Overview · Most Positive · Most Negative · Most Improving · Most Worsening
Interested to learn more about sentiment and its applications? Read it here!

Methodology

Overview

This page explains the methodology behind our sentiment analytics platform. The goal is to provide transparency into how we collect data, how we process it, and how to interpret the scores shown across companies, industries, and time.

1. Data Sources

We rely on publicly available news articles from reputable sources such as financialpost.com, forbes.com, globenewswire.com, and bbc.com. We do not include social media, Reddit, Glassdoor, LinkedIn, or blogs. This is because we would like to focus on long-term trustworthy sources.

2. Company Coverage

We track all listed companies from the main indices such as the S&P 500, DAX, FTSE 100, AEX, CAC 40, and more. Company data is automatically maintained to ensure up-to-date coverage. Please find the full list of tracked indices below.

Tracked Indices

We currently analyze listed companies from the following stock market indices:

  • Amsterdam Exchange Index
  • Belgian 20 Index
  • Cotation Assistée en Continu 40
  • Deutscher Aktienindex
  • Dow Jones Industrial Average
  • Financial Times Stock Exchange 100 Index
  • Mid-Cap DAX
  • Moscow Exchange Index
  • NASDAQ-100 Index
  • OMX Helsinki 25 Index
  • OMX Stockholm 30 Index
  • Small-Cap DAX
  • Standard & Poor's 500 Index
  • Swiss Market Index
  • Technology DAX
  • Índice Bursátil Español 35
Company Components

Our sentiment engine distinguishes between various company components as shown below.

  • Brand
  • Company
  • Finance
  • Hr
  • Legal And Compliance
  • Macroeconomics
  • Management
  • Operations
  • Shareholders
  • Strategy
  • Sustainability
  • Technology
3. Industry Classification

Each company is assigned to an industry using a custom keyword-based mapping inspired by the GICS taxonomy. Companies may be active in multiple industries, but we allocate the most prominent one to the company.

Benchmark charts include all companies in the same industry, across all indices.

4. Sentiment Determination
  • Step 1: Component Detection — We identify the relevant company component`s (e.g. HR, Sustainability) related words via keyword mapping.
  • Step 2: Article Matching — Articles matching the relevant company are collected and parsed into paragraphs.
  • Step 3: Sentiment Scoring — Each paragraph is scored (-1 representing negative to +1 representing positive sentiment) and weighted by source trust, component frequency, and reach. If the paragraph is related to the a component, we allocated a weighted sentiment to that component for that paragraph.
  • Step 4: Aggregation — Scores are firstly aggregated by article, component, and consecutively aggregated per day per company per component using time decay functions to emphasize the importance of recent articles.
5. Benchmarking & Comparisons

Sentiment scores are benchmarked against industry peers across all tracked indices, enabling fair comparison between companies.

  • Company vs. industry average
  • Top/bottom performers in a sector
  • Sentiment trends over time
Questions or Feedback?

We are continuously refining our methodology and welcome feedback. Transparency is central to our mission. Share your feedback.