# Worksheet: compare two market-size estimates
New Eruditions • reusable reader template

**Question or decision:** ____________________  
**Research date:** ____________________  
**Researcher / editor:** ____________________

## Record the claims before explaining them

| Field | Estimate A | Estimate B | Same / different / unknown | Evidence URL + passage |
|---|---|---|---|---|
| Original source, author and title | | | | |
| Publication / update date and data vintage | | | | |
| Printed number and units | | | | |
| Measurement period / calendar mapping | | | | |
| Geography and customer population | | | | |
| Market / product definition | | | | |
| Included services | | | | |
| Explicit exclusions | | | | |
| SaaS versus underlying infrastructure | | | | |
| Public / hosted private / customer-owned infrastructure | | | | |
| Managed and professional-service fees | | | | |
| Vendors and eligible business lines | | | | |
| Customer spend / supplier revenue / gross-net rules | | | | |
| Currency, FX translation and inflation basis | | | | |
| Growth rate and whether its basis matches the level | | | | |
| Quarterly / annual / annualized / forecast | | | | |
| Source data, estimation method and uncertainty | | | | |
| Revisions, coverage changes and rounding | | | | |
| Vendor-share denominator and scope | | | | |
| Upstream origin of downstream references | | | | |

Write **“not disclosed”** where applicable. A blank field is unfinished work; “not disclosed in inspected material” is an evidence finding. Absence from a description is not a documented exclusion.

## Calculate only after checking comparability

| Calculation | Formula | Your inputs / result | Assumptions and rounding |
|---|---|---|---|
| Absolute difference | A − B | | |
| Relative to B | (A − B) / B × 100 | | Name B as denominator |
| Relative to A, if useful | (A − B) / A × 100 | | Name A as denominator |
| Prior-period level | Current / (1 + growth) | | Only if growth and level share basis and vintage |
| Vendor diagnostic | Total × share / 100 | | Derived eligible amount, not a company disclosure |
| Category bridge needed | A − B | | Only under a matched remainder; not a measured category |
| Rounding sensitivity | Lower/upper inputs from display precision | | State rounding assumptions; not confidence intervals |

**Do not:** combine constant-currency growth with a nominal level without a bridge; replace a quarter with annual revenue/4; equate run rate with observed annual revenue; average estimates into a “correct” figure without a justified error model.

## Classify explanations

| Hypothesis | Confirmed / plausible but unconfirmed / not supported / unresolved | Supporting evidence | Evidence needed to resolve |
|---|---|---|---|
| Category boundaries | | | |
| Geography / customer population | | | |
| Period or annualization | | | |
| Vendor eligibility and allocation | | | |
| Revenue vs expenditure | | | |
| Currency basis | | | |
| Method/model differences | | | |
| Rounding / revisions | | | |

**Confirmed observations:** ____________________  
**Our calculations:** ____________________  
**Interpretation:** ____________________  
**Unresolved differences:** ____________________  
**What could change the conclusion:** ____________________

## Final check
- Follow important numbers to their original source.
- Reopen originals; compare dates, units, definitions and growth qualifiers.
- Recalculate formulas and record source inputs.
- Count source origins, not pages repeating the same statistic.
- Keep model estimates distinct from observed facts and your own calculations.
- Make assumptions and unresolved items visible beside the conclusion.
- State the fit of the selected estimate to your question; do not claim unsupported accuracy.
- Provide sources, author, real publication date and correction path.

Suggested conclusion: “A reports ___ for ___; B reports ___ for ___. We confirmed ___. The numerical difference is ___ on the stated basis. ___ is plausible but unconfirmed. We cannot yet reconcile ___; resolving it requires ___.”