Strategy and responsible systems workshop

Allow 90–120 minutes. Prerequisites: percentages, introductory Python and the linked chapters. All numbers, costs, orders and emission factors are invented.

Competitive strategy: customer → alternatives → activities → constraints → review

Calculate A:100×(1000−600)−20000=20000 and B:80×(1200−700)−25000=15000. Explain why these results do not establish demand, available capacity or full value. Submit a customer problem, substitute table, activity map and sensitivity case. Identify the promise branding communicates and the operations needed to deliver it.

Platforms: participants → interactions → incentives → rules → recovery

100 transactions averaging500 produce gross value50000. Commission10% produces revenue5000. Support20 per transaction and fixed cost2000 leave1000 before excluded costs. Submit a participant map, unit-economics table and dispute/recovery scenario. Run the greedy matcher: R1→S1, R2→unmatched, R3→unmatched. It processes input order and one slot per compatible request; it is not optimal, fair by definition or evidence of real market acceptance. Reorder requests and discuss whose opportunity changes.

AI strategy: purpose → baseline → evaluation → oversight → recovery

TP3,FP2,FN1,TN4 give precision60%, recall75%, accuracy70%. With invented FP loss10 and FN loss50, error loss70. Five flags exceed review capacity4. Submit a baseline comparison, confusion table, review rule and stop/recovery plan. Explain how reviewers can reject output and what happens to unflagged errors. This exercise does not train a model or call an API.

Sustainability: boundary → baseline → options → tradeoffs → review

1000kWh/100jobs=10;900/120=7.5kWh/job. Absolute use falls10%, intensity25%. A rebound case1500/200=7.5 improves intensity but total rises50%. The invented factor0.5kg/kWh gives500,450,750kg respectively; it is not a verified factor for any place or period. Submit a boundary statement, unit dictionary, absolute/intensity table and uncertainty memo. No full emissions inventory or current reporting obligations are established.

Run and test

Save strategy_systems.py and tests together.

python3 strategy_systems.py
python3 -m unittest strategy_systems_test.py

Predict each output first, run it, then change one assumption. Invalid negative quantities, commission above100%, zero completed jobs, repeated request IDs and invalid seller capacity should fail. Write the reason before reading the error. The scripts use exact fractions for arithmetic and preserve the supplied seller capacity.

Department and question-word review

Discuss how marketing, operations, finance, IT, customer support and management must exchange definitions. Who owns the decision; whom does it affect; whose assumptions apply; what is measured; where does the boundary end; when is review due; why is the option suitable; which alternative remains; how does the system recover? Include employee effort, customer experience, local conditions and global alternatives.

Chapters

• mba competitive strategy
• mba digital platforms
• mba ai strategy
• mba sustainability strategy

Free primary further reading

• OpenStax: competition and strategy
• MIT OpenCourseWare: network effects lecture material — historical graduate course material, not enrolment or credit.
• NIST: voluntary AI risk framework — check the current official version and revision notices.
• GHG Protocol: calculation guidance — use verified context-specific factors for real work.

Links inspected 4 October 2026. These resources complement the original lessons rather than certify exhaustive programme coverage.