Please enable JavaScript.
Coggle requires JavaScript to display documents.
Business Analytics Cycle 1 - Coggle Diagram
Business Analytics Cycle 1
"The forecast is always wrong"
Driver tree
- which KPDs matter?
Sensitivity analysis:
Change 1 KPD at a time & show its impact on KPI in data table & tornado diagram
How sensitive is profit to demand at 3 different production levels?
Rule of ranges:
use ranges for KPIs when KPDs are uncertain; reflect uncertainty in width of range
Rule of sensitivity:
same level of confidence across all KPDs
Managing uncertainty
Diversification
Averages are more useful to a CEO (diversified products) than a product manager (concentration around 1 product)
Uncertain number can be represented by an average if it is diversified
Flaw of averages
- probabilities
Don't pull averages from a 50/50 chance
Scholtes Revenue Fallacy:
AR ≠ Avg Price x Avg Sales Volume
Flaw of averages
- average demand isn't helpful with D2D variability
Options/flexibility
Reduces risk of downside & unlocks upside opportunity - need to know at what cost
Probability distributions
Demand distribution
Cumulative demand distribution
Stages of problem solving
What's the problem? (problem solving is problem articulation)
What are the key choices & uncertainties? (assess sensitivity of KPI to ranges of KPDs)
Build a model to analyse the issue (consider probabilities with distribution charts)
Validate & test outcomes with others (explore & determine level of risk)
Try again
Thrashing
is normal - creating useful models requires a lot of modelling effort
Risk of being wrong - 1) bad data or 2) bad model (
worse
)
