About the Data

The Company Dataset

Lumino's benchmark dataset covers 100 global fashion and apparel companies representing $0B in combined annual revenue and an estimated $0.0B in combined IT spend. These companies were selected to represent the full spectrum of business models in the industry — from vertically integrated fast fashion to luxury outerwear, pure-play e-commerce to value retail — giving Lumino users a meaningful and diverse peer comparison. The full company list is available in Benchmark Explorer.

Business Model Categories

Lumino classifies all 100 companies into seven business model categories. IT spend intensity varies systematically across these categories — driven by digital channel mix, supply chain complexity, and technology as a competitive differentiator.

Data Points Available Per Company

For each of the 17 core companies in Lumino's detailed dataset, a consistent set of financial and operational data points is available at three levels of geographic granularity: global, regional, and market (country). The remaining 83 companies in the benchmark dataset provide company-level data only.

The number of active markets varies by company. Inditex has data across 30+ countries. Aritzia covers US and Canada only. A country appears in the dataset only if that company has meaningful commercial operations there.

How We Collected the Data

Primary sources — public financial filings

Revenue figures at global and regional level are sourced directly from each company's most recent annual report or SEC filing (10-K for US-listed companies, 20-F for foreign private issuers, and local equivalents for European companies). Regional breakdowns follow each company's own reporting segments — for example, Inditex reports into Spain, Europe ex-Spain, Americas, and Asia/Rest of World; lululemon reports Americas, China Mainland, and Rest of World. Where a company's reporting segments do not align exactly with Lumino's five-region model (North America, South America, Europe, Africa, Asia), figures are reclassified using the geographic definitions standard in the industry.

All revenue figures are converted to USD at the average exchange rate for the relevant fiscal year. EUR/USD 1.08, GBP/USD 1.27, CAD/USD 0.74, CHF/USD 1.12, SEK/USD 0.095 were the primary rates applied.

IT spend — estimation from benchmarks

Unlike revenue, IT spend is rarely disclosed as a standalone line item in fashion company financial reports. Lumino's IT spend figures are estimated using a layered benchmarking approach.

The foundation is the McKinsey State of Fashion Technology report (2022, updated with 2024 trajectory data), which established that fashion companies invested 1.6–1.8% of revenues in technology in 2021, trending toward 3.0–3.5% by 2030. For FY2024, Lumino uses a baseline range of 1.8–2.2% for traditional retailers, with adjustments applied for business model.

HG Insights retail technology data (2024) provided category split benchmarks: IT Services (labor) 44% of total IT spend, Software 28%, Hardware/Communications (Other) 28%. These splits are adjusted for digital maturity — DTC-led companies carry a higher labor and software share.

Gartner IT Key Metrics Data (retail vertical) provided cross-validation for IT spend intensity ranges and category mix.

Business model adjustments

A flat benchmark is not appropriate across a peer set that spans Primark (no e-commerce, store-only) to Shein (technology company that manufactures fashion). Lumino applies the following directional adjustments:

Actual averages by category (100-company dataset)

Country-level data

Country revenue figures are estimated from regional totals using store count data, disclosed country-level figures, and proportional allocation based on market size and brand presence. Country-level IT spend is calculated by applying the global IT% to the country revenue figure — a simplification that provides a consistent and comparable baseline.

Why IT Spend is Reported Globally

We report each company's IT spend as a single global figure — not split by region or country. No company discloses its IT budget broken down by geography, so any such split would be invented: false precision dressed up as detail. Revenue is different. Where a company genuinely reports revenue by segment we carry those segments through; IT spend, we don't, because the disclosure to support it doesn't exist.

How we arrive at the global figure depends on what the company discloses. When it reports its IT or technology spend directly, we read it from the filing (Reported). When it discloses something we can reason from — a cost breakdown that implies it — we derive it and label it as such (Inferred). When it discloses nothing, we estimate from public-source sector benchmarks, adjusted for the company's business model — a stated assumption that, say, a luxury house and a value retailer don't spend alike, not a measurement — and say plainly the number is modelled (Modelled). Where even a grounded estimate isn't possible, we decline to publish a figure rather than guess.

Nothing reaches you unreviewed: every figure passes a person before it's published. How much trust each tier earns, and how it rolls into a company's score, is set out in full on the DQI methodology page.

How We Score Data Quality

Every figure in Lumino carries a provenance label — not just a number, but a record of where that number came from and how much trust it deserves. The Data Quality Index (DQI) turns that provenance into a single 0–100 score you can compare across companies and watch improve as you add your own verified data.

Why provenance matters

Two companies can show identical IT spend percentages while drawing on very different evidence bases. One figure might come from a published annual report with explicit IT disclosure; another might be modelled from a peer median. The DQI makes that difference visible so you can calibrate how much weight to place on each comparison.

The four evidence tiers

Three-stage weighted roll-up

The DQI is not a flat average. It weights evidence along two axes — geographic level and field importance — then combines them into a single score.

Within each level, fields are further weighted by decision relevance. Revenue and IT spend percentage — the figures most directly used in benchmarking — carry the most weight. Category splits (labour, software, other) and internal allocation ratios carry less, because they are rarely disclosed and their absence is the norm rather than the exception.

Score bands

Calculated fields

IT spend as a percentage of revenue is a calculated field — derived live as IT spend (USD) ÷ revenue × 100. It is displayed throughout the platform for readability but is never scored independently by the DQI. Scoring it separately would double-count the same underlying evidence already captured by the revenue and IT spend figures. Calculated fields carry a grey ƒ Calculated badge in the Data Foundations page.

Improving your score

Registered companies can enter their own figures directly in the Data Foundations page. Any field you verify with a source link moves from Modelled (25) or Inferred (55) to Verified by you (100), immediately raising your DQI. The biggest gains come from confirming global IT spend (USD) — it accounts for roughly 29% of the global sub-score.