is usually: "I don't know", "I don't know what I am going to search for, and what I am going to find and going to decide, so I cannot estimate". This may be true, but should not be used as a licence to freely spend time.
What should you do? "I don't know."
Then what are you going to do tomorrow? "I don't know it yet. Tomorrow I will see".
Try to define important things that could be done and define time boxes, like "Use 4 hours to define what the project actually is about", or "Use 6 hours on Requirements collection". Then put these tasks on the Candidate Task list, define priorities and let everybody take 26 hours of the top of the list, get commitments and that's it.
Still, in some cases the team members cannot set their mind to commit to these tasks. Then, as a last resort, ask the team members to do whatever they want, provided that during the work they record what they are doing and for how long. This is learning material for the next weeks' meeting.
Estimation and tracking (measurement) is a chicken and egg question (What was earlier?). Some people think they cannot estimate and need experience (measurements) before they can estimate. They rather start with the chicken than with the egg. Because most designers are not completely void of experience, this normally is not necessary: let them estimate/guesstimate and then compare the actual results with the estimates. If they refuse to even try to estimate, don't push it too hard or you will lose them altogether. Still, insist on measuring the effort, so that they will learn. Insist, insist, insist! Otherwise they will never learn!