Now my brain is less fuddled due to not being in the heat quite so much, I can explain my theory on estimating project tasks.
Start with the basics. What needs to happen? How long will each task take? The answers to these questions will allow you to pull together a timeline.
Try to use estimates gleaned from the people who will actually be doing the work. If you cannot get hold of them, or someone with a similar profile or background to them, make an educated, conservative guess. Find out if a similar project has been done before. Can you speak to the manager? How long did that piece of work take to complete?
Be realistic about what tasks need to be done and the length of time each task will take. Having a realistic timeline – your plan – is critical to whether your project is seen at the end to be a success.
A word of warning: it’s a fine balance between spending too long planning and estimating at this stage and meeting the expectations of your stakeholders: you can’t keep them waiting forever. Aim for a best estimate end date, and tell them that’s all it is until you have a chance to plan in more detail. The more accurate you can be now the easier it will be later, as the finish date should not move too much. Don’t get sucked into the trap of promising an end date before you have really worked out the tasks involved, but if you get caught in a corridor and are asked when your project will finish, make something up with plenty of contingency. It’s always easier to manage the communication around a project that delivers earlier than expected than one that delivers late!