Designing Enterprise Decisions
Research contact: Victor Tang (Graduate Student), Massachusetts Institute of Technology
Research Advisor, Prof. Warren P. Seering Mechanical Engineering and Engineering Systems Division
victang@mit.edu MIT E60-256, Center for Innovation in Product Development, 30 Memorial Drive, Cambridge, MA 02142
Research goals
Enable executives to make high quality decisions by effective exploration of the decision and solution spaces using
engineering methods …
particularly, design of experiments (DOE)
A model of a real company, ADI
new product development
R&D spending
stock market market potential and market share
financial stress competitors management accounting manufacturing pricing product price costs competitors’ price unit cost
costs and expenses
market value of the firm
Total Quality Management new products shipments orders
industry growth balance sheet, I&E, flow of funds yield prod. price financial accounting % contribution 878.1 906.6 603.3 761.9 output Predictions best factor-levels 1,3,1,3 In best environment In worst environment BAU output best worst BAU 715.8 862.6 566.9 717.9 1 2 3 3 829.9 969.9 670.1 849.6 3 1 2 3 680.2 813.3 532.9 694.4 2 3 1 3 812.2 952.7 653.5 830.5 2 1 3 2 650.4 748.3 511.5 691.5 1 3 2 2 776.6 927.4 613.1 789.3 3 2 1 2 805.5 954.1 641.9 820.6 3 3 3 1 767.8 912.5 607.3 783.7 2 2 2 1 708.0 852.1 553.0 718.9 1 1 1 1 orders = compt = indust = pr ice C O G S yie ld R & D environment price 59.5% COGS 24% yield 14.8% R&D 1.7%
9 experiments, 27 points predictions
Configuration of controllable variables
can favorably impact a firm’s performance,
under good and bad environments
source: experiment L27 including full factorial Indicated Market Value of the firm
1.215 B 1.091 B 967.22 M 843.13 M 719.03 M 594.94 M 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 2 4 6 8 10 12 14 16 18 20 22 24 Time (Month)
Indicated Market Value of the firm : 110L81-61 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
Indicated Market Value of the firm : 110L81-41 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Indicated Market Value of the firm : 110L81-21 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
best BAU worst output best worst BAU 848.4 982.6 689.8 872.7 3 1 3 3 864.4 989.9 693.2 910.0 3 1 3 2 817.2 965.7 650.7 835.2 3 1 1 1 870.1 1003. 707.5 899.8 3 1 3 1 781.5 931.3 617.3 795.9 3 3 2 1 812.5 960.7 647.7 829.1 3 2 2 1 736.1 885.8 581.8 740. 8 1 1 2 1 842.3 984.0 677.6 865.1 3 1 2 1 799.1 944.0 937.4 815.7 2 1 2 1 orders = compt = ind = p ric e C O G S y ie ld R & D environment
9-step hill-climbing: a very effective process
Our study suggests that this engineering method for designing decisions can be applied to enterprise decisions.
Sparcity, hierarchy, and inheritance all properties of complex engineering
systems - are also exhibited by an enterprise – a socio-technical system.
Summary of key findings
Case studies underway
How to optimize client satisfaction for a risky Web-based development project for a global manufacturing company?
How to raise profit level by $xx M in the next six months for a global electronics outsourcing company?
Decisions … decisions
What can I control? How?
By how much?
What cannot be controlled? What can I do about that? Can I limit my down-side risks?
The problem in DOE normal form
uncontrollable variables controllable variables outcomes problem industry growth competitors’ attractiveness R&D IC yield firm value stock price gross profit
threat of hostile take-over
SG&A
op. income
ADI orders
COGS