R.P. Churchill

Many practitioners identify and define items that go in the backlog. Product owners work with stakeholders to elaborate on and prioritize items within the backlog. Designers and implementors then work the highest priority items.

Items may be prioritized based on many criteria including include business value, stakeholder needs, technical dependencies, available resources, and others.

Items in the backlog may exist at different levels of maturity, granularity, and clarity. The items at the top (or front) of the backlog queue should be clear, defined, granular, and complete so they may be worked on with minimal clarification or additional guidance.

Item tracking is how participants in an effort monitor what concerns, issues, and tasks are valid and need to be addressed, and who has responsibility. Items can arise in any phase of an engagement and be tracked through any other phase.

Elements may be tracked using many methods. Jira and Rally are oft-used shared systems, but many others exist. Punch lists are common near the end of an engagement.

Elements of tracked items may include Identifier, Summary, Category, Type, Date Identified, Identified By (Person), Impact, Priority, Resolution Date, Owner, Resolver, Agreed Strategy, Status, Resolution Update, and Escalation Matrix.

Per the BABOK, "The purpose of the lessons learned process is to compile and document successes, opportunities for improvement, failures, and recommendations for improving the performance of future projects or project phases."

Lessons can be discussed and recorded at any time, but tend to be considered at the end of some increment like a sprint, phase, deliverable, project, or program.

Lessons can be learned about all facets of an environment or engagement.

Reviews happen continuously as part of the iteration within and between phases.

Reviews are used to keep all items in synch as they are tracked in a traceability matrix.

They are similarly employed to align work in larger efforts involving many parallel teams (esp. using different methods as in the SAFe framework).

Acceptance criteria are used to determine whether items meet needs identified by users and organizations.

Acceptance criteria are evaluated through every phase of interaction with the customer and through the process of traceability.

Functional requirements are more likely to be evaluated using discernible go/no-go criteria; non-functional requirements less so.

Evaluation criteria are used to compare many prospective solutions.

If potential solutions are very different and provide different benefits, a hybrid, multiplicative scoring approach may be needed to incorporate individual weightings.

Objective evaluation criteria may not exist, in which case entrepreneurial judgment must be used.

Metrics and KPIs indicate how an operation performs relative to trend or a defined goal. They may be point values (95% uptime), ranges (within 1% of target), probabilities (< 15 min wait, 85% of the time), or derived from other calculations.

Data items may drive decisions but that doesn't make them good metrics or KPIs.

These are usually derived from the operation of the system. They may be be simple (properties & uptime) or complex (aircraft readiness).

Care must be taken to ensure the metric or KPI measure what you think it measures, and provides appropriate and effective guidance (see Goodhart's Law).

Benchmarking involves learning about activities and characteristics across industries, organizations, products, methodologies, and technologies to identify best practices, product options, and competitive requirements.

Market Analysis involves studying customers, competitors, and the market environment to determine what opportunities exist and how to address them.

This is similar to the balanced scorecard technique on the previous slide, but to my mind is geared toward analyzing organizations at a higher level, possibly involving more senior people.

It seeks to evaluate organizations along nine dimensions: Key Partnerships, Key Activities, Key Resources, Value Proposition, Customer Relationships, Channels, Customer Segments, Cost Structure, and Revenue Streams.

Many, many different methods of observation can be employed, and they may be used in almost every phase of an engagement. That said, they are most likely to be used in conceptual modeling and testing.

Potential methods can involve guided and unguided walk-throughs; review of drawings; internet research; collection of real-time and historical data recorded from sensors and other programmed activities; direct visual results recorded using notes, logsheets, checklists, and mobile apps; interviews; surveys; photo and video; review of manuals and specs; document review; calculations; general research; and procedures and policies.

The focus can be qualitative (discovery => nouns and verbs) or qualitative (data collection => adjectives and adverbs).

A concept model, or conceptual model, is an abstract representation of an organization, process, system, or product. It relates the nouns and verbs and other categorizations within and between elements.

Concept models are constructed to represent existing systems and potential future systems. They can represent the As-Is state or the To-Be state in any level of abstraction.

Concept models can represent all or part of a system or solution.

Functional decomposition is similar to brainstorming. A problem should be broken down many ways to identify a clear and effective solution.

Several categorizations may apply: Decomposition Objectives, Subjects of Decomposition, Level of Decomposition, and Representation of Decomposition Results.

Understanding standard components of a problem helps. This comes from specific industry and organization experience and ad hoc analysis.

Think about the effects changes to every element might have (like an equation with many variables).

A mind map is a particular type of diagram used for taking notes, organizing thoughts, and understanding hierarchies. Business analysts can use any type of diagram that aids understanding and communication between participants in any engagement.

Mind maps are sometimes defined as variations of tree diagrams (which themselves can take on many forms), but in practice people call a lot of things mind maps.

I originally thought mind maps were only products generated by certain software programs meant to show associations between concepts in a relatively freeform way. I've seen people take notes using such apps. Some find this inefficient, but use what works!.

This is the processing of large quantities of data to glean insights and support decisions.

Techniques may be descriptive, diagnostic, or predictive.

Top-down methods are used to develop algorithms. Bottom-up methods are used to discover patterns. Supervised methods look for something specific. Unsupervised methods are applied blindly to see what emerges.

The process follows five general steps: Requirements Elicitation, Data Preparation: Analytical Dataset, Data Analysis, Modeling Techniques, and Deployment.