The cost-estimate badge

THE ESTIMATE TELLS YOU:
  Total additional cost for the event
  Per-target breakdown
  Confidence band (e.g., ±15%)
  Whether it's "estimated" or "calculated"
  ROI framing vs business value

EVENT: Black Friday 2026 (24 hours)
SCOPE: 3 production ASGs

COST ESTIMATE:
  Pre-warm cost (24h of event capacity):  $4,800 ±15%
  Scale-down cost (1 hour ramp-back):        $50
  Total additional cost for event:        $4,850

  Baseline cost continues normally:        $X,XXX (unchanged)

  Per-target breakdown:
    asg-checkout (40 instances × 24h):    $2,880
    asg-api (60 instances × 24h):          $1,680
    asg-search (30 instances × 24h):       $240

isEstimated badge: ⚠ YES

1. ACTUAL PEAK DURATION UNKNOWN
   Event might be exactly 24h, or might extend by hours
   Marketing/business may make last-minute adjustments

2. CLOUD RATES CAN CHANGE
   Pricing API gives current rates
   Spot prices fluctuate
   On-demand can change rarely (but possible)

3. PER-INSTANCE COST VARIES
   RI coverage affects effective per-instance cost
   Savings plans cover variable share
   Sustained-use discounts kick in at certain durations

THE ESTIMATE USES current rates and defined duration
ACTUAL COST may vary

COST ESTIMATE: $4,800 ±15%

Lower bound:  $4,080
  (Best case: spot pricing held; event ended on time;
              RI coverage fully utilized)

Upper bound:  $5,520
  (Worst case: rates rose; event extended;
              spot capacity limited)

BAND CALCULATION:
  Recent pricing volatility
  Event duration variance
  Capacity flexibility
  Historical accuracy on similar events

For LARGER EVENTS with more uncertainty: bands widen to ±25%
For ROUTINE EVENTS with historical data: bands tighten to ±10%

PER-TARGET BREAKDOWN:
─────────────────────────────────────────────────────────────────
asg-checkout
  Current:  12 × $4/hr = $48/hr
  Event:    40 × $4/hr × 24h = $3,840
  Additional: $3,840 - ($48 × 24) = $2,688
  Note: assumes $4/hr rate per instance

asg-api
  Current:  20 × $3/hr = $60/hr
  Event:    60 × $3/hr × 24h = $4,320
  Additional: $4,320 - ($60 × 24) = $2,880

asg-search
  Current:  5 × $2/hr = $10/hr
  Event:    30 × $2/hr × 24h = $1,440
  Additional: $1,440 - ($10 × 24) = $1,200

NO BADGE:
  Cost is confidently calculated from pricing data + capacity plans
  Rarely true for new events
  Common for repeated events with history

isEstimated = YES:
  Estimate is best-effort
  Actual cost can vary ±15% (or larger for big events)
  Most realistic for any first-time event
  Customer should plan for upper bound

EVOLUTION:
  First execution: badge always YES
  After first execution: real cost data available
  After multiple executions: confidence builds; band tightens
  Eventually: badge removes; calculation is reliable

EVENT CONFIG CHANGES:
  Capacity multiplier adjusted
  Duration changes
  Scope added/removed

PRICING DATA UPDATES:
  Weekly sync of pricing_cache
  Spot rates change
  RI coverage adjustments

CUSTOMER triggers re-estimate:
  Manual refresh
  After config changes
  Before final commit

EVENT: Black Friday
ESTIMATED EVENT READINESS COST: $4,800
EXPECTED ADDITIONAL REVENUE: $250,000+
  (Based on smooth performance maintaining conversion)

ROI: 50x

THE ESTIMATE IS SMALL relative to the value of avoiding
lost sales from latency spikes

COMMON MISTAKE: present only the cost
  "Event Readiness will cost $4,800. Approve?"
  Finance: "$4,800 is a lot for a one-time scheduling thing."
  Decision delayed; may not approve

BETTER: present cost + value
  "$4,800 cost; projected $250K revenue impact from smooth event"
  "50x ROI"
  "Without this: estimated $25K loss from latency-driven bounces"
  Decision is clear

DOCUMENTATION:
  Save the ROI calculation
  Save the comparison post-event
  Build a track record for future events

AFTER EVENT, reconcile estimate vs actual:

ESTIMATED COST: $4,800 ±15%
ACTUAL COST: $4,920
WITHIN ESTIMATE BAND: yes (within $4,080 - $5,520)

ACCURACY: estimate was within 3% of actual
CONFIDENCE: tightens for next time

OVER MULTIPLE EVENTS:
  Estimate accuracy improves
  Band tightens
  Eventually: isEstimated badge removes
  Calculations become precise

EVENT: Marketing campaign Nov 30 (24 hours)

ZopNight shows in wizard:

  PRE-WARM 1 HOUR BEFORE: minimal cost (~$200)

  EVENT WINDOW (24 hours):
    asg-marketing-api:    25 → 60 instances = $2,160
    asg-search:           10 → 25 instances = $360
    asg-notification:      5 → 15 instances = $180
    TOTAL EVENT WINDOW:   $2,700

  SCALE-DOWN (1 hour ramp-back): ~$30

  GRAND TOTAL:            ~$2,930 ±15%

  isEstimated: YES
  Confidence: 85% (first-time event for this scope)

CUSTOMER REASONING:
  Compare to estimated impact:

  ESTIMATED REVENUE IMPACT FROM AVOIDING LATENCY SPIKES:
  ~$45,000 (conservative: based on marketing's projected conversions
   without latency-related dropoff)

  ROI: $45K / $3K = 15x

  Plus: brand experience preserved
  Plus: customer trust maintained

DECISION: approve Event Readiness for $3K to capture $45K opportunity

POST-EVENT (Dec 1):
  Actual cost: $3,060 (close to estimate)
  Actual revenue impact: $52K (above projection)
  ROI confirmed: 17x

  Lessons:
    Estimate accurate within 5%
    Capacity adequate for actual peak
    No issues during event

  For next event: confidence higher; band can tighten

□ STEP 1: Open Event Readiness; pick an event
  Event: __________
  Cost estimate: $_____
  Confidence band: ±___%

□ STEP 2: Per-target breakdown
  Target 1: __________   Cost: $_____
  Target 2: __________   Cost: $_____
  Target 3: __________   Cost: $_____

□ STEP 3: Estimate vs revenue impact
  Expected revenue if successful: $_____
  Expected loss if latency spikes: $_____
  ROI: _____x

□ STEP 4: Decision
  □ Approve (ROI strong)
  □ Reject (ROI weak; investigate options)
  □ Adjust capacity (cost too high; reduce multiplier)

□ STEP 5: Post-event commit
  When to reconcile: __________
  Owner: __________