Decision Table Testing — Handling Complex Business Rules Systematically

Some features are driven by simple input ranges that EP and BVA handle well. But many real-world features involve combinations of conditions — if the user is a premium member and the order is over $100 and the coupon is valid, then apply a 20% discount; otherwise apply 10%; unless the item is already on sale, in which case no additional discount applies. Decision table testing is the technique designed for exactly this kind of multi-condition business logic, ensuring every combination is tested and no rule is missed.

Decision Tables — Systematic Combinatorial Testing #

A decision table lists all possible combinations of conditions (inputs) and the corresponding actions (outputs) in a grid format. Each column is a rule — a unique combination of conditions that produces a specific action. Each rule becomes a test case.

# Decision Table: Free shipping eligibility
# Conditions:
#   C1: Order total >= $50?
#   C2: Customer is a loyalty member?
#   C3: Shipping destination is domestic?
# Actions:
#   A1: Free standard shipping
#   A2: Discounted shipping ($3)
#   A3: Standard shipping rate ($8)

# Build the decision table
conditions = ["Order >= $50", "Loyalty member", "Domestic shipping"]
actions = ["Free shipping", "Discounted ($3)", "Standard ($8)"]

# Each rule is a combination of condition values → action
rules = [
    # Rule 1: All conditions true
    {"C1": True,  "C2": True,  "C3": True,  "action": "Free shipping",
     "desc": "Big order + loyalty + domestic = free shipping"},
    # Rule 2: Big order + loyalty + international
    {"C1": True,  "C2": True,  "C3": False, "action": "Discounted ($3)",
     "desc": "Big order + loyalty + international = discounted"},
    # Rule 3: Big order + non-member + domestic
    {"C1": True,  "C2": False, "C3": True,  "action": "Free shipping",
     "desc": "Big order + domestic = free (even non-member)"},
    # Rule 4: Big order + non-member + international
    {"C1": True,  "C2": False, "C3": False, "action": "Standard ($8)",
     "desc": "Big order but international non-member = standard rate"},
    # Rule 5: Small order + loyalty + domestic
    {"C1": False, "C2": True,  "C3": True,  "action": "Discounted ($3)",
     "desc": "Loyalty member domestic = discounted regardless of amount"},
    # Rule 6: Small order + loyalty + international
    {"C1": False, "C2": True,  "C3": False, "action": "Standard ($8)",
     "desc": "Loyalty but international small order = standard rate"},
    # Rule 7: Small order + non-member + domestic
    {"C1": False, "C2": False, "C3": True,  "action": "Standard ($8)",
     "desc": "No qualifiers = standard rate"},
    # Rule 8: All conditions false
    {"C1": False, "C2": False, "C3": False, "action": "Standard ($8)",
     "desc": "Nothing qualifies = standard rate"},
]

# Display the decision table
print("Decision Table — Free Shipping Eligibility")
print("=" * 75)
header = f"{'Condition':<22}"
for i in range(len(rules)):
    header += f" R{i+1:>2}"
print(header)
print("-" * 75)

for c_name in conditions:
    row = f"{c_name:<22}"
    for rule in rules:
        key = {"Order >= $50": "C1", "Loyalty member": "C2",
               "Domestic shipping": "C3"}[c_name]
        row += f" {'Y':>3}" if rule[key] else f" {'N':>3}"
    print(row)

print("-" * 75)
row = f"{'→ Action':<22}"
for rule in rules:
    label = {"Free shipping": "Free", "Discounted ($3)": "$3", "Standard ($8)": "$8"}
    row += f" {label[rule['action']]:>3}"
print(row)

print(f"\n3 conditions × 2 values each = {2**3} rules = {len(rules)} test cases")
print(f"Each rule becomes one test case with specific test data.")

Common Mistakes #

Mistake 1 — Only testing the “interesting” rules and skipping defaults #

❌ Wrong: Testing Rules 1-3 (the free shipping and discounted cases) and skipping Rules 7-8 (standard rate defaults).

✅ Correct: Testing every rule because each represents a unique combination that the code must handle. Default cases are where “fall-through” bugs hide — the code handles special cases but crashes or produces wrong results for the ordinary path.

Mistake 2 — Not validating the decision table with the product owner #

❌ Wrong: Building the decision table from the requirements document alone and assuming every rule is correct.

✅ Correct: Reviewing the completed decision table with the product owner to confirm that every rule’s action is the intended business behaviour. This review frequently uncovers undocumented rules and corrects misinterpretations.