Communicating uncertainty

HONEST COMMUNICATION:
  "Forecast: $162K/mo for Q2, with ±10% confidence band
  ($146K - $178K). Driven by uncertainty in feature launch
  timing and Q2 marketing campaign impact. Our forecasts have
  averaged 97% accurate over the past 4 quarters."

DISHONEST COMMUNICATION:
  "Q2 cost: $162K/mo. Exactly."

1. STATE THE BAND
   "Forecast $X with ±Y%"
   "Most likely: $X, range: $W - $Z"

2. EXPLAIN THE BAND'S DRIVERS
   "Driven by:
    - Feature launch impact (timing uncertain)
    - Growth rate variance (10-15% expected)
    - One-time efficiency project (sustain rate unknown)"

3. PROVIDE SCENARIOS (when audience benefits)
   "Best case: $X - 5% (efficiency exceeds plan)
    Likely:   $X (current trajectory)
    Worst:    $X + 10% (launch overruns)"

4. SHOW HISTORICAL ACCURACY
   "Our forecasts have been within ±5% for the past 4 quarters."
   "Track record suggests 92% confidence band is realistic."

FINANCE wants:
  - A single number for budget setting
  - Plus a band for safety margin
  - Plus history of accuracy
  - Plus the underlying methodology

LEADERSHIP wants:
  - The range for planning
  - Best-case and worst-case for risk planning
  - Driver analysis (what would push us each way)
  - Trust signals (our forecasts are reliable)

ENGINEERING wants:
  - All of the above
  - Plus assumptions in detail
  - Plus the per-team breakdown
  - Plus the validation methodology

MISTAKE                                   FIX
──────────────────────────────────────────────────────────────────
Hiding uncertainty                         State the band. "$162K"
("we said $162K, period")                  alone is dishonest;
                                          "$162K ±10%" is honest.

Too much detail (audiences get lost)       Lead with the headline;
                                          drill into detail if asked

No driver explanation                      Explain what would push
                                          higher or lower; helps
                                          audience plan

Updating without explanation                When forecasts change,
                                          explain why; otherwise
                                          looks like incompetence

Confidence over-stated (false precision)    "$162.37K" is false
                                          precision; "$162K range
                                          $146K-$178K" is calibrated

Hiding bad forecasts                       Variance happens; surface
                                          it honestly with driver
                                          analysis; lessons learned

Confusing forecast with budget              Forecast is projection;
                                          budget is commitment;
                                          they're related but distinct

TIMING                                    BAND
──────────────────────────────────────────────────────────────────
Annual planning (9-12 months out)         ±20-25%
                                          Larger band; many unknowns

Quarterly planning (3-6 months out)        ±15%
                                          Tighter as plans firm

Quarterly commit (start of quarter)        ±10%
                                          Standard quarterly band

Mid-quarter (6 weeks in)                    ±5-8%
                                          Half the period elapsed;
                                          patterns visible

End-of-quarter (last 2 weeks)               ±2-3%
                                          Mostly known; small tail

After quarter close                         actual (no band)

SLIDE: Q2 Forecast (made March 1)
─────────────────────────────────────────────────────────────────

FORECAST: $162K/month for Q2
CONFIDENCE BAND: ±10% ($146K - $178K)

DRIVERS (what makes the actual land in the range):
  + Feature launch (April): timing uncertain
  + MAU growth: assumed 12% Q-o-Q
  + Marketing campaign: committed (in plan)
  - One-time efficiency project: sustain rate uncertain

HISTORICAL ACCURACY: 97% average over last 4 quarters
                    (range 96.1% - 98.8%)

RISKS to monitor:
  Launch could slip 2-4 weeks → cost shifts to Q3
  Growth could exceed expectations → cost up at top of band
  Efficiency sustain rate could exceed model → cost lower

DECISION REQUESTED:
  Budget approval for $165K/mo (forecast + slight buffer
  consistent with past 4 quarters' variance pattern)

─────────────────────────────────────────────────────────────────

CFO: "What will our cloud cost in Q2?"

WEAK ANSWER (false precision):
  "$162,000 per month, exactly. We've calculated it."

WHAT HAPPENS WHEN ACTUAL IS $171K:
  CFO: "You said $162K. We're off by $9K. Why?"
  Trust erodes.

STRONG ANSWER:
  "Forecast: $162K with ±10% band ($146K-$178K).
  Likely drivers of variance: feature launch timing,
  marketing campaign size. Our forecasts have been ±5%
  accurate over the past year. I recommend budgeting at
  $165K-$170K for safety."

WHAT HAPPENS WHEN ACTUAL IS $171K:
  CFO: "Within your band. You called it. What drove it?"
  Trust grows.

"The number you'd want is the most likely scenario: $162K.
The reason for the range is so you can plan for variance.
If Q2 lands at $178K, you've already planned for that ceiling.
If we hadn't given you the band, the variance would look like
a surprise. The range is the planning input."

MOST LEADERSHIP ACCEPTS this framing once explained.

FINANCE QUARTERLY REVIEW (CFO + FinOps Lead + Engineering Leader)
TOPIC: Q2 2026 cost forecast

[ HEADLINE ]
COST FORECAST: $162K/month for Q2 2026
CONFIDENCE BAND: $146K - $178K (±10%)

[ DRIVERS ]
  + Feature launch (April): timing uncertain
  + MAU growth: assumed 12% Q-o-Q
  + Marketing campaign in May: committed
  - One-time efficiency landing in March: sustain TBD

[ HISTORICAL ACCURACY ]
  Q1 2026: 98.1%
  Q4 2025: 96.1%
  Q3 2025: 98.8%
  Q2 2025: 97.5%
  Average: 97.6% (range 96.1-98.8%)

[ ASK ]
  Approve budget at $165K/mo (forecast + small buffer
  consistent with past variance)

[ CFO QUESTION ]
  "Why is the band 10% and not 5%? Can we tighten?"

[ FINOPS RESPONSE ]
  "10% reflects the actual uncertainty at this point: feature
  launch timing is the biggest variable. As we get closer to
  the launch date and that uncertainty resolves, we'll
  re-forecast monthly and the band will tighten to ±5%.
  At quarter-end, the band will be ±2-3%. If you want a
  tighter commit number for budget purposes now, $165K/mo
  is the right anchor."

[ CFO RESPONSE ]
  "Got it. Approved $165K. Send me the next re-forecast
  in 30 days."

[ OUTCOME ]
  Budget committed at $165K
  Re-forecast scheduled for April 1
  CFO understands the methodology
  Trust established

FORECAST PERIOD:    __________

ONE-LINE HEADLINE:
  "$______ per month for __________"

CONFIDENCE BAND:    $______ to $______

KEY DRIVERS (what would push outside the band):
  + __________
  + __________
  - __________

HISTORICAL ACCURACY (your actual track record):
  Last 4 quarters: ____% average
  Or: track record not yet established (acknowledge this honestly)

ASK / DECISION REQUESTED:
  __________

PRACTICED RESPONSE if CFO asks "just give me the number":
  __________________________________________________________