Estimation of Test Efforts:

“Size, Complexity (Domain/Technical), Priority/Severity, Time (Preparation/Research), Cost, Risk” these are the factors which always directly affects to the Estimation, then let it be development or testing effort estimation.

This is Bottom Up approach for estimations.

Calculate the Size of the Test Project, currently we use a Function/Use Case Point technique for estimations. Now what is FP/UCP technique? In FP/UCP technique we consider the above mentioned attributes/parameters

1. Complexity: (Measured by your previous experience and understanding of the system) for e.g. Simple, Average or Complex.

It depends on the following points

- Knowledge of the System/Domain (rate resource in between 1-5)

- No. of Use Cases (Considering HLD and LLD is ready)

- Number of steps to write and execute and the verification points. (>5 Simple, >10 Average, <15 Complex)

- Navigation points throughout the system while executing. (>5 Simple, >10 Average, <15 Complex)

- Complex Use Case would be with High test case steps and navigation points with high understanding complexity.

2. Priority/Severity: (Decide up on understanding of the System and Domain) rate as High-15, Medium-10, Low-5 and give the points accordingly.

3. Time: Now Complexity, Priority and Severity are directly proportional to Time. If time increases accordingly the Complexity and Priority and Severity increases. We will need to consider the time required for the R&D which is intangible (bit tuff) but need to do so with prior experiences.

Can rate as Max-15, Medium-10, Min-5 and give the points accordingly.

4. Cost: Again Cost is also directly proportional to the Time.

5. Risk: This is a very important and critical factor of the effort estimation. We need to identify the tangible and intangible risks involved in the project and accordingly keep the preventive action plan.

- As we can’t find/see all the risks involved into a project so we need to identify accordingly with prior experience and proper forecasting with your understanding and knowledge of the system/project.

- We will need to consider the availability of the resources (Human/Technical).

- Always keep a buffer to execute plan B (Alternate plan) if plan A fails.

- We can rate risk as High-15, Medium-10, Low-5 and give the points accordingly.

So now we can include specific parameters for the estimation according to the requirement and nature of the project and assign points accordingly.

Put in all these parameters in a excel sheet and with a suitable format which would be easy and understandable and rate(assign) points for each parameter as per the use cases. Calculate the total no. of points and decide upon Points:Time ratio per resource as per the understanding of the points assignment for the parameters.

We can decide upon the Points:Time ratio as per the skills/capabilities of resources. We consider here (pts:hrs) 10:2 for writing the test cases. And (pts:hrs) 10:1 for executing the test cases and 0.5 hrs for reporting purpose per person.

In addition we will need to consider the Deliverables preparation and Reporting time required and at the same time we will need to consider the issue resolution or discussion time required for the issues too.

E.g.

Use Case-1: Complexity + Priority/Severity + Time + Risk = Total points

Complex + High + Max+ High = Calculated pts

i.e. 15+15+15+15=45 Points

So time required for the Use Case one to test would be:

Time Required for Test (TRT) = Time Required for Writing the Test Case + Time Required for Execute the Test Case + Reporting time

***Time Required TC Writing (TRW) = [Calculated pts*2(hrs)]/10(pts)

i.e. TRW = [45*2]/10 = 90/10 = 9 hrs for writing the test case

***Time Required TC Execution (TRE) = [Calculated pts*1(hrs)]/10(pts)

i.e. TRE = [45*1]/10 = 45/10 = 4.5 hrs for writing the test case

***Time Required (TC Reporting)= 0.5 hrs

So,

***TRT = 9+4.5+0.5 = 14.0 hrs per person.

So accordingly we can calculate the Points for all the scenarios or use cases and calculate the exact time required for the Testing where in you will need to add up time for the intangible Risks and finally calculate the final effort estimation.

Note: Your prior experience play a major role in the effort calculation.

This method I have designed through my experiences and leanings and currently we are executing it successfully in projects.

and this methodology is Open ended (just not restricted to the SDLC/Agile and Dev/Test). Why because we can create a template according to the needs (needs- I mean SDLC/Agile/Dev/Test) put these parameters use your previous experience to design the points/parameter. And we can use it for the Dev cycle, Test Cycle in SDLC, we can use it for the Sprints in Agile or for a Modules in JAD/RAD. Depends how you customize it!

To answerer the question “how long does it typically take for you to estimate a given sprint with this formula?” It depends on the size J

E.g.

Suppose I have 20 Use-cases/functionalities to be developed and test in Sprint in this case it takes 20-30 mins to have the estimation J Coz I don’t need to do anything other than ratings and then calculations. You can try it out J

If anyone feels or have any suggestion to improve this methodology please let me and everyone know which will be very helpful. It may possible that I am using a wrong technique and putting my extra money/time/efforts.

Please suggest if I can improve in the methodology or missing any important parameter for the effort estimation.

Awaiting the responses!!!

-- Prashant.