The Faraday Identity Graph: 1,400+ data points on 240MM US consumers

The Faraday Identity Graph (FIG) is a continuously maintained consumer dataset covering 240 million U.S. adults across 1,400+ verified data points — demographics, financial signals, property data, life events, and lifestyle indicators. It's the foundation underneath every prediction, data append, persona, and audience Faraday produces.

This guide covers what's in FIG, how it's built, why it works the way it does, and what brands actually do with it.

What's in FIG

FIG's 1,400+ data points are organized into categories that together describe who a person is, where they are in life, and how they behave economically:

• Demographics — age, gender, household income, education level, marital status
• Property and housing — homeownership status, home value, roof type, garage spaces, length of residence
• Financial signals — Predictive datapoints like estimated income, spending percentiles, investment behavior, debt indicators
• Life events — recent move, new homeowner, new parent, marriage, retirement
• Lifestyle and interests — hobbies, brand affinities, vehicle ownership, charitable giving patterns
• Contact and identity — verified name, address, email, phone, household composition

The full catalog is browsable at faraday.ai/catalog, with documentation, value distributions, and metadata for every data point.

Why coverage and depth matter

But having a massive catalog of attributes only matters if it actually applies to your customers. The two numbers that define FIG — 240 million adults, 1,400+ data points — aren't just marketing figures. They're functional thresholds that determine what FIG can actually do.

Coverage drives match rates. When you bring customer records to Faraday, the first thing that happens is matching them against FIG. If coverage is thin, a meaningful portion of your records come back unenriched — fewer leads scored, smaller addressable audiences, weaker models. At 240M U.S. adults, FIG covers effectively the entire adult consumer population. Match rates stay consistently high—ranging from ~60% on email-only lists up to 99% with full addresses—because we prioritize real, verifiable matches over inflated vanity numbers.

Depth drives model accuracy. Predictive models are pattern-finders. A model trained on 50 data points per person can only find patterns within those 50 dimensions. A model trained on 1,400 can find combinations of signals no human would think to look for — and the more raw material the model has, the more predictive lift it can deliver. This is also why CRM-native scoring tools tend to underperform: they're working with whatever data happens to be in your system, not the full picture of who your customers are.

Coverage without depth gives you reach without insight. Depth without coverage gives you insight on too few people to matter. FIG delivers both — the difference between a model that works in theory and one that works in production.

The historical advantage

But even the deepest data falls short if it only shows you who a customer is right now.

That’s where historical (longitudinal) data comes into the picture. Most consumer data vendors give you a snapshot — who someone appears to be today, based on whatever signals were available at the moment the data was assembled. FIG v2 introduced something fundamentally different: full historical data, with every data point carrying its complete timeline of values, each one timestamped and annotated with precision and source metadata.

Instead of knowing someone's income is $85,000 today, you know it was$62,000 five years ago, $74,000 three years ago, and$85,000 today.

This sounds incremental. It isn't. Here's why it matters for predictive modeling specifically.

When you train a model to find prospects who'll convert, you're showing it examples of people who already did. The problem is that converted customers don't look the same after they convert as they did before. A home remodeling buyer, post-purchase, has a contractor transaction on their credit, possibly a HELOC, elevated home-improvement spend across categories. None of those signals existed when they were still a prospect. A model trained on current-state customer data is actually learning what people look like after the thing you care about has already happened — and then it goes hunting for prospects who have already bought.

With historical data, you can train the model on what a customer looked like just before they became a lead. The model learns the pre-purchase signature, not the post-purchase one. The result is dramatically better at finding genuine future prospects, because it's finally been shown the right examples.

This is the single sharpest differentiator between FIG and every other consumer dataset on the market. Snapshot data tells you who someone is right now. FIG tells you who they were, who they are, and the trajectory between the two. For predictive modeling, trajectory is what matters. Capturing that trajectory at scale, however, requires serious infrastructure.

How FIG is built

FIG is sourced through partnerships with leading consumer data providers and refined into a single normalized, ML-ready dataset before it ever touches a customer's data. That refinement is the part most teams underestimate when they consider building this kind of dataset in-house.

Raw consumer data from vendors arrives in inconsistent formats with overlapping fields, gaps, encoding differences, and varying definitions. Turning it into something a model can train on requires deduplication across sources, identity reconciliation, feature engineering, statistical type tagging, and ongoing quality monitoring. Faraday has been doing this work continuously for over a decade. What customers access through FIG is the output of that pipeline — pre-cleaned, pre-engineered, and ready to use from day one.

All FIG data is responsibly sourced. No third-party cookies, no social scraping, no inferred behavior. Every record is tied to a verified consumer identity — which matters for compliance with modern privacy regulations, and for accuracy. The signals that survive in a privacy-first world are the ones worth building models on.

FIG is also continuously refreshed. With the launch of FIG v2 in 2026, Faraday moved from scheduled monolithic data releases to continuous ingestion — vendor data flows into FIG as soon as it's received. For time-sensitive signals like recent movers, life events, and income changes, freshness is the difference between reaching someone at the right moment and missing the window entirely. Teams that need tighter control over update timing can pin their account to a specific FIG release and move the pin forward on their own schedule.

How FIG connects to your data

FIG is only useful if Faraday can match it accurately to the customers and leads you care about. That's the role of identity resolution.

The process works in three steps:

1. Secure matching. Faraday takes the identifiers you have — names, emails, phone numbers, addresses — and resolves them against FIG using deterministic matching across multiple PII fields. Not guesswork: actual identity matches.
2. Enrichment. Once a record is matched, the FIG data points you've selected are appended to it. You can pick individual fields or use Faraday's pre-built essentials packages curated for verticals like home services, retail, and financial services.
3. Delivery or modeling. Enriched records flow seamlessly into your existing stack — whether that's Salesforce, HubSpot, Snowflake, or Meta — or directly into a Faraday-built predictive model that generates lead scores, churn predictions, personas, or product recommendations. More integrations can be found on our integrations page.

For a deeper technical walkthrough of identity resolution — including how it handles sparse identifiers, household-level matching, and change-of-address — see how Faraday unifies your data with FIG.

What FIG unlocks

FIG is the foundation underneath nearly everything Faraday does. The same data asset powers a wide range of customer-facing capabilities:

• Predictive lead scoring — ranking leads by likelihood to convert, surfacing patterns CRM-native scoring can't see.
• Customer personas — clustering your customer base into data-driven segments built from real data points, not survey responses.
• Data appends and enrichment — purchasing specific FIG data points to fill gaps in your CRM, expand into new channels, or research a market before you commit to it.
• Market opportunity analysis — sizing addressable populations geographically before you decide where to expand.
• Product recommendations — predicting which products or services a given customer is most likely to buy next.
• Churn prediction — identifying which customers are at risk of leaving, before they leave.

Each of these is built on the same underlying graph — and the list isn't exhaustive. Any custom predictive model built on Faraday runs on FIG, so the same data asset extends to whatever outcome a team needs to predict. That's part of what makes FIG operationally efficient — brands work with one dataset that powers the entire stack instead of stitching together separate data sources for each use case.

B2C brands are seeing results like a tripled conversion rate, 22x ROI, or even CAC reduced by 67% with FIG. See our customer stories page to see more successes.

Why brands don't build this themselves

Some teams attempt to assemble this kind of data infrastructure in-house — licensing data from multiple vendors and integrating it internally. The challenges that surface, in order:

• Contracting complexity. Sourcing offline consumer data typically involves long vendor negotiations, minimum spend commitments, and overlapping coverage that's hard to reconcile.
• Security and compliance. Handling sensitive consumer data at scale requires SOC-2 certification and ongoing compliance investment.
• Regulatory overhead. State-level privacy laws — opt-outs, deletion requests, residency requirements — add operational load that grows as you scale.
• Data science overhead. Raw vendor data needs cleaning, feature engineering, bias mitigation, and ongoing maintenance before it's usable. This is often the biggest hidden cost.

Faraday handles all of this before the data reaches you. The choice isn't FIG vs. DIY — it's whether your team spends its time building and maintaining data infrastructure, or running better campaigns with infrastructure that already works.

Ready to put FIG to work?
If you want to see what FIG unlocks for your business, there are two ways to start:

• Self-serve: Browse the data catalog and purchase what you need at buy.faraday.ai.
• Enterprise: Talk to a Context Consultant — we'll walk through your data, your use case, and what FIG can do for your specific business.

FAQ

How big is the Faraday Identity Graph?

FIG covers approximately 240 million U.S. adults across 1,400+ verified consumer data points spanning demographics, financial signals, property data, life events, and lifestyle indicators.

Where does FIG data come from?

FIG is sourced through partnerships with leading consumer data providers. All data is responsibly sourced — no third-party cookies, no social scraping, no inferred behavior. Every record is tied to a verified consumer identity.

How often is FIG updated?

With FIG v2, Faraday ingests fresh data from vendors continuously as it's received, rather than waiting for scheduled releases. Customers can either accept updates automatically or pin their account to a specific FIG release for tighter change management.

What's the difference between FIG and other consumer data providers?

FIG's primary differentiator is historical (or longitudinal) data — every data point carries its complete historical timeline, which dramatically improves predictive model quality by letting models train on pre-conversion signals rather than post-conversion noise. FIG is also production-ready out of the box: normalized, feature-engineered, and accessible through Faraday's platform with no separate licensing or preprocessing required.

Can I purchase FIG data directly without using Faraday's predictive models?

Yes. All FIG data points are available and can be appended directly to any file you upload. You can browse the catalog and buy directly at buy.faraday.ai or chat through your data goals with a Context Consultant.

How does Faraday match my records to FIG?

Faraday uses deterministic identity resolution—meaning we require exact matches across name, email, phone, and address fields. We never rely on probabilistic guesswork, so you can trust that the data you receive maps exactly to the right person.