Answer.

Historically, effort estimation was built on expert assessments or analogies with past projects. Under conditions of limited time and information, a systems analyst is forced to work with high-level, vague requirements, often encountering incompleteness and inflated expectations.

Problem: Uncertainty leads to risks of underestimation, conflicts with the client and technical team, and budget overruns. Estimation is very challenging due to changes in initial inputs after the contract is signed.

Solution:

Use requirement decomposition techniques and quick estimation games (planning poker, T-shirt size)
Identify and document all assumptions and constraints
Keep a log of questions and unknown areas
Include risk buffers (uncertainty coefficients/buffers)
Most importantly — fix completion criteria (Done) and boundary conditions

Key features:

Brief artifacts: high-level scenarios, user stories, C4 diagrams
Visualization of main value streams
Focus on prototypes and wireframes to reach agreements faster

Tricky questions.

Is it possible to conduct estimation without risking quality if the requirements have not yet been fully clarified?

No, any estimation at this stage will have to be marked as preliminary with risk and reserve documentation. Otherwise, the responsibility for budget overruns will fall on the executor.

Should only those objects that are explicitly defined by the client be included in the estimation?

No. Everything that is not defined is estimated through a "buffer of uncertainty" or special Story Points for future clarifications; it is important to state: "other requirements are outside the estimation."

Is it necessary for a systems analyst to participate in preparing TCO (total cost of ownership)?

Yes, the analyst formulates the input data - the list of requirements, a list of scenarios, risk areas, limitations, which are critical for accurate TCO calculation.

Common mistakes and anti-patterns

"Flat" project estimation without considering unknown areas
Ignoring possible changes after the implementation starts
Not securing assumptions and constraints in writing

Real-life example

Negative case: The systems analyst accepted the requirements from the manager "as is", quickly estimated without delving into details and not addressing constraints and hidden areas.

Pros:

Quick, convenient for the client

Cons:

Budget overruns, dissatisfaction from the client and team
Refactoring solutions and deadlines are "burning" from the very start

Positive case: The analyst held a working session with key stakeholders, even addressed general requirements, created a map of areas of uncertainty, indicated assumptions, and introduced reserves.