What is BrandSignals?
BrandSignals is an intelligence research and analytics platform built around the Behavioural Business Signals framework, a dual-score model designed to evaluate publicly traded consumer companies through two complementary lenses: financial performance and behavioural business quality.
It is built for investors, analysts, and commercially curious professionals who want to understand beauty and fashion businesses more holistically. Rather than relying on financial statements alone, BrandSignals explores how a company performs both on paper and in practice — through the combined analysis of financial strength, consumer sentiment, employee feedback, brand visibility, and other behavioural indicators.
The platform currently covers a curated set of publicly traded companies across the beauty, beauty retail, fashion, apparel, and performance lifestyle sectors. The aim is not simply to rank businesses by profitability, but to examine where financial strength and experiential strength align — and where meaningful gaps emerge.
Companies & Sectors
The current company set spans four broad groupings:
- Beauty — companies centred on cosmetics, skincare, fragrance, haircare, and adjacent beauty products, where brand trust, consumer loyalty, and product sentiment play a significant role in long-term business quality.
- Beauty Retail — retailers and distributors whose performance depends not only on product mix and margins, but also on customer experience, repeat purchasing behaviour, and in-store or digital engagement.
- Fashion & Apparel — branded apparel and accessories businesses where identity, positioning, customer perception, and operational execution are closely tied to commercial performance.
- Performance Footwear & Lifestyle — companies operating at the intersection of fashion, function, and brand culture, where growth is often shaped by both financial execution and behavioural signals such as loyalty, aspiration, and public momentum.
Methodology
Companies are included according to three main criteria.
First, they are publicly traded, which makes their financial disclosures accessible, comparable, and sufficiently standardised for structured analysis.
Second, they are consumer-facing and brand-sensitive, meaning that behavioural signals such as customer sentiment, employee reviews, and public attention are both visible and analytically meaningful.
Third, they operate in sectors where the relationship between financial performance and experiential quality is especially interesting. In beauty and fashion, business quality is shaped not only by revenue growth, margins, and cash flow, but also by trust, loyalty, perception, and execution at the customer and employee level. These characteristics make them especially well suited to a dual-score framework.
Data Sources
BrandSignals draws on a combination of automated and manually collected public data sources. Financial data is refreshed quarterly. Behavioural data is refreshed quarterly for automated sources and every three to six months for manually collected sources.
Powers the BrandSignals chat assistant ✦ Ask using the Gemini 2.5 Flash-Lite model via Google's Generative Language API. Current usage limit: 1,000 requests per day.
yfinance is an open-source Python library that provides access to Yahoo Finance data. It serves multiple purposes across BrandSignals: it collects all 10 financial metrics from annual income statements, balance sheets, and cash flow statements for all companies in the BrandSignals universe; it provides live stock prices (15-minute delayed) shown on individual company pages and the visualise quick-view panel; it provides live market capitalisation used to size bubbles on the visualise page (refreshed every 24 hours); and it provides real-time valuation multiples (P/E, EV/EBITDA, Price/Sales) and market sentiment data (price returns, 52-week range, volatility) shown on individual company pages. All financial scoring data reflects the most recently available FY2025 annual filing, collected in May 2026. Fiscal year-end dates vary across the universe; full detail is provided on the Financial page.
FMP remains configured as the primary source in the financial data waterfall, with yfinance as fallback. It is currently inactive due to API changes. The waterfall structure is preserved so FMP can reactivate automatically if access is restored and to support future expansion of the company universe.
Glassdoor is used to collect employee sentiment data for each company. Eight fields are collected: Overall Rating, CEO Approval, Recommend to a Friend, Culture and Values, Work-Life Balance, Diversity and Inclusion, Senior Management, and Compensation and Benefits. Three are used directly in the Behavioural Score: Overall Rating, CEO Approval, and Recommend to a Friend. The remaining fields are shown on individual company pages for context. Data is collected manually and refreshed every three to six months. Last updated May 2026.
Trustpilot is used to collect consumer satisfaction data for each company's primary consumer brand page. Two fields are collected: overall star rating and total review count. The rating is used directly in the Behavioural Score, while review count acts as a confidence modifier rather than a standalone metric. Trustpilot tends to skew negative in retail and beauty because dissatisfied customers are more likely to leave reviews, so it is most useful as a relative peer signal rather than an absolute quality measure. Data is collected manually and refreshed quarterly. Last updated May 2026.
App ratings are collected from the Google Play Store or Apple App Store depending on data availability, using whichever source provides the most complete and representative rating for each company's primary consumer-facing mobile app. Google Play Store is the primary source; Apple App Store is used where Google Play data is unavailable or insufficient. Data is gathered through automated scraping and manual collection. Companies without a meaningful consumer app are excluded from this signal, and the weight is redistributed within the Consumer Experience and Trust bucket. Data is refreshed quarterly. Last updated May 2026.
Google Trends data is collected via pytrends, an unofficial Python interface for Google Trends. Two scores are collected for each company: a US search interest score and a global search interest score, each based on average weekly search interest over the past 12 months. Which score is used in the Behavioural Score depends on the company's geographic classification: US-focused, global, or mixed. Full detail is provided on the Behavioural page. Data is refreshed quarterly. Last updated May 2026.
YouTube subscriber counts are collected via the YouTube Data API v3, Google's official API for channel data. Each company's primary consumer-facing YouTube channel is used where available. For holding companies with no meaningful direct corporate YouTube presence, the largest fully-owned active consumer brand channel is used as a proxy. Proxy channels must meet an activity threshold — dormant channels are excluded regardless of subscriber count. A modest confidence discount is applied to proxy-based metrics. Full methodology and current proxy assignments are detailed on the Behavioural page. Data is refreshed quarterly. Last updated May 2026.
Follower counts for each company's primary brand account, collected manually. Used as a brand momentum signal reflecting chosen ongoing followership. Last updated May 2026.
Total likes for each company's primary brand account, collected manually. Used as a brand momentum signal reflecting cumulative engagement and cultural relevance. Last updated May 2026.