Practical test effort estimation



The objective of this masterclass is to enable the participants “a scientific estimation ability” to enable correct estimation with available information.

Target audience

Test manager, Project manager, Test lead, Test automation staff.


Estimation is scientific guessing. The keyword is “Scientific”.

Estimation is closer to reality when done close to end of the project. The challenge is in being accurate is with limited information. Estimation requires us to do two things (1) clearly know the various elements of work based on information that is more complete (2) know what questions to ask to identify the various elements of work, when information is not complete or missing.

Hypothesis Based Methodology (HBT) powered by the defect detection technology STEM (STAG Test Engineering Method) is used to estimate effort in a scientific manner.

know more

HBT is a six-stage personal test methodology with the first two stages focused on “Understanding expectations and context” and “Defect hypothesis”. The third stage is about “Devising proof” that includes test strategy formulation and e!ort estimation. This stage is powered by the “Test strategy & planning” discipline of STEM, the underlying defect detection technology that powers HBT. This discipline consist of about ten steps with seven scientific concepts to enable strategy development and planning. The two scientific concepts of “Defect-based activity breakdown” and “Approximation principle” is useful in test e!ort estimation. The discipline of “Tooling” enables estimation of e!ort for automation.

The process of application of this discipline enables the participant to scientifically estimate e!ort, identify aspects that are unclear and pave the way for extracting information that is needed for estimation.



To enable good testing, a priori estimation of effort that is close to the intended is critical. Correct estimation is typically understood as the outcome of a rich experience. In short this is good guesswork. The perennial question is “Is there a way to scientifically guess?”. Can there be a set of techniques, principles, guidelines that can allow us to rationally arrive at good estimates? Is there a science of estimation?

Effort estimation is about identifying the various activities in test lifecycle and then estimating the effort for each of the activities and then summing them. The first step of detailing the various activities can be done precisely if we have a clear goal-focused approach to validation. Goal focused validation means

know more

(1)what defects are we planning to uncover (2) what types of test should we perform (3)how many stages of validation do we need to perform (4) how many cycles do we anticipate (5) how many test scenarios do we anticipate (6) how many cycles should we plan for (7) how do we plan to execute these tests (8) what test infrastructure do we need to be develop (9) what scenarios do we plan to automate and (10) what competencies do we have.

This workshop takes a scientific approach to “guessing” the effort. This is done by identifying the potential defect types and then identifying the various activities for uncovering these defects across all the test entities. The workshop takes a pragmatic approach by detailing a clean activity breakdown approach based on potential defect types. This approach allows us to understand clearly aspects of estimation that are fuzzy, enabling us to question better to arrive at the correct estimates

Program content

  • Elements of test effort estimation
  • Basis of effort estimation
  • The laundry list of various activities for estimating effort
  • Science of approximations
  • Defect centered activity breakdown
  • Identifying the types of tests involved
  • How to estimate the number of scenarios
  • How to estimate effort for a scenario
  • Aspects of automation in estimating effort
  • Volatility and its impact on estimation
  • Historical data and its value in estimation
  • Understanding variation, and how to tame this demon