Methodology · v1.0 · Q1 2026

Measure what companies do, not what they say.

The AIR APAC Mid-Market Readiness Index is built from observable, multi-source signals – not surveys or self-reports. This page documents the methodology used in the Q1 2026 Index: scoring architecture, confidence framework, tier vocabulary, and the relationship between the Index and the qualitative six-dimension diagnostic.

---

Why behavioural signals

Survey data on AI readiness has a well-known reliability problem: respondents over-report progress, under-report obstacles, and answer differently depending on who is asking. For mid-market organisations across Asia Pacific – where boards, family stakeholders, and customers all pay close attention – the gap between stated intent and observable behaviour is structural, not occasional.

AIR APAC’s methodology inverts the standard approach. Instead of asking what an organisation thinks it is doing, we measure what it is publicly doing – the technology stack it operates, the talent it is hiring, the public narrative it is choosing to invest in. The result is comparable, longitudinally trackable, and resistant to the most common reporting biases.

Three-layer scoring

Each company in the Index is scored 0–100 on the Total Readiness Score (TRS). The TRS aggregates three weighted layers:

Weights are hypothesis-driven. They reflect the analytical judgment that talent signals are the most reliable and discriminating data currently available for APAC mid-market companies. The weighting is held constant across the Q1 2026 Index and will be reviewed annually against longitudinal outcome data.

Confidence framework

Every TRS is paired with a confidence rating. A score is not a verdict; it is an estimate of the underlying readiness state, conditional on the density of public signal available.

Confidence	Q1 2026 share	Definition
High	5.5% (28 / 510)	Full signal density across all three layers; TRS is a direct read of public behaviour.
Medium	24.5% (125 / 510)	Partial signal density; scoreable with caveat. Used in cross-market comparisons but flagged.
Low	70.0% (357 / 510)	Insufficient publicly observable signal. The score reflects observability as much as readiness; not used for tier-level conclusions.

The 70% Low-confidence finding is itself a Q1 2026 result – what AIR APAC calls asymmetric visibility. The headline statistic of the Q1 Index is not an average TRS; it is the share of mid-market APAC companies that cannot be assessed with confidence at all. Cross-market comparisons must therefore be read as a composite of readiness and observability.

Tier vocabulary

Within the confident subset, companies are placed into one of four tiers based on TRS thresholds calibrated to Q1 2026 sample distribution.

Tier thresholds are reviewed each Index cycle. A company with a Low-confidence rating is not assigned a tier – it is held in the unplotted set until signal density rises.

The Index and the six-dimension diagnostic

AIR APAC operates two complementary instruments. They are calibrated to different questions and should not be confused.

	Mid-Market Readiness Index	Six-Dimension Diagnostic
Question answered	Where does this organisation sit relative to peers across APAC?	What specifically is blocking this organisation from scaling AI?
Method	Three-layer behavioural-signal scoring (Digital Footprint, Talent, Strategic Intent)	Six-dimension qualitative assessment (Leadership, Data, Skills, Process, Governance, Culture)
Inputs	Public signals only	Executive interviews, document review, stakeholder assessment
Output	TRS 0–100 + tier + confidence	12-page action framework with prioritised barriers and 90-day milestones
Cadence	Quarterly Index, annual flagship	One-time per engagement (or repeat by design)
Use	Benchmarking, market intelligence, public research	Advisory engagements, board readouts, transformation roadmaps

The Index ranks. The diagnostic explains. An organisation can read its Index position to know where it stands; it commissions a diagnostic to know what to do next. Most engagement journeys begin with the Index and proceed to a diagnostic when leadership commits to act.

Limitations

The Index is honest about what it cannot do.

• Observability is not readiness. A Low-confidence rating does not prove a company is unprepared – only that it cannot be distinguished from one. The Q2 2026 brief, The Confidence Floor, will examine the cheapest interventions to raise observability.
• Sector-level effects. Manufacturing’s 88% At-Risk rate in the full sample partly reflects weaker observability of shop-floor AI activity, not only weaker readiness. The confident-subset manufacturing finding is more balanced.
• Cross-market comparison. Markets with high Low-confidence rates (Malaysia and Thailand at 2% High-confidence each) are not directly comparable to Australia (10% High-confidence) without controlling for visibility.
• Hypothesis-driven weights. The 25/45/30 weighting reflects analytical judgment, not regression-derived optimisation. Weights are reviewed annually as longitudinal outcome data accumulates.
• Time-windowed. Each Index is a snapshot. Signal density and AI behaviour evolve quickly – a Q1 2026 score is not interchangeable with a Q4 2026 score.

Versioning

This methodology page tracks the canonical version of the Index methodology. Material changes are recorded below.

• v1.0 · Q1 2026 (April 2026): First publication. 510 companies, five markets (AU, SG, MY, ID, TH), six sectors. Three-layer scoring with 25/45/30 weights. Confidence framework introduced. Pacesetter / Strategist / Pragmatist / Traditionalist tier vocabulary established.

---

Methodology v1.0 · Q1 2026 · A research publication of the Center for AI Readiness, Asia Pacific.