Alternative data for investments on demand.
Enrich investment models with satellite and geospatial activity, anonymized card-spend and transaction trends, web-traffic and app-engagement data, and news and social sentiment — delivered cleaned, normalized, and point-in-time, so your team builds alpha instead of pipelines. Pull a single ticker, a trial sample, or a one-off name on demand. Charged per pull, no full feed required.
Built for investment teams that need the signal, not the pipeline.
Four patterns where pulling alternative data on demand is faster, cheaper, or lower-risk than licensing a full feed upfront.
Test a new signal before committing to a full feed.
Pull the history for a specific signal type — card-spend, satellite activity, web-traffic, or sentiment — and run it through your backtesting framework against the tickers or sectors you care about. If the signal survives — shows genuine predictive power on revenue, demand, or price — you have the evidence to justify a full feed license. If it doesn't, you've spent a fraction of what the full license would have cost to find out.
Extend coverage to a name or region outside your existing feed.
Your current alternative data subscriptions cover your primary universe. When a name outside that universe becomes relevant — a new idea, a client request, an emerging-market name your feed doesn't include — pull its signal history on demand rather than upgrading your subscription for one name.
Get the current alternative-data read on a name, right now.
When a holding catches your attention — ahead of earnings, after a news event, or during a sector review — pull the latest card-spend trend, web-traffic reading, or sentiment score for that ticker. A single on-demand pull gives you the alternative-data view without a quant team intermediary or a wait for the next data delivery.
Validate signal quality on a sample before negotiating a license.
Alternative data vendors typically offer full-universe feeds on annual contracts. Pull a representative sample — a sector, a date range, a signal type — on demand and evaluate quality, coverage, and predictive power against your own benchmarks before entering any negotiation. The evaluation cost is the pull cost, nothing more.
The signal, not the pipeline.
Pull a single ticker's history or a trial sample today. No feed contract, no minimum spend — you only pay for the data you actually pull.
- Charged per pull
- No full feed required
The honest answers.
If something below doesn't cover your case, ping us — we answer directly, no sales process.
What signal types are available?
+
Four categories: satellite and geospatial activity (facility usage, shipping movements, industrial output indicators); anonymized card-spend and transaction trends (consumer spending by merchant category, company, and region); web-traffic and app-engagement data (site visits, session trends, app downloads); and news and social sentiment (NLP-scored news and social media, by issuer and sector).
What does 'point-in-time' mean and why does it matter?
+
Point-in-time means the data record reflects what was known at that date — no later revisions, no restated figures folded in. Without it, a backtest will appear to perform better than it would have in live trading, because it incorporates information that wasn't available at the time. All data delivered here is point-in-time, so backtest results reflect what your model would actually have seen.
What does 'charged per pull' mean?
+
You pay for the data you request. A pull covering one ticker's card-spend history costs less than a pull covering a full sector. There is no minimum spend, no subscription fee, and no charge for data you don't request.
How is this different from licensing a full alternative data feed?
+
A full feed gives you the entire signal universe on a recurring delivery schedule, priced for that scale and typically requiring an annual contract. This service returns the signal for the specific tickers, date ranges, and signal types you request, charged per pull. For research and evaluation work — before you know which signals survive your backtesting process — the per-pull model is almost always cheaper and carries no commitment risk.
Is the data backtest-ready out of the box?
+
Yes. Data is delivered normalized to a consistent format, point-in-time, and free of survivorship bias. You can pipe it directly into your backtesting framework without a cleaning or transformation step.
Can I pull data for names outside major indices?
+
Yes. Coverage includes names across global markets, including smaller-cap and emerging-market issuers that many alternative data vendors either exclude or charge a premium to cover. If a specific name is outside coverage, the service returns a clear no-coverage response at no charge.
Can this be used inside an automated research workflow?
+
Yes. The service is designed for both ad-hoc analyst use and automated research pipelines. A quant team can integrate it into a systematic evaluation framework; a discretionary analyst can query it manually. Both pay only for what they pull.