Agile Estimation

Characteristics :

1. Relative : Where items are compared against each other or are grouped together
2. Collaborative : Estimate independently and come to consensus later
3. Faster : Faster than other traditional techniques

Three aspects of any project management efforts are :

1. Scope
2. Effort
3. Time

Difference from Waterfall

1. Waterfall is a bottom up approach. Agile is a top down approach.
2. Waterfall is plan driven. Agile is value driven.

How often to estimate

1. Bid level
   • Arrive at a ballpark figure based on availability of features or epics available at the time
   • High Level Vision may be able to help us get to it
   • Epic/Feature level clarity may be lacking at this stage
2. Program Increment level(PI)
   • Main objective is to plan for the product features based on dependencies
   • Estimation focus is on the size of work which helps in identifying capacity/team size
   • Adoption of T-Shirt sizing or story point estimates depend on clarity of features
3. Iteration
   • Detailed planning based on decomposed user stories from backlog
   • Planning poker estimation is predominantly used
4. Retrospective
   • for objective and data-driven retrospection of efforts actual values of all estimations may be compared with the estimated values at the start of the sprint

Planning Poker

Teammates use the Fibonacci or exponential series to assign a story point. If there is a large gap, the team members with the largest gap are asked to present their opinions, till a consensus is reached.

A high estimate means that the story point is not well understood, and may need to be further broken down into smaller units.

T-Shirt sizing

XS, S, M, L, XL are used. There can be any number of sizes used.

A relative marker and it's value is chosen(usually M) after discussion and agreement with team members or estimators. Then the size is assigned to each backlog items relative to the size assigned to medium.

Dot Voting

When there are a small number of items, team members are assigned a limited number of votes(dots) that they assign to items. The more number of dots an items receives, the more work it requires.

Affinity Mapping

Team collaboratively arranges backlog items relative to each other in a measure of size of effort. The buckets are then associated with a numerical value desired.

Bucket System

Team arranges items in buckets of different sizes based on complexity and effort. A numerical estimate is placed on each of the bucket based on that.

Estimations can be based on Story points or Functional Points

Story Point based Comprehensive Estimation(SPACE) technique

Has 3 parts, Input, Process and Output.

Requires 3 inputs :

1. Requirements(FR and NFR)
2. Technology
3. Team Sizing

Processing involves picking a story, assigning a probable story point, select parameters and respective values, get story points from SPACE, review comfortable, likely and stringent sizes suggested by SPACE, converge on size based on discussions and review.

Outputs :

1. Size
2. Epic Identification
3. Effort based on Tech
4. Capacity utilization
5. Velocity

Other estimation methods and their factors are :

1. Cloud Apps
   • Service-Service communication
   • Number of services
   • DB impact
   • Configuration
   • Integration
   • Security
2. RPA
   • input type
   • Total number of clicks required
   • Type of development
   • involvement of Citrix/OCR
   • Log and reporting
3. SAP
   • Solution and Strategy
   • Architectural Change
   • Configuration
   • Inter modular integration
   • global process impact
   • cutover impact
4. SFDC
   • Configuration
   • Customization
   • Standalone deployment
   • Automated deployment
5. Mobility
   • Mix of platforms
   • Ui Complexity
   • Processing
   • Connectivity
   • Middleware
   • Special Requirement
   • Online/Offline
6. IoT
   • Edge Connectivity
   • IoT protocol adapter
   • platform device management
   • authentication
   • data ingestion
   • alerts/notifications
   • smart app interfaces