When does the shared component stop being cheaper?
A lower fixed charge can lose once transfer cost is included.
The question
Larkspur is comparing two invented network-component models for one stated connectivity need. The component budget is $4,500 per 30-day period. Neither price is a provider quote or proof of technical equivalence.
| Fixed USD per period | USD per decimal GB | |
|---|---|---|
| Shared | 1200 | 0.04 |
| Local | 3200 | 0.01 |
Total = fixed charge + rate × volume for the same period. Technical and operating acceptance remain separate.
“Choose from the actual workload and say what must remain under review.”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Which comparison includes the whole modeled cost?
What would you do here? Open the move you’d make.
Compare only fixed charges
Tempting, but it misses something. This omits transfer cost.
Compare only per-GB rates
Tempting, but it misses something. This omits fixed charges.
Include fixed and variable terms for the same volume
The strong move. The costs now describe the same workload and period.
See how a strong candidate says it
“Compute fixed charge plus per-GB rate times the period’s volume for both options. Neither the lower fixed charge nor the lower rate alone decides the result.”
Why it works: A design choice depends on both cost terms and the workload.
Step 2: At 20,000 GB, what do the two models cost?
What would you do here? Open the move you’d make.
Shared $2,000; local $3,400
The strong move. Both fixed and variable charges are included.
Shared is always cheaper
Tempting, but it misses something. This extrapolates one workload into a universal claim.
Local costs only $3,200
Tempting, but it misses something. This leaves out its variable term.
See how a strong candidate says it
“Shared is $2,000; local is $3,400. Shared is cheaper for this row, not for every future workload.”
Why it works: The result belongs to the displayed traffic volume.
Step 3: Observed volume is now 100,000 GB. What changes?
What would you do here? Open the move you’d make.
Keep the earlier choice because it was cheaper before
Tempting, but it misses something. The workload changed; the earlier result does not decide this row.
Keep reporting $2,000
Tempting, but it misses something. That reuses the lower traffic volume.
Recompute $5,200 shared and $4,200 local
The strong move. The revised calculation exposes the component-budget conflict.
See how a strong candidate says it
“Shared becomes $5,200; local becomes $4,200. Only local fits the $4,500 component budget under this model.”
Why it works: Changed traffic can reverse the cost preference.
Step 4: What closes the recommendation beyond price?
What would you do here? Open the move you’d make.
Choose only by price
Tempting, but it misses something. The technical and operating claims remain unverified.
Verify the claims and assign a recurring cost review
The strong move. This connects the actual design to its customer and operating requirements.
Move every customer to the local option
Tempting, but it misses something. The evidence concerns this workload, not all accounts.
See how a strong candidate says it
“Verify the connectivity, resilience and operating claims. Assign the cost owner and review the volume boundary near 66,666.67 GB. A dashboard alone does not supply that ownership.”
Why it works: A favorable cost model does not establish technical equivalence or future workload.
The answer, in short
- Decision
- The shared model wins at the lower volume; the local model fits the higher-volume budget.
- Because
- $2,000 is below $3,400, but $5,200 exceeds both $4,200 and the $4,500 cap.
- What would change it
- A changed volume, price model, technical requirement or operating result can reopen the choice.
Common traps
- Ignoring fixed or variable charges.
- Using an old volume for a new workload.
- Treating a lower price as technical acceptance.
Check your answer
- Did I include both cost terms?
- Did I use the observed workload?
- Did I check the component cap?
- Did I name the technical evidence and recurring owner?
Last reviewed Nov 10, 2026