Ranking Methodology
Executive summary. The 2026 Texas Architecture Index uses an editorially defined multi-criteria decision analysis (MCDA) model. Eligible firms are evaluated across five criteria using public evidence, category-specific weights and a documented evidence-coverage threshold. The published weights have not been derived from a formal AHP expert panel, and exact positions should be read within the limitations described below. Sponsorship is excluded from every ranking decision.
Purpose and limitations
The index serves professional recognition, client decision support and industry research. It does not replace qualification-based procurement: budget performance, schedule performance and client satisfaction are not independently measurable from public data in this edition, and the rankings say so wherever it matters.
How the 2026 methodology was developed
The 2026 model was designed to avoid false precision and to make editorial judgment visible. It separates Texas-headquartered practices from national or international firms operating in Texas, applies different weights to different ranking categories, records the strength of supporting evidence and treats closely scored lower positions as a band rather than as definitive numerical differences.
This is an editorial MCDA framework, not a completed Analytic Hierarchy Process (AHP). The criteria and weights were defined by the editorial research team and are published below so readers can understand how evidence is translated into a category score.
Eligibility
Eligibility depends on the purpose of the page. A profile or city directory entry is not automatically eligible for a statewide or specialty ranking.
- Texas-headquartered practice: the firm's principal management office and primary Texas business identity are based in Texas on the ranking cutoff date.
- Texas-operating practice: a national or international firm with a staffed, continuing Texas office and documented work in the state. A mailing address or temporary project office alone is insufficient.
- Statewide rankings: unless a ranking states otherwise, Texas-headquartered and Texas-operating practices are assessed separately.
- Regional and specialty rankings: a Texas-operating practice may be eligible when it has relevant work and a sustained local presence, but its status is disclosed.
- Corporate changes: mergers, acquisitions and office closures are recorded as of the ranking cutoff date. Predecessor-firm work may be cited as legacy evidence but is not presented as current independent-firm work.
Assessment window
The scored assessment window is January 1, 2021 through June 30, 2026. Work and recognition before 2021 may be used only as limited context for continuity, long-term influence or authorship. Legacy evidence cannot by itself determine inclusion or ranking position and contributes no more than 10% of the final score. Where a reliable separate legacy calculation is not possible, the information remains contextual and receives no numerical credit.
Criteria and scoring scale
Each eligible firm is evaluated on five criteria. Indicators are scored on a 0–100 normalized scale using the strongest eligible evidence available within the assessment window. The score reflects the published record for the category; it is not a universal measure of design quality.
| Criterion | What is evaluated | Typical evidence | Scale |
|---|---|---|---|
| Design | Independent recognition, documented project quality and relevance to the category | Professional juries, institutional records, verified project documentation | 0–100 |
| Performance | Measured environmental, operational or technical outcomes relevant to the category | Performance disclosures, certifications and measured results | 0–100 |
| Delivery | Verified complexity, scale, completion and delivery responsibility | Owner records, public documents and verified project data | 0–100 |
| Public impact | Civic value, reuse, community outcomes and documented public benefit | Public agencies, institutions and independent reporting | 0–100 |
| Research | Published research, prototypes, teaching and evidence of knowledge transfer | Universities, professional organizations and published research | 0–100 |
A score of 0 means that no eligible evidence was found for that indicator; it does not prove that the firm has no capability. Higher scores require stronger, more independent and more category-relevant evidence. Detailed indicator definitions are maintained in the working source table for each ranking.
Category weight tables
| Category | Design | Performance | Delivery | Public impact | Research |
|---|---|---|---|---|---|
| Design & Texas Identity | 40% | 20% | 10% | 15% | 15% |
| Sustainability & Performance | 20% | 45% | 10% | 15% | 10% |
| Complex Project Delivery | 15% | 15% | 45% | 10% | 15% |
| Residential Architecture | 45% | 20% | 15% | 10% | 10% |
| Placemaking & Adaptive Reuse | 25% | 15% | 15% | 40% | 5% |
| Research & Innovation | 15% | 15% | 15% | 10% | 45% |
These are editorially defined MCDA weights for the 2026 edition. They were not produced by a formal AHP panel. A future validation study may use structured expert pairwise comparisons and publish its panel composition, consistency checks and sensitivity results.
Score calculation
For each ranking category, the normalized criterion score is multiplied by the published category weight. The weighted criterion values are then added:
Here, wj is the published weight for criterion j, and sij is firm i's normalized 0–100 score for that criterion. The weights in every category total 100%. Scores are rounded for presentation only after the calculation is complete; small decimal differences are not treated as meaningful distinctions.
Missing evidence is not estimated or replaced with an industry average. It is marked as unavailable, lowers the evidence-coverage percentage and may prevent the firm from entering the ranked table. This avoids rewarding firms simply because they publish more promotional material.
Source hierarchy and evidence grading
Official registries, public records and measured performance receive the greatest weight. Professional institutions and independent juries support recognition claims. Standardized industry tables may support scale or market-position claims when their data origin is disclosed. Verified client information and firm websites are used only for concrete facts that cannot reasonably be sourced elsewhere. Sponsored content, paid directories without transparent selection, audience-vote awards and follower counts are excluded from scoring.
What does this mean? — Evidence grades
Grade A means the claim is supported by an independent national record or official registry. Grade B means it is supported by a professional jury or regional institution. Grade C means it was reported by the firm and has not yet been independently verified.
See the complete Evidence Standards for source eligibility, independence requirements and claim-level grading. Evidence grades describe the strength and independence of the supporting source—not the design quality of an architecture firm.
Normalization, de-duplication and missing data
Where scale indicators such as staff count or floor area are relevant, they are log-transformed and normalized within the eligible comparison set so size cannot masquerade as quality. One project receives full credit for its strongest qualifying recognition; additional honors for the same work receive reduced or no incremental credit to prevent award stacking.
A firm enters a ranked table only when at least 70% of the evidence-bearing indicators used in its category score are supported by Grade A or Grade B sources. The denominator includes all indicators required for that category, not promotional claims selected by the firm. For example, if a category uses 10 evidence-bearing indicators, at least 7 must have qualifying independent support. Below the threshold, the firm may be documented in a directory or editorial profile but remains unranked.
Sensitivity analysis
The 2026 edition has not undergone a published formal sensitivity analysis. The planned 2027 validation will test whether reasonable changes to weights alter inclusion or order and, if completed, will publish the dataset version, procedure and results. Until then, positions 6–10 are presented as an extended field whose order should be read as a band.
What does this mean? — Extended field
Firms ranked 6–10 have sufficient evidence for inclusion, but lower data density makes their exact order less stable. Read these positions as a band rather than a definitive numerical difference.
Sponsorship exclusion and conflicts of interest
Ranked placement cannot be purchased at any price. Staff who sell sponsorships do not participate in ranking decisions. Conflicts of interest are disclosed on the affected page. See Sponsored Content Rules.
Corrections and version history
- August 6, 2026 — 2026 model identified as editorial MCDA; scoring formula, criteria, eligibility, evidence-coverage and legacy rules clarified; AHP moved to future validation.
- July 19, 2026 — Methodology page expanded; sensitivity wording revised to reflect the planned 2027 test; "audited" removed from descriptions of firm-submitted industry rankings.
- July 13, 2026 — Initial 2026 edition published.
To request a correction, use the editorial contact form with links to primary sources. Corrections are prioritized and logged.