Earnings season concentrates research effort and alpha opportunity. Alternative data is used to set expectations, validate or challenge consensus, and interpret results. Not every dataset or platform is equally useful for earnings, and the "best" mix depends on your strategy, coverage, and existing stack. This post compares the main categories of alternative data used around earnings and how leading platforms fit in, so you can choose what to add or prioritize.
What institutional investors use around earnings
Demand and behavioral data. Search volume, web and app traffic, and social or trend data gauge interest and demand before the print. For consumer, retail, and digitally exposed names, these signals often lead revenue by weeks. They are used to form a view on whether demand is strengthening or softening and to compare that view to consensus. Platforms that offer normalized, ticker-mapped behavioral data (e.g. Paradox Intelligence with search, social, news, traffic, and 15+ sources) let you run multi-signal checks in one place and align with an earnings calendar so you know when each name reports.
Sentiment and narrative. News sentiment, social sentiment, and earnings-call tone help capture narrative shift and event risk. Sentiment alone is rarely sufficient; it is more useful when combined with behavioral data (e.g. "demand up, sentiment stable" or "sentiment turning negative ahead of print"). Providers like RavenPack and Quiver focus on news and disclosure; for how they compare to behavioral platforms, see Paradox vs RavenPack vs Quiver.
Fundamentals and earnings automation. Some platforms specialize in speeding up the post-print workflow: model updates, estimate revisions, and source-linked accuracy. These are not "alternative data" in the search/social sense but are often used in the same earnings process. Daloopa, for example, provides earnings and fundamental data automation (Excel, API, MCP) for fast model updates. Paradox complements that by providing behavioral and demand signals before and around the print; the two can be used together (behavioral for pre-print view, fundamentals automation for post-print update).
Transaction and spending data. Aggregated card or receipt data can support revenue or same-store expectations for consumer and retail. Coverage and latency vary; providers like YipitData and Earnest Analytics are commonly cited. Transaction data is outcome-focused (what people bought); behavioral data (search, traffic) is intent-focused. Many funds use both: behavioral for lead, transaction for confirmation. For how they differ, see Best Consumer Demand Data for Hedge Funds.
Comparison at a glance
| Category | What it does for earnings | Example providers | Best when |
|---|---|---|---|
| Behavioral (search, traffic, social) | Demand and interest lead; multi-signal check | Paradox, Similarweb, Thinknum | You need a pre-print demand view and ticker-mapped series |
| Sentiment and news | Narrative and event risk; tone of coverage | RavenPack, Quiver, AlphaSense | You want to combine narrative with behavioral |
| Fundamentals / earnings automation | Fast model updates, estimate accuracy post-print | Daloopa, Visible Alpha | You need speed and accuracy after results |
| Transaction / spending | Revenue and same-store expectations | YipitData, Earnest Analytics | You cover consumer/retail and have access |
Stay up to date on our best ideas
Paradox Intelligence (behavioral around earnings)
Paradox provides multi-source behavioral data (Google Search, Images, News, Shopping, YouTube, Wikipedia, TikTok, Amazon, Reddit, web traffic, app, and others) normalized and mapped to tickers. For earnings, you get: demand and trend series to set or challenge expectations; Analyse to compare multiple signals (search, traffic, sentiment, etc.) for the same name; an earnings calendar and history per ticker; and API/MCP for pipelines and alerts. Best for: pre-print demand and narrative checks and continuous monitoring, with one platform for discovery, comparison, and calendar. For detail on what helps around earnings, see Using Alternative Data Around Earnings: What Actually Helps.
Sentiment and document platforms
RavenPack and Quiver Quantitative focus on news, sentiment, and public disclosures (congressional, insider, etc.). They are used for narrative and event risk around earnings, not for demand time series. AlphaSense adds AI search over transcripts, filings, and research. Use these when you need sentiment or document intelligence; combine with behavioral data (e.g. Paradox) for a full picture. See Paradox vs RavenPack vs Quiver and Paradox vs AlphaSense.
Fundamentals and earnings automation
Daloopa and Visible Alpha–style platforms focus on fundamental and earnings data: model updates, estimate accuracy, and workflow speed after the print. They are complementary to behavioral alternative data: use Paradox (or similar) for demand and narrative before and around earnings; use Daloopa/Visible Alpha for fast fundamental updates after. No need to choose one; they serve different steps in the process.
Transaction data (consumer/retail)
YipitData, Earnest Analytics, and similar providers deliver transaction and spending data. They are strongest for consumer and retail names where you need revenue or same-store proxies. Use them when you have coverage and budget; combine with behavioral data so you have both intent (search, traffic) and outcome (spend). Paradox does not provide transaction data; it positions as the behavioral layer that leads and validates.
How to combine them
A practical earnings workflow often looks like this:
- Pre-print. Use behavioral data (search, traffic, sentiment) to form a demand and narrative view; align with earnings calendar so you know which names to focus on.
- Post-print. Use fundamentals/earnings automation for fast model and estimate updates; use behavioral and transaction data to interpret beats/misses and update forward views.
- Ongoing. Use the same behavioral platform for screening and prioritization (e.g. which names deserve deeper pre-print work next quarter).
No single provider does everything. Many institutions use: one behavioral platform (e.g. Paradox) for demand and multi-signal checks; one sentiment or document platform if needed; one fundamentals/earnings automation tool; and optionally transaction data for consumer/retail. For a broader platform comparison, see Best Alternative Data Platforms 2026.
Summary
- Behavioral data (search, traffic, social) = demand and intent lead; best for pre-print expectations and multi-signal validation. Paradox is built for this.
- Sentiment and news = narrative and event risk; combine with behavioral for a full view.
- Fundamentals/earnings automation = post-print speed and accuracy; complementary to behavioral.
- Transaction data = revenue/same-store for consumer/retail; use with behavioral for intent and outcome.
Choose by use case: if earnings is a key focus, prioritize a behavioral platform with ticker mapping and an earnings calendar, then add sentiment, fundamentals, or transaction data as needed.
This post is for institutional investors and research professionals. It is not investment advice.
