Estimating a project by time when it is not clear what the project is?

Question

Project management

Good day.
I work as a user interface designer and I had a problem and a question at the same time - how to estimate time.
When there is a specification (TOR), the developer can apply a bunch of methods for evaluation. It can be some formulas, or maybe planning poker, but one thing is always clear - it has specifics. That is, he sees the task, it is scheduled for him.
But in my case, I start a project when the customer himself does not know what will be there, what he wants, etc., but requires an assessment of the end of the analytics (prototypes + specifications). I can give a rough estimate, but more often it happens that either I miss (giving more than I needed), or the project grows, becomes large and I do not have time. Accordingly, the customer is indignant, the boss is indignant, and I work on weekends and 20 hours a day.
So the question is - how to calculate the time for this stage, if it is difficult to understand what will happen and how? Maybe change / introduce some practice?
Thanks in advance for your replies.

Answer 1

There are such customers. As mentioned above, without the full amount of information, you will not be able to give a normal estimate. However, there is a way out of this situation. Surely there are some basic implementation requirements that will definitely need to be done? Give an estimate for this piece of work, set the correction by 2.5 times and make a note that this estimate is only for a specific piece of work for which you have information. Further, we will simply make edits as the task arrives. To a question like “I need a full figure”, answer “half a year and we will significantly reduce as information becomes available” :)
According to your own situation - advice, multiply your estimate by 2.5 times. If the authorities ask why so many? Answer that you give the numbers for which you will have time to do this or that work, and not what they want to hear. From your work for 20 hours a day and weekends, no one is better off. You will burn out in six months or a year of such work, it has been verified from personal experience.
If something is not taken into account, warn the manager / client immediately on the situation, do not postpone the conversation for later. Don't be afraid to say you need more time. Believe me, it is much worse not to complete the work in the time promised to the client than to lay down more time, taking into account the risks.
When planning, consider risks. There is a task, divided into 2 stages. The first stage can be done in 1 day if everything goes well, but here and there difficulties may arise, which may take another 1-2 days to solve. But the second point is unclear. Maybe 1 day, maybe 2 days, depending on the success of the first item. But let's take here 3 days + 1 for our own testing / verification of the implementation of the business requirement.
It is this detailed estimate that will allow you to avoid many problems and switch to a normal work schedule. The client will understand why you need more time (or may need it) and you will have peace of mind knowing that you still have time to spare.

Answer 2

It is impossible to estimate how long such work will take, just as it is impossible to estimate how long it will take a scientist to make an invention or develop, for example, a medicine.

Answer 3

take the most pessimistic assessment :)

Answer 4

Very interesting, but you also set the price from the ceiling, not knowing what exactly will have to be done?
No task - no time to complete it, no cost. And point.
First, TK, at least a brief, then everything else. The project is growing - write additions to the TOR with your own deadlines and a separate estimate.