## At a glance

- **The core challenge:** Advia Credit Union knew their members needed auto loans and other financial products — but couldn't identify which members to target, when, or with what offer, and couldn't risk fair lending violations in the process.
- **The approach:** Enriched historical member data with the Faraday Identity Graph, built a propensity model trained on past auto loan outcomes, and ran bias mitigation to ensure fair treatment across protected classes before deployment.
- **The results:** Application rate climbed from 1.19% to 5.18% in 90 days — generating $2.7M in new auto loans without hiring a data scientist or building proprietary modeling infrastructure.

—

[Advia Credit Union](https://www.adviacu.org/) had a wallet share problem. Their members were financially active, but they weren't always channeling that activity through Advia. The average American holds eight financial products — but only [19% of Americans](https://www2.deloitte.com/content/dam/Deloitte/us/Documents/financial-services/us-kickingitupanotch-092614.pdf) have three or more products with their primary financial institution. Advia knew their members needed auto loans, mortgage refinancing, and debt consolidation products. What they didn't know was which members to talk to, when, and with what offer — and how to make sure they captured that business before another institution did.

Adding to the challenge, Advia's lead data scientist had recently left the organization — making it clear that building this kind of targeting capability in-house wasn't a realistic near-term option. What they needed was a way to get smarter about their existing member base without rebuilding a data science team from scratch or running afoul of fair lending regulations.

## The wallet share problem

[Credit unions and banks](/blog/banks-credit-unions-complete-context-guide) face a structural challenge that most consumer brands don't: their best growth opportunity is sitting right inside their existing member base, but activating it requires navigating a uniquely complex data and compliance environment.

First-party data tells you what a member has done with your institution. It doesn't tell you where they are in their financial life, what products they might need next, or how likely they are to act on a given offer. That missing context is what turns "grow wallet share" from a strategy into a guessing game — and what leaves most institutions relying on broad campaigns that erode member trust faster than they build revenue.

## Balancing growth with fair lending compliance

Financial services is one of the few industries where getting targeting wrong doesn't just waste budget — it creates legal exposure. The [Equal Credit Opportunity Act (ECOA)](https://www.consumerfinance.gov/compliance/compliance-resources/other-applicable-requirements/equal-credit-opportunity-act/) prohibits discrimination in credit transactions based on race, color, religion, national origin, sex, marital status, age, or receipt of public assistance. Any scoring model used for member targeting must be built and validated with these constraints in mind.

The risk isn't always intentional. Historical member data often reflects past patterns of underrepresentation, and a model trained on that data without bias mitigation can quietly perpetuate those patterns at scale. This is why many financial institutions either avoid data-driven targeting altogether or deploy it without the guardrails they need.

Faraday treats [bias mitigation](https://faraday.ai/blog/faraday-bias-mitigation-machine-learning) as a foundational step, not an afterthought. Before any model is trained, Faraday rebalances [historical](https://faraday.ai/blog/fig-v2-historical-data) data to correct for underrepresentation across demographic groups and [monitors the resulting scores for fairness across protected classes](https://faraday.ai/blog/faraday-bias-reporting). The result is a model that's both more accurate and more compliant — not a tradeoff between the two.

For Advia, this was as important as the performance numbers. A campaign that drives loan applications but creates fair lending exposure isn't a win.

## How Faraday built Advia's member scoring

Advia's auto loan campaign started with a data connection. They linked their historical member data — including records of which members had previously taken out auto loans — to the [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph): a continuously maintained dataset covering 240M U.S. adults across [1,400+ verified consumer data points](https://faraday.ai/catalog), spanning demographics, financial signals, property data, life events, and lifestyle indicators.

That enrichment step is what transforms a member record from a transaction history into a full [consumer profile](/blog/what-data-do-you-need-for-marketing). FIG adds the context that first-party data alone can't provide — household income, life-stage signals, financial behavior patterns, and more — giving the scoring model far more signal to work with.

With enriched data in place, Faraday built a [propensity model](/blog/propensity-models-predict-customer-behavior) trained on Advia's historical auto loan outcomes. The model learned the profile of members who had taken out auto loans in the past — and used that pattern to score every current member by their likelihood to apply for one now. Bias mitigation ran in parallel: Faraday identified demographic imbalances in the training data and rebalanced the model to ensure fair treatment across protected groups before any scores were deployed. Faraday validated model performance using holdout testing before deployment.

The output was a ranked list of members by auto loan propensity — with the top percentile identified as the highest-priority targets for Advia's campaign.

## The results

Advia used Faraday's scores to segment their member base and target the top percentile through [personalized](/blog/personalization-tactics-that-drive-conversions) email and direct mail campaigns (see what integrations are available for these types of campaigns on our [integrations page](https://faraday.ai/integrations)). The benchmark they were working to beat: a 1.19% auto loan application rate. And here’s what they got:

**30-day results:** Application rate reached 2.41% — more than double the baseline — generating **$1.1M in new auto loans**.

**90-day results:** The application rate continued climbing to 5.18% after 90 days, producing a total of **$2.7M in new auto loans**.

That's more than a 4x improvement in application rate, achieved without hiring a new data scientist, without building proprietary modeling infrastructure, and with a compliance posture Advia could stand behind.

## Expanding predictive scoring across the entire portfolio

The Advia story started with auto loans, but the same approach applies to any financial product a credit union or bank wants to grow. FIG enrichment, propensity scoring, and bias mitigation can be applied to mortgage refinancing, debt consolidation, credit cards, savings products, and more.

Each product becomes its own scored outcome: who among your current members is most likely to need this right now? Layered together, those scores give your team a clear view of which offer is most relevant for each member at each stage of their financial life — and the confidence to act on it without compliance risk.

For a broader look at how banks and credit unions use customer context across their member base, see our [guide to predictive AI for banks and credit unions](/blog/banks-credit-unions-complete-context-guide).

## Ready to grow wallet share?

If you want to talk through how customer context could work for your institution, [talk to a Context Consultant](https://faraday.ai/talk-to-sales). Or if you'd rather get started on your own, [try it on Faraday Pro](/signup).

Learn how Advia Credit Union used Faraday’s consumer data to get the context they needed on their member base, which allowed them to navigate fair lending compliance and generate $2.7M in new auto loans in 90 days.

How Advia Credit Union used Faraday’s data and AI to generate $2.7M in new auto loans in 90 days

## At a glance

- **The core challenge:** Boll & Branch's customer segments had no meaningful correlation to products or marketing channels — and an untested assumption that their core audience skewed older was driving significant spend toward the wrong channels.
- **The approach:** Connected their BigQuery warehouse to the Faraday Identity Graph, built data-driven personas from enriched customer data, and assigned every contact to a segment — no survey required.
- **The result:** The data revealed their highest-value customers were wealthier, settled homeowners who were heavy social media users — not radio listeners. Rebuilding campaigns around that insight drove a 30% lift in email conversion rates.

—

[Boll & Branch](https://www.bollandbranch.com/?srsltid=AfmBOoqF9IYaLXZdIdME-bCesQlrETL6zfLbHzW8Zhiq-n479qkOpmh5) knew their customers weren't all the same. The DTC linens brand had built a loyal following across a wide demographic range — but their segmentation couldn't reflect that. Their initial customer segments had no meaningful correlation to specific products or marketing channels, which made [personalization](/blog/personalization-tactics-that-drive-conversions) guesswork rather than strategy.

The problem wasn't effort. It was context. First-party transaction data tells you what customers bought. It doesn't tell you who they are — their financial situation, life stage, household makeup, or the signals that actually predict how they'll respond to a given message. Without that context, even well-intentioned personalization can only go so deep.

## The limits of traditional personas

Brands typically build [personas](https://faraday.ai/docs/abstractions/persona_sets) one of two ways. The first is the qualitative route: surveys, focus groups, which produce rich profiles but can't scale. An additional problem here is that surveys often see single-digit response rates, which means the majority of your customer base might never be assigned to a segment at all.

The other option is pre-computed [personas](/blog/what-are-ai-customer-personas) like PRIZM codes: these are universal, but generic, built on broad demographic clusters rather than patterns specific to your brand. They tend to produce segments with little variation in actual customer behavior.

Boll & Branch needed something different: a [truly personalized](/blog/personalization-tactics-that-drive-conversions) [retail](/blog/retail-ecommerce-complete-context-guide) [persona framework](/blog/what-are-ai-customer-personas) built on their own customer data, enriched with enough depth to be universally assignable across their entire audience — not just the customers who filled out a survey.

## From transaction history to full customer profiles

To start, Boll & Branch connected their [BigQuery](https://cloud.google.com/bigquery) data warehouse to the [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph).

Instead of relying solely on purchase history, they matched their first-party customer records — [names, emails, physical addresses](https://faraday.ai/blog/wheres-my-99-match-rate) — against FIG's [continuously maintained](https://faraday.ai/blog/fig-v2-freshness-release-pinning) dataset covering 240M U.S. adults across [1,400+ verified consumer](https://faraday.ai/catalog) data points. That match transformed each customer from a list of past transactions into a rich household profile: financial signals, life-stage indicators, property data, lifestyle data points, and more.

## How Faraday built Boll & Branch's personas

With that enriched foundation in place, Faraday's [clustering algorithm](https://faraday.ai/blog/k-means-clustering-customer-personas) built personas directly from Boll & Branch's own customer data — automatically identifying [three distinct segments](https://faraday.ai/blog/how-many-personas-do-you-need) based on socio-demographic patterns in the enriched dataset.

Because the personas are grounded in FIG data covering nearly every U.S. adult, they could be assigned universally — to every customer in the database, not just those who had self-reported preferences. That universal assignability is what made the personas operationally useful. Faraday pushed the persona assignments back into Boll & Branch's BigQuery warehouse, giving their analytics team the metadata they needed to cross-reference segments against their attribution system, seeing which customers came from which channels, how different personas responded to discount codes, and what products each segment over-indexed on.

## The insight that changed their strategy

Before the persona analysis, Boll & Branch assumed their core demographic skewed older — a belief that had driven significant investment in radio advertising.

[The data told a different story](/blog/retail-ecommerce-complete-context-guide). Their highest-value segment turned out to be wealthier, settled homeowners who preferred clean, modern aesthetics. And crucially, FIG data showed this cohort were heavy social media users — not radio listeners.

That single insight gave Boll & Branch the confidence to realign their marketing around this audience: from branding and product design to prospecting, engagement, and retention. Isolating the segment also enabled highly targeted focus groups, where they discovered these buyers were heavily influenced by interior design personalities — a finding that directly informed their successful partnership with designer Nate Berkus.

## The results: 30% lift in conversion rates

With every contact assigned a persona, Boll & Branch redesigned their email campaigns around each segment. Creative aligned to the socio-demographic makeup of each persona. Product recommendations drawn from what each persona actually over-indexed on. Offers calibrated to what each segment responded to.

The result was clear : a **30% lift in email conversion rates**.

And as their Chief Digital Officer, Katia Unlu, put it herself: _"Faraday is in our DNA."_

## Ready to lift conversion rates?

If you want to talk through how customer context could work for your brand, [talk to a Context Consultant](https://faraday.ai/talk-to-sales). Or if you'd rather get started on your own, [try it on Faraday Pro](/signup).

Learn how Boll & Branch used the Faraday Identity Graph to move beyond generic demographic segments, build data-driven personas that revealed a surprising high-value customer, and generate a 30% lift in email conversion rates.

How Boll & Branch achieved a 30% conversion lift with customer context

## At a glance

- **The core challenge:** A leading subscription box brand had outgrown manual curation — rigid rules, no way to test what was working, and a growing catalog that the merchandising team couldn't keep up with.
- **The approach:** Built a two-sided recommendation engine using Faraday, scoring both customer affinity and product performance simultaneously, with subscriber records enriched from the Faraday Identity Graph (1,400+ data points across 240M U.S. adults).
- **The results:** \~5% lift in revenue per customer, \~3% reduction in churn, and 5x+ monthly ROI — across four consecutive months of algorithmic recommendations outperforming manual rules.

—

In [subscription commerce](/blog/retail-ecommerce-complete-context-guide), the wrong product doesn't just go unappreciated — it cancels subscriptions. And when a customer pays for a curated experience month after month, relevance isn't just nice-to-have. It's the entire product.

The math on retention makes this concrete: it costs [5–6x](https://www.americanexpress.com/en-us/business/trends-and-insights/articles/retaining-customers-vs-acquiring-customers/) more to acquire a new customer than to keep an existing one, and returning customers spend [67%](https://www.business.com/articles/returning-customers-spend-67-more-than-new-customers-keep-your-customers-coming-back-with-a-recurring-revenue-sales-model/) more on average than first-timers. A subscriber who receives consecutive boxes that don't match their lifestyle doesn't just disengage — they [churn](https://faraday.ai/blog/churn-prevention-machine-learning). And unlike a bad email, a bad shipment is hard to ignore.

That's the challenge a leading subscription box brand set out to solve. They had outgrown their manual curation process and needed a way to introduce real customer context into their monthly product matching at scale.

## The problem: manual rules and missed opportunities

Before working with Faraday, this brand was manually assigning monthly boxes using written rules. The logic was informed by surface-level customer traits and preferences, but it was fundamentally rigid and prone to human bias. As the catalog and subscriber base grew, the cracks showed fast: the merchandising team found themselves buried under conflicting rules, manually balancing inventory constraints against subscriber preferences with no reliable way to test whether any of it was actually working.

They were operating on instinct rather than context — and leaving significant lifetime value on the table.

## Bridging the context gap with FIG

To move beyond rigid rules, the brand needed to stop treating subscribers as simple profiles and start understanding them as fully realized households. First-party transaction data tells you what a customer bought. It doesn't tell you their household makeup, financial trajectory, or broader lifestyle affinities — the signals that actually predict what they'll want next month.

To fill that gap, the brand enriched their subscriber records using the [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph): a continuously maintained dataset covering 240M U.S. adults across [1,400+ verified consumer data points](https://faraday.ai/catalog), spanning demographics, financial signals, property data, life events, and lifestyle indicators.

Crucially, FIG also provides [historical data](https://faraday.ai/blog/fig-v2-historical-data) — not just a current-state snapshot, but a full timeline for each data point. The model doesn't just know a household's income today; it knows how their lifestyle and spending patterns have shifted over the last five years. That trajectory is what lets a recommendation engine find combinations of signals no human rule-maker would ever think to look for.

## The methodology: two-sided matching

With the enriched data foundation in place, Faraday implemented a two-sided matching framework — a meaningful departure from traditional recommendation engines that rank products by overall popularity.

Instead of one score, the model evaluates two vectors simultaneously:

- **Customer affinity:** How likely a specific person is to engage with a product
- **Product performance:** How well that product has performed with similar audiences

By evaluating both sides, the system predicts the mutual affinity between a person and a product — not just "is this product popular?" but "is this product right for this person, given everything we know about them and everyone like them?"

Behind the scenes, this is powered by explainable [propensity models](/blog/propensity-models-predict-customer-behavior). The practical payoff for a brand: you can [see exactly](https://faraday.ai/blog/faraday-score-explainability) which data points are driving each product match — household signals, life events, spending trajectory — rather than trusting a black box.

### Tuning for business reality

A common failure point for recommendation engines is ignoring operational constraints. An algorithm might identify the perfect product for a large cohort, but the warehouse might not have the inventory to support it.

Faraday solves this by letting the merchandising team manage global program constraints — inventory availability, product diversity, subscription cadence — rather than writing individual rules. The result is recommendations that are both deeply [personalized](/blog/personalization-tactics-that-drive-conversions) and operationally feasible.

## The results

To validate the approach, the brand ran a clean [A/B test](https://faraday.ai/blog/holdout-testing-with-faraday) comparing Faraday's recommendations against their manual rules. And the results were clear:

- **\~5% lift** in revenue per customer
- **\~3% reduction** in overall churn rate
- **5x+ monthly ROI** generated with Faraday
- **Four consecutive months** of algorithmic recommendations outperforming manual rules

These weren't marginal gains — they were statistically significant improvements that validated the shift away from manual curation. And because the models improve as more data flows through them, these aren't ceiling numbers. They're a starting point.

For a deeper look at how to predict and prevent churn across your subscriber base, see our [complete guide to churn prediction](https://faraday.ai/blog/churn-prevention-machine-learning).

## Ready to optimize your recommendations?

If you want to talk through how a two-sided recommender could work for your subscription business, [talk to a Context Consultant](https://faraday.ai/talk-to-sales). Or if you'd rather get started on your own, [try it on Faraday Pro](/signup).

Learn how a leading subscription box company replaced rigid manual product rules with a two-sided recommendation engine built on Faraday's customer context — boosting revenue per customer, reducing churn, and generating 5x monthly ROI.

How a leading subscription box brand boosted revenue and reduced churn with customer context

## At a glance

- **The core challenge:** An established auto dealer marketing platform was building AI-powered audience features but had no visibility into who car buyers actually were — only what happened inside dealerships.
- **The integration:** Embedded Faraday's API across 1,000+ dealer accounts, provisioning a dedicated prediction environment for each dealer with custom propensity models trained on their own transaction history.
- **The business impact:** Audience generation now runs programmatically across every connected account with automated monthly retraining — no data science team or manual intervention required.

—

Every [platform](/blog/powered-by-faraday-complete-guide-embedded-predictive-ai) today knows it needs better data and smarter AI tools. But the question is how to get it.

For [newer companies](/blog/real-time-automotive-lead-enrichment-faraday), the answer is often baked in from day one — consumer intelligence built into the architecture before the first customer signed on. But for established platforms — ones that have been delivering real value for years, with hundreds or thousands of customers already depending on them — the path looks different. You can't rebuild from scratch. You need to add intelligence to what already works.

That tension is at the center of what Faraday calls the [context gap](/blog/consumer-context-data-ai): brands know how their customers behave inside their product, but they don't know who those customers actually are. Their financial picture. Their life stage. Whether they're likely to be in-market right now. That kind of consumer context doesn't live in a SaaS platform's database. It doesn't come from a CRM feed. And building the infrastructure to generate it — the data sourcing, the modeling, the ongoing maintenance — is simply off the table for most established teams.

That's exactly where a leading automotive software and marketing technology company found itself. Years of real customers, real trust, and real product depth — a system that worked. But as they built out their AI-powered audience features, they kept running into the same wall: their product knew everything about what happened inside a dealership, and almost nothing about the real people interested in buying cars — beyond their clicks and views.

## Why do established SaaS platforms face a "Consumer context gap"?

The instinct for many platforms at this point is to go [get the data themselves](/blog/powered-by-faraday-complete-guide-embedded-predictive-ai). But sourcing production-ready consumer data isn't a weekend project. It means negotiating with data vendors, normalizing and unifying records across sources, building the modeling infrastructure to turn raw data points into useful predictions, and then maintaining all of it on an ongoing basis as data ages and customers change. Most established platforms don't have a data science team sitting idle waiting for that project. And even if they did, it would take years to reach the quality and coverage that purpose-built consumer data infrastructure can deliver from day one.

Building it in-house wasn't the answer. Embedding it was.

## The solution: Faraday as the embedded context layer

Rather than rebuilding from scratch, the platform embedded Faraday directly into its product architecture via [API](https://faraday.ai/docs/reference).

Faraday's role was to supply what the platform was missing: real consumer context. That means access to [the](https://faraday.ai/catalog) [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph), which contains [1,400+ data points](https://faraday.ai/catalog) on 240M+ U.S. adults — wealth indicators, life stage signals, behavioral data points, geographic context. But [consumer profiles](/blog/what-data-do-you-need-for-marketing) alone aren't enough to build smart audiences. What dealers actually need to know isn't just who their customers are — it's who among the thousands of consumers in their market is most likely to be ready to buy right now. So on top of that consumer context layer, Faraday builds custom [predictive models](/blog/propensity-models-predict-customer-behavior) trained on each dealership's own transaction history — turning static profiles into forward-looking propensity scores that power every audience the platform generates.

Critically, this wasn't a one-size-fits-all enrichment layer. The platform operates at the individual dealer level, and Faraday mirrors that structure. Using Faraday's multi-account API architecture, the platform provisions a dedicated prediction environment for each dealer — its own data connections, its own cohorts, its own trained outcomes. Audiences are built on models that understand what a buyer looks like for that dealer, in that market — not a generic consumer profile applied uniformly across accounts. And because franchise businesses can only sell to customers they can actually reach, each dealer's audiences are geofenced to their specific area, so targeting never spills outside the radius they can serve.

When an account's data updates — new transactions, new leads, new service visits — Faraday retrains. The consumer context layer stays current automatically, with no manual intervention required.

## What changed

Before Faraday, the platform's AI audiences were built from the inside out: start with what the dealer knows, and try to find patterns. After Faraday, audiences are built from both directions — the dealer's first-party signals enriched and extended by real-world consumer context that no CRM can generate on its own.

For the platform, the operational effects were meaningful:

- Audience generation, which previously required manual rules-based configuration, now runs programmatically across every connected account
- Monthly data refreshes trigger automated model retraining — no human intervention required to keep predictions current
- Every audience the platform generates now reflects real consumer behavior — not just dealership CRM activity.

None of this required the platform to hire data scientists or build modeling infrastructure. Faraday handled that complexity behind the API.

### Key results

- Audience generation automated across 1,000+ dealer accounts
- Monthly model retraining runs with no human intervention
- Every audience now reflects real consumer behavior, not just CRM activity

## The bigger pattern: why established platforms are embedding consumer intelligence

This case reflects something happening across B2B software right now. Established platforms are under real pressure to make their AI claims real — not just in the product copy, but in the actual outputs. And the honest answer is that most AI features, left to first-party data alone, produce mediocre results. The data simply isn't there.

What Faraday offers platforms in this position isn't a dataset to license. It's a consumer intelligence infrastructure they can embed without rebuilding their product. The AI story stays theirs. The context that makes it credible comes from Faraday.

For a platform serving hundreds of accounts, each with their own customers and markets, that's the difference between AI as a feature name and AI as a real capability.

## Ready to optimize your platform?

If you want to talk through how customer context and embedded AI could work for your platform, [talk to a Context Consultant](https://faraday.ai/talk-to-sales) or explore our [developer documentation](https://faraday.ai/docs) to see how easily we could integrate into your existing stack. Or if you'd rather get started on your own, [try it out yourself on Faraday Pro](/signup).

Learn how a leading auto dealer marketing platform closed the consumer context gap in its AI product — embedding Faraday to power individualized predictions for every account it serves, without building a data science team.

How an established marketing platform delivered AI-powered audiences across 1,000+ accounts

## At a glance

- **The core challenge:** A real-time automotive sales intelligence platform needed consumer context for its AI co-pilot — without the six-figure costs and months of engineering required by traditional data brokers.
- **The tech stack integration:** Embedded Faraday's Lookup API directly into the co-pilot's CRM lead intake flow.
- **The business impact:** Achieved instant access to the Faraday Identity Graph (1,400+ attributes across 240M+ U.S. adults), dropping deployment time from months to days while maintaining strict privacy compliance.

—

Every [platform](/blog/powered-by-faraday-complete-guide-embedded-predictive-ai) today knows it needs better data. The question is how to get it — and how fast.

That's the position one automotive AI startup found themselves in. Their product is a real-time AI co-pilot for automotive sales teams. When a lead comes in — someone who filled out a form on a dealer's website, or clicked through from a listing — the co-pilot builds a psychological profile and generates a live call script on the rep's screen before they dial. The rep knows how to open the conversation, which questions to ask, how to handle objections, and what the person on the other end of the line is likely to respond to.

The product works because it knows who that lead actually is. And getting that consumer intelligence quickly enough to ship a working product was the problem the team needed to solve.

## The challenge: scaling real-time consumer Data enrichment for AI co-pilots

The platform's value proposition depends entirely on the quality of the [consumer profile](/blog/what-data-do-you-need-for-marketing) it can build in the seconds between a rep receiving a lead who submitted info on a dealership's site and dialing them back. The team had access to credit bureau data from [Equifax](https://www.equifax.com/) — useful for financial signals — but it was missing the consumer preferences, life events, lifestyle indicators, and psychographic signals that make a profile genuinely actionable.

The traditional path to filling that gap runs through data brokers. But for a startup, that path comes with real costs: [six-figure upfront licensing fees, months of engineering work to build ETL infrastructure, identity resolution pipelines, and ongoing privacy compliance overhead](/blog/powered-by-faraday-complete-guide-embedded-predictive-ai). And at the end of it, a static dataset with no [real-time API](https://faraday.ai/docs/features/lookup-api) — which meant limited use for a product that needs to enrich a lead in near-real-time, while a rep is about to make a call.

For a company focused on signing dealers and validating the product, that tradeoff didn't make sense.

### Faraday vs. traditional data brokers

| Feature / metric            | Traditional data Brokers              | Faraday lookup API integration           |
| :-------------------------- | :------------------------------------ | :--------------------------------------- |
| **Upfront costs**           | Six-figure licensing fees             | Scalable API pricing                     |
| **Implementation timeline** | Months of engineering (ETL/pipelines) | Live in days                             |
| **Data delivery**           | Static datasets (No real-time API)    | Near-instantaneous, real-time enrichment |
| **Compliance overhead**     | Manual privacy tracking               | Built-in compliance infrastructure       |

## The solution: Embedding Faraday’s Lookup API for instant identity resolution

The team embedded Faraday directly into the co-pilot's lead intake flow via the [Lookup API](https://faraday.ai/docs/features/lookup-api).

When a lead arrives in a dealer's CRM, the co-pilot fires a real-time call to Faraday. Within seconds, it gets back a consumer profile drawn from the [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph) — [1,400+ data points on 240M+ U.S. adults, covering demographics](https://faraday.ai/catalog), financial signals, lifestyle indicators, life events, and household context. That profile feeds directly into the co-pilot's AI, which uses it to generate the psychological profile and call script the rep sees on screen.

Privacy compliance, identity resolution, and data infrastructure came included. The engineering team focused on building the product, not on the data plumbing underneath it.

## What changed: real consumer context on every lead, in real time

Before Faraday, the co-pilot’s profiles were built primarily from credit bureau signals and whatever the CRM had on the lead. Useful, but incomplete — missing the consumer dimensions that actually shape how someone responds in a conversation.

With Faraday, every lead that comes through a connected dealership arrives enriched with a full consumer profile before anyone picks up the phone. The rep sees not just financial signals but lifestyle context, life stage indicators, and psychographic patterns — enough context to shape how a rep opens the conversation and handles objections.

For the platform, the operational results were what mattered at this stage:

- Lead enrichment that previously would have required months of infrastructure work was live in days
- A real-time pipeline that works within the seconds-level latency the co-pilot requires
- Built-in compliance infrastructure that would have taken significant engineering time to replicate

## The bigger pattern: consumer context as infrastructure, not a dataset

This startup's story is a version of a problem that shows up across the software industry. Whether you're building from scratch or retrofitting an established product, the moment you try to make your AI or your agents genuinely intelligent about the people they're engaging, you hit the same wall: the consumer context you need doesn't exist in your system, and building the infrastructure to generate it is a distraction from building the product itself.

Faraday’s consumer intelligence infrastructure acts as a plug-and-play data layer that handles identity resolution, data normalization, real-time API delivery, and privacy compliance out of the box.

For this team, that meant going from concept to production in weeks instead of months. For the dealers using the platform, it means every rep on every call has a complete picture of who they're talking to before the first word is spoken.

## Ready to optimize your platform?

If you want to talk through how customer context could work for your platform, [talk to a Context Consultant](https://faraday.ai/talk-to-sales) or explore our [developer documentation](https://faraday.ai/docs) to see how the integration works. Or if you'd rather get started on your own, try it out yourself on [Faraday Pro](/signup).

Learn how a real-time automotive sales intelligence platform embedded Faraday's consumer data into its AI co-pilot — giving automotive BDC reps a real-time psychological profile and call script for every incoming lead, without building data infrastructure from scratch.

How a real-time automotive sales platform embedded consumer intelligence — without building data infrastructure

"If one record is wrong, how can I trust the rest?" 

We sometimes hear this from clients, and we get why. You open a freshly enriched customer file, spot something you know is wrong — your own name at an address you left five years ago, a coworker tagged with an income that's clearly outdated — and suddenly the whole file feels suspect.

But that instinct — to judge the whole file by one bad record — is exactly the **wrong way to evaluate consumer data**. And it leads to worse decisions, not better ones.

## Spot-checking is the wrong test for data quality

First things first: spot checking for 100% accuracy is not a valid way to review data quality in a customer file.

Whether you have spotted an incorrect record in a 100k person audience for direct mail or you've identified an incorrect attribute from your customer file, the key is to go back to first principles: your ultimate goal is likely to drive an increase in conversion rates or reach more people. One bad row isn't going to spoil the bunch.No dataset at scale is 100% accurate, however, our mission is to have the best data set on the market. We rely on many upstream vendors and getting feedback on their data quality only helps us get better at our mission. 

## What to do when you spot a bad record

When you find one bad record, here’s what to do: 

- First, consider your use for the data. If the file is 95% accurate instead of 100% accurate, can you still achieve your goal?  
- Second, report it to us via [faraday.ai/data-corrections](https://faraday.ai/data-corrections). When a client flags something, the issue almost always originates with one of our upstream vendors — but because it ended up in your file, it's our job to act on it. Our team reviews every submission, escalates with the upstream vendor when needed, and applies corrections back into the pipeline. It's not a quality guarantee — it's a correction loop, and it's how the dataset stays trustworthy over time.

Our number one goal is to provide data that drives meaningful lift on your campaigns, and ultimately delivers significant ROI for your company. (Side note: we strive to be assessed by the ROI we deliver, check out some of our [case studies](/stories) to see just how much lift we generate). 

The reality is that chasing 100% accuracy at this scale would mean slower files, longer QA cycles, higher prices, and less freshness — all in service of a number that doesn't change the campaign outcome. So the right question isn't "is every record correct?" — it's "does this file help me reach more of my best customers than a random list would?"

## How to actually evaluate consumer data

If you want to know whether your file is doing its job:

* **Run a holdout test.** Pick a random subset of your enriched audience and a comparable subset without enrichment. Run the same campaign to both. The lift between them is your real answer.  
* **Look at downstream metrics.** Response rate, conversion rate, CPA, ROAS. These are the only metrics that reflect whether the file is working.  
* **Evaluate match rates in context.** Match rates depend on what identifiers you bring to the table (check out our [thoughts on this](https://faraday.ai/blog/wheres-my-99-match-rate)). Compare against realistic benchmarks, not vanity numbers.

None of these require inspecting individual rows. All of them tell you something a spot-check cannot: whether the file is actually delivering value.

## The bottom line

It’s just like life, don’t let one data point throw you off your game\! Stay focused on measuring what matters: response rate, CPA, lift versus a random list. If the numbers clear your benchmarks, the file did its job. If they don't, that's the signal worth investigating — not the one weird row that caught your eye.

When you do find a record that's off, [flag it](/data-corrections). We want to know. And if your campaign results aren't where they should be, [talk to us](/talk-to-sales). That's the conversation we want to be having.

When you spot a bad record in a 150,000-person file, your instinct is to question the whole dataset. The math says otherwise. Here's how to actually evaluate data quality.

Found a wrong record in your data? Here's what it actually means

**See our complete financial services guides:**
- [Predictive AI for banks and credit unions](/blog/banks-credit-unions-complete-context-guide)
- [Predictive AI for financial services (debt, mortgage, lending)](/blog/debt-mortgage-lending-complete-context-guide)
- [Predictive AI for insurance brands](/blog/insurance-complete-context-guide)
---

If you’re leading marketing at a financial services brand—credit unions, regional banks, fintech lenders, mortgage providers—you already know how critical data is to keeping your funnel moving. You’ve built strong programs to generate leads, pre-qualify applicants, and feed loan officers. But there’s always another level of efficiency and ROI to unlock when you have richer context on the households you’re targeting.

Think about prioritizing applications in the call center: better data means fewer unqualified prospects, more productive reps, and stronger portfolio growth. That’s where Faraday comes in.

## The Faraday data platform

To take your efforts further, you don’t just need more data—you need the right data, delivered in a way that’s immediately usable across your CRM, loan origination system, and campaigns.

That’s where [Faraday’s three-part data platform](/blog/what-data-do-you-need-for-marketing) comes in:

- **Identity completion**: [Fill in the blanks](/blog/what-data-do-you-need-for-marketing#layer-1-identity-completion-building-the-contact-foundation) with verified names, addresses, emails, and phone numbers so every record maps to a real household.
- **Consumer data**: [Enrich your customer records](/blog/what-data-do-you-need-for-marketing#layer-2-consumer-profile-enrichment-closing-the-context-gap) with the [Faraday Identity Graph (FIG)](https://faraday.ai/blog/whats-in-faraday-identity-graph), which includes over 1,400 responsibly sourced [datapoints](https://faraday.ai/catalog) covering 240M U.S. adults and households. These datapoints span demographics, lifestyle, property, and behavioral attributes that make segmentation, targeting, and prioritization easy—even before layering on custom predictions.
- **Custom predictive datapoints**: [AI-trained scores](https://faraday.ai/blog/predictive-lead-scoring-b2c) [built on your own records](/blog/what-data-do-you-need-for-marketing#layer-3-custom-predictive-data-anticipating-customer-next-action)—like Likelihood to Refinance, Default Risk Proxy, or ValueScore—so you can prioritize and personalize at scale.

## Essential data packages

One way we make the consumer data layer easier to use is by curating **industry essential data packages**: sets of 20 datapoints that we’ve determined to be most predictive for your industry. We identify these datapoints using our Faraday [predictor leaderboard](/leaderboard), which ranks datapoints by their predictive power in each particular industry vertical. For financial services, these packages are tuned to help you qualify applicants, segment portfolios, and drive smarter outreach.

- **How they’re chosen**: We analyze all 1,400+ datapoints in the FIG and surface the ones with the strongest predictive signals for your vertical. That way you get the attributes that matter most, without paying for data you don’t need.

- **How they’re delivered**: Essential data packages show up just like any other Faraday datapoint—directly appended to your records in your CRM, call center platform, or lead management system in [real time through our API or in scheduled batch](/blog/real-time-vs-batch-which-drives-roi). No extra setup required.

- **How they’re different**: Most data vendors hand you a pile of raw attributes. Faraday packages are curated, validated, and aligned to your outcomes. Our data is high-quality, responsibly sourced, and refreshed regularly—and we take on the work of figuring out what’s actually useful for you.

## The financial services essential data package

Here’s what’s included in the standard package for financial services brands:

| Data Trait | Description |
| :----------------------------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [Is millennial](https://faraday.ai/catalog/v2/is_millennial) | Indicates whether a household contains an adult from the Millennial generation |
| [Shopping styles](https://faraday.ai/catalog/v2/shopping_styles) | Segments households by preferred shopping behavior—online, in-store, luxury, deal-driven, and more—based on transactions, media habits, and survey data |
| [Household income](https://faraday.ai/catalog/v2/household_income) | Estimated annual household income, based on modeled narrow- and broad-band income ranges |
| [Gender](https://faraday.ai/catalog/v2/gender) | Gender of the individual |
| [Dwelling type](https://faraday.ai/catalog/v2/dwelling_type) | Structural classification of the residence (single-family, duplex, townhouse, apartment, mobile home) |
| [Child presence in household indicator](https://faraday.ai/catalog/v2/child_presence_in_household_indicator) | Indicates whether a child is present in the household |
| [Family composition](https://faraday.ai/catalog/v2/family_composition) | Classifies the household by the marital status of adults and the presence or absence of children |
| [Miscellaneous online topic spending interest](https://faraday.ai/catalog/v2/miscellaneous_online_topic_spending_interest) | Flags households with reported online spending in topics not covered by other interest categories |
| [Age](https://faraday.ai/catalog/v2/age) | The age of the individual |
| [Is gen x](https://faraday.ai/catalog/v2/is_gen_x) | Indicates whether a household contains an adult from Generation X (born 1965–1980) |
| [Is baby boomer](https://faraday.ai/catalog/v2/is_baby_boomer) | Indicates whether a household contains an adult from the Baby Boomer generation (born 1946–1964) |
| [Home market value change direction](https://faraday.ai/catalog/v2/home_market_value_change_direction) | Indicates whether a household's estimated home market value has recently increased or decreased |
| [Individual marital status](https://faraday.ai/catalog/v2/individual_marital_status) | Marital status |
| [Purchased via internet](https://faraday.ai/catalog/v2/purchased_via_internet) | Likely to make purchases via the internet channel |
| [Credit card payment indicator](https://faraday.ai/catalog/v2/credit_card_payment_indicator) | Indicates whether the individual has made a purchase using a credit card |
| [Household income change direction](https://faraday.ai/catalog/v2/household_income_change_direction) | Indicates the household income has had an increase or decrease |
| [Childrens spending quintile](https://faraday.ai/catalog/v2/childrens_spending_quintile) | Ranks an individual into five tiers based on spending on children's products including apparel, toys, furniture, games, and educational materials |
| [New adult to file indicator](https://faraday.ai/catalog/v2/new_adult_to_file_indicator) | Indicates a new young adult has recently been added to the household record—an early signal of first-time credit or account activity |
| [Income tier derivation](https://faraday.ai/catalog/v2/income_tier_derivation) | Indicates the type of information used to derive a household's income tier assignment |
| [Low to median household income pct](https://faraday.ai/catalog/v2/low_to_median_household_income_pct) | Estimates the share of households in the surrounding postcode earning at or below 80% of the postcode median household income |

These datapoints give you immediate visibility into whether a lead or applicant is likely to be a fit for your business, before your loan officers ever pick up the phone.

## Why this matters

When you add these essentials to your leads and customer records, you:

- **Save money** by rejecting unqualified applicants earlier in the funnel.
- **Increase conversions** by prioritizing the right households for your products.
- **Boost efficiency** by making sure your reps and loan officers aren’t wasting time on poor fits or low-probability approvals.
- **Prove ROI** to the CFO with data-driven evidence that marketing spend is driving real portfolio growth.

## Smarter data, real results

Industry essential data packages are how Faraday makes your records AI-ready, without heavy lifting from your team. Combined with verified identity completion and custom predictive datapoints, they give you the full modern data stack your financial services brand needs to compete.

Ready to see how these datapoints can transform your funnel? [Get started on Faraday Pro](/signup) for self-serve access, or [talk to a Context Consultant](/contact) if you're evaluating for a larger contract or want to talk through activation angles.

Faraday’s financial services essential data package delivers curated, high-impact datapoints that help banks, credit unions, and fintechs qualify applicants faster, prioritize the right households, and drive portfolio growth.

Just the essentials: get the datapoints financial services brands need to make better decisions

**See our complete guide:** [Predictive AI for home services brands](/blog/home-services-complete-context-guide).

---

If you're leading marketing at a home services brand—kitchen, bath, remodeling, roofing, solar—you already know how critical data is to keeping the funnel moving. You've built strong programs to generate leads, feed the call center, and keep installers busy. But there's always another level of efficiency and ROI to unlock when you have richer context on the households you're targeting.

Think about qualifying appointments in the call center: better data means fewer no-shows, more productive reps, and more installs. That's where Faraday comes in.

## The Faraday data platform

To take your efforts further, you don't just need _more_ data—you need the _right_ data, delivered in a way that's immediately usable across your CRM, call center, and campaigns. That's where Faraday's three-part data platform comes in:

- **Identity completion**: [Fill in the blanks](/blog/what-data-do-you-need-for-marketing#layer-1-identity-completion-building-the-contact-foundation) with verified names, addresses, emails, and phone numbers so every record maps to a real household.
- **Consumer data**: [Enrich your customer records](/blog/what-data-do-you-need-for-marketing#layer-2-consumer-profile-enrichment-closing-the-context-gap) with the [Faraday Identity Graph (FIG)](https://faraday.ai/blog/whats-in-faraday-identity-graph), which includes over 1,400 responsibly sourced [datapoints](https://faraday.ai/catalog) covering 240M U.S. adults and households. These datapoints span demographics, lifestyle, property, and behavioral attributes that make segmentation, targeting, and prioritization easy—even before layering on custom predictions.
- **Custom predictive datapoints**: [AI-trained scores](https://faraday.ai/blog/predictive-lead-scoring-b2c) [built on your own records](/blog/what-data-do-you-need-for-marketing#layer-3-custom-predictive-data-anticipating-customer-next-action)—like Likelihood-to-Purchase, Best-First-Product, or Credit Score Proxy—so you can prioritize and personalize at scale.

## Essential data packages

One way we make the consumer data layer easier to use is by curating industry essential data packages: sets of 20 datapoints that we've determined to be most predictive for your industry. We identify these datapoints using our Faraday predictor [leaderboard](/leaderboard), which ranks datapoints by their predictive power in each particular industry vertical. For home services, these packages are tuned to help qualify leads, score calls, and prioritize installs.

**How they're chosen:** We analyze all 1,400+ datapoints in the FIG and surface the ones with the strongest predictive signals for your vertical. That way you get the attributes that matter most, without paying for data you don't need.

**How they're delivered:** Essential data packages show up just like any other Faraday datapoint—directly appended to your records in your CRM, call center platform, or lead management system in real time through our api or in scheduled batch. No extra setup required.

**How they're different:** Most data vendors hand you a pile of raw attributes. Faraday packages are curated, validated, and aligned to your outcomes. Our data is high-quality, responsibly sourced, and refreshed regularly—and we take on the work of figuring out what's actually useful for you.

## The home services essential data package

Here's what's included in the standard package for home services brands:

| Data Trait | Description |
| :----------------------------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [Shopping styles](https://faraday.ai/catalog/v2/shopping_styles) | Segments households by preferred shopping behavior—online, in-store, luxury, deal-driven, and more—based on transactions, media habits, and survey data |
| [Is baby boomer](https://faraday.ai/catalog/v2/is_baby_boomer) | Indicates whether an adult born between 1946–1964 (the Baby Boomer generation) is present in the household |
| [Family composition](https://faraday.ai/catalog/v2/family_composition) | Classifies the household by the marital status of adults and the presence or absence of children |
| [Purchased via sweepstakes](https://faraday.ai/catalog/v2/purchased_via_sweepstakes) | Likely to make purchases via the sweepstakes channel |
| [Gender](https://faraday.ai/catalog/v2/gender) | Gender of the individual |
| [Mail order buyer interest](https://faraday.ai/catalog/v2/mail_order_buyer_interest) | Likely to make purchases via the mail order channel |
| [Miscellaneous online topic spending interest](https://faraday.ai/catalog/v2/miscellaneous_online_topic_spending_interest) | Flags households with reported online spending in topics not covered by other interest categories |
| [Individual marital status](https://faraday.ai/catalog/v2/individual_marital_status) | Marital status |
| [Facebook interest](https://faraday.ai/catalog/v2/facebook_interest) | Flag for household interest in or active use of Facebook, derived from social media engagement signals |
| [Dwelling type](https://faraday.ai/catalog/v2/dwelling_type) | Structural classification of the residence (single-family, duplex, townhouse, apartment, mobile home) |
| [Low to median household income pct](https://faraday.ai/catalog/v2/low_to_median_household_income_pct) | Estimates the share of households in the surrounding postcode earning at or below 80% of the postcode median household income |
| [Household income](https://faraday.ai/catalog/v2/household_income) | Estimated annual household income, based on modeled narrow- and broad-band income ranges |
| [High dollar belief causes donor likelihood](https://faraday.ai/catalog/v2/high_dollar_belief_causes_donor_likelihood) | Ranks likelihood to donate $500 or more annually to religious causes |
| [Income tier derivation](https://faraday.ai/catalog/v2/income_tier_derivation) | Indicates the type of information used to derive a household's income tier assignment |
| [Home year built](https://faraday.ai/catalog/v2/home_year_built) | Four-digit year the primary building on the property was originally constructed |
| [Available home equity amount](https://faraday.ai/catalog/v2/available_home_equity_amount) | Current market value of the home minus outstanding mortgage balance or liens |
| [Prominent household profession](https://faraday.ai/catalog/v2/prominent_household_profession) | The most prominent profession associated with the household |
| [Child presence in household indicator](https://faraday.ai/catalog/v2/child_presence_in_household_indicator) | Indicates whether a child is present in the household |
| [Credit score proxy](https://faraday.ai/catalog/v2/credit_score_proxy) | Groups individuals into 12 ranks across 5 tiers (Best to Poor) that approximate credit score; rank 1 (~760) is best, rank 12 (~590) is lowest |
| [Credit card payment indicator](https://faraday.ai/catalog/v2/credit_card_payment_indicator) | Indicates whether the individual has made a purchase using a credit card |

These datapoints give you immediate visibility into whether a lead is likely to be a fit for your business, before your sales team ever picks up the phone.

## Why this matters

When you add these essentials to your leads and customer records, you:

- **Save money** by rejecting unqualified leads earlier in the funnel.
- **Increase conversions** by prioritizing the right households.
- **Boost efficiency** Make sure your reps aren't wasting time on unqualified leads, low-fit or no show appointments.
- **Prove ROI** to the CFO with data-driven evidence that marketing spend is driving installs and revenue.

## Closing: smarter data, real results

Industry essential data packages are how Faraday makes your records AI-ready, without heavy lifting from your team. Combined with verified identity completion and custom predictive datapoints, they give you the full modern data stack your home services brand needs to compete.

Ready to see how these datapoints can transform your funnel? [Get started on Faraday Pro](/signup) for self-serve access, or [talk to a Context Consultant](/contact) if you're evaluating for a larger contract or want to talk through activation angles.

Faraday’s essential home-services data package gives brands 20 curated datapoints—like household income, property type, and credit score proxy—to qualify leads, cut waste, and boost installs.

Just the essentials: get the datapoints home services brands need to make better decisions

**See our complete guide:** [Predictive AI for e-commerce and retail brands](/blog/retail-ecommerce-complete-context-guide).

---

If you’re leading marketing at a retail or e-commerce brand—like apparel, furniture, or subscription boxes—you already know how critical data is to keeping the funnel moving. You’ve built strong programs to drive traffic, grow carts, and keep customers coming back. But there’s always another level of efficiency and ROI to unlock when you have richer context on the households you’re targeting.

Think about segmentation for campaigns: better data means fewer wasted impressions, more personalized offers, and higher conversion rates. That’s where Faraday comes in.

## The Faraday data platform

To take your efforts further, you don’t just need more data—you need the right data, delivered in a way that’s immediately usable across your CRM, ESP, and ad platforms. That’s where Faraday’s three-part data platform comes in:

- **Identity completion**: [Fill in the blanks](/blog/what-data-do-you-need-for-marketing#layer-1-identity-completion-building-the-contact-foundation) with verified names, addresses, emails, and phone numbers so every record maps to a real household.
- **Consumer data**: [Enrich your customer records](/blog/what-data-do-you-need-for-marketing#layer-2-consumer-profile-enrichment-closing-the-context-gap) with the [Faraday Identity Graph (FIG)](https://faraday.ai/blog/whats-in-faraday-identity-graph), which includes over 1,400 responsibly sourced [datapoints](https://faraday.ai/catalog) covering 240M U.S. adults and households. These datapoints span demographics, lifestyle, property, and behavioral attributes that make segmentation, targeting, and prioritization easy—even before layering on custom predictions.
- **Custom predictive datapoints**: [AI-trained scores](https://faraday.ai/blog/predictive-lead-scoring-b2c) [built on your own records](/blog/what-data-do-you-need-for-marketing#layer-3-custom-predictive-data-anticipating-customer-next-action)—like Likelihood-to-Purchase, Best-First-Product, or Credit Score Proxy—so you can prioritize and personalize at scale.

## Essential data packages

One way we make the consumer data layer easier to use is by curating industry essential data packages: sets of just the datapoints that we’ve determined to be most predictive for your industry. We identify these datapoints using our Faraday predictor [leaderboard](/leaderboard), which ranks datapoints by their predictive power in each particular industry vertical. For retail/e-commerce, these packages are tuned to help segment audiences, tailor offers, and boost conversions.

**How they’re chosen**: We analyze all 1,400+ datapoints in the FIG and surface the ones with the strongest predictive signals for your vertical. That way you get the attributes that matter most, without paying for data you don’t need.

**How they’re delivered**: Essential data packages show up just like any other Faraday datapoint—directly appended to your records in your CRM, ESP, or ad platform in real time through our API or in scheduled batch. No extra setup required.

**How they’re different**: Most data vendors hand you a pile of raw attributes. Faraday packages are curated, validated, and aligned to your outcomes. Our data is high-quality, responsibly sourced, and refreshed regularly—and we take on the work of figuring out what’s actually useful for you.

## The retail/e-commerce essential data package

Here’s what’s included in the standard package for retail & e-commerce brands:

| Data Trait | Description |
| :----------------------------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [Fashion accessories and beauty spending quintile](https://faraday.ai/catalog/v2/fashion_accessories_and_beauty_spending_quintile) | Average dollars spent on the fashion accessories and beauty category, bucketed into 5 quintile groupings |
| [Fashion accessories and beauty purchase recency](https://faraday.ai/catalog/v2/fashion_accessories_and_beauty_purchase_recency) | Recency of purchase in the fashion accessories and beauty category |
| [Gender](https://faraday.ai/catalog/v2/gender) | Gender of the individual |
| [Online amount spent on purchases quintile](https://faraday.ai/catalog/v2/online_amount_spent_on_purchases_quintile) | Total dollars spent on online purchases within lifetime activity, bucketed into 5 quintile groupings |
| [Shopping styles](https://faraday.ai/catalog/v2/shopping_styles) | Segments households by preferred shopping behavior—online, in-store, luxury, deal-driven, and more—based on transactions, media habits, and survey data |
| [Miscellaneous online topic spending interest](https://faraday.ai/catalog/v2/miscellaneous_online_topic_spending_interest) | Flags households with reported online spending in topics not covered by other interest categories |
| [Purchased via internet](https://faraday.ai/catalog/v2/purchased_via_internet) | Likely to make purchases via the online channel |
| [Home market value](https://faraday.ai/catalog/v2/home_market_value) | Modeled market value of the property in dollars |
| [Length of residence](https://faraday.ai/catalog/v2/length_of_residence) | The number of months the resident has lived at this location |
| [Retail spend share quintile](https://faraday.ai/catalog/v2/retail_spend_share_quintile) | Ratio of transactions within the retail channel compared to overall transactions, bucketed into 5 quintile groupings |
| [Online purchase count quintile](https://faraday.ai/catalog/v2/online_purchase_count_quintile) | Total lifetime number of online purchases made, bucketed into 5 quintile groupings |
| [Home and fashion trend spending quintile](https://faraday.ai/catalog/v2/home_and_fashion_trend_spending_quintile) | Average dollars spent on the latest trends in home decor and fashion, bucketed into 5 quintile groupings |
| [Household education](https://faraday.ai/catalog/v2/household_education) | Highest education level attained by any adult member of the household |
| [Child presence in household indicator](https://faraday.ai/catalog/v2/child_presence_in_household_indicator) | Indicates whether a child is present in the household |
| [Home and fashion trend purchase recency](https://faraday.ai/catalog/v2/home_and_fashion_trend_purchase_recency) | Recency of purchase in the trendsetters category |
| [Total amount spent on purchases quintile](https://faraday.ai/catalog/v2/total_amount_spent_on_purchases_quintile) | Total dollars spent on purchases within lifetime activity, bucketed into 5 quintile groupings |
| [Sports merchandise activewear spending quintile](https://faraday.ai/catalog/v2/sports_merchandise_activewear_spending_quintile) | Average dollars spent on the sports merchandise & activewear category, bucketed into 5 quintile groupings |
| [Beauty and spa spending quintile](https://faraday.ai/catalog/v2/beauty_and_spa_spending_quintile) | Average dollars spent on the beauty and spa category, bucketed into 5 quintile groupings |
| [Low to median household income pct](https://faraday.ai/catalog/v2/low_to_median_household_income_pct) | Estimates the share of households in the surrounding postcode earning at or below 80% of the postcode median household income |
| [High price female apparel accessories spending quintile](https://faraday.ai/catalog/v2/high_price_female_apparel_accessories_spending_quintile) | Average dollars spent on the high ticket female apparel/accessories category, bucketed into 5 quintile groupings |

These datapoints give you immediate visibility into whether a prospect is likely to be a fit for your brand and how best to reach them, before you spend on acquisition or campaigns.

## Why this matters

When you add these essentials to your records, you:

- **Save money** by excluding low-fit audiences before they eat up ad budget.
- **Increase conversions** by personalizing campaigns to the right households.
- **Boost efficiency** by matching offers to the channels households actually prefer.
- **Prove ROI** to the CFO with data-driven evidence that marketing spend is driving sales and lifetime value.

## Closing: smarter data, real results

Industry essential data packages are how Faraday makes your records AI-ready, without heavy lifting from your team. Combined with verified identity completion and custom predictive datapoints, they give you the full modern data stack your retail or e-commerce brand needs to compete.

Ready to see how these datapoints can transform your funnel? [Get started on Faraday Pro](/signup) for self-serve access, or [talk to a Context Consultant](/contact) if you're evaluating for a larger contract or want to talk through activation angles.

Faraday’s essential retail/e-commerce data package gives brands the curated datapoints—like household income, online vs. offline purchase ratios, and channel preferences—to qualify customers, cut waste, and boost conversions.

Just the essentials: get the datapoints retail & e-commerce brands need to make better decisions

A holdout evaluation is the industry-standard method data scientists use to validate predictive models: train on historical data, set aside a portion the model has never seen, then check whether the highest-scoring leads actually converted at higher rates. If they did, the model works.

If you're evaluating predictive data vendors, this is the test worth asking for. With tight budgets and increasing pressure to show ROI, you can't afford to bet on unproven tools or spend months figuring out if a vendor will live up to the promise. You need a way to separate legitimate predictive tools from the noise.

That's what our holdout evaluation does. It's not a methodology we invented — it's an industry standard used to validate [custom predictive data](https://faraday.ai/blog/predictive-lead-scoring-b2c), including [propensity-to-buy scores](https://faraday.ai/blog/propensity-to-buy-machine-learning) used across lead buying, direct mail, and [personalization](/blog/personalization-tactics-that-drive-conversions).

Below we've outlined how it works and what to look for when evaluating whether it's delivering real value.

## What is a holdout evaluation?

A holdout evaluation is a method for validating predictive models by comparing their predictions to real-world outcomes. We **set aside a portion of recent conversion data**—data the model has never seen before—and use it as an independent benchmark, against which predictions can be compared.

This "holdout set" acts as a real-world proving ground. It is untouched until after the predictive model has been trained.

Meanwhile, the rest of your historical data is combined with rich consumer data from the [Faraday Identity Graph (FIG)](/blog/whats-in-faraday-identity-graph) ([which includes over 1,400 data points on 240M+ U.S. adults and their households](https://faraday.ai/features/faraday-identity-graph)). We use this combined data to train the new predictive model.

Training a predictive model simply means finding patterns in historical conversions. In other words, what can we learn from your past customers to predict future outcomes? Once trained, the model generates predictions—scoring leads based on their likelihood to convert and only after these predictions are made do we compare them to the **real conversion rates** in the holdout set.

If the highest-scoring leads actually convert at higher rates than lower-scoring ones, the model is validated as accurate and reliable.

Holdout evaluations ensure predictive models aren't just theoretical—they work in the real world, using real data, to deliver real results.

## How holdout validation works at Faraday

We've streamlined this validation process to be transparent and straightforward:

1. ![customer conversion data](/images/blog/holdout_testing_diag/holdout_testing_diag-01.png)

   #### Step 1: Set aside a holdout group

   We start with your recent [lead conversion data](https://faraday.ai/blog/first-party-vs-third-party-data-marketing). Before any processing occurs, we reserve a portion of your most recent leads as a "holdout set"—data that the model will never see during training. This ensures we have an independent benchmark for evaluating predictions.

2. ![merged with FIG](/images/blog/holdout_testing_diag/holdout_testing_diag-02.png)

   #### Step 2: Enrich with consumer data

   With the holdout set reserved, we then combine remaining data with Faraday's [built-in consumer data](https://faraday.ai/features/faraday-identity-graph) sourced from the Faraday Identity Graph (FIG). This expanded dataset provides a deeper view into key predictive signals, like property and demographic information, life events, buying habits, and more.

3. ![model is trained and predictions begin](/images/blog/holdout_testing_diag/holdout_testing_diag-03.png)

   #### Step 3: Train the model and score leads

   Using this enriched dataset (excluding the holdout set), we train a predictive model to identify patterns in conversion behavior—learning from consumer signals, demographic data, and behavioral patterns.

   Once the model is trained, it scores each lead based on its likelihood to convert, generating a ranked list of opportunities—from high-value leads to those less likely to convert.

4. ![model is validated](/images/blog/holdout_testing_diag/holdout_testing_diag-04.png)

   #### Step 4: Compare predictions to real outcomes

   Finally, we put the model to the test: we compare its predictions against the real outcomes in that untouched holdout set. If the highest-scoring leads don't convert at significantly higher rates, we go back to the drawing board.

## How to know if a holdout validation was successful

A holdout validation succeeds when predicted scores align with actual conversion rates across segments. Look for these three signs:

1. **Clear correlation between predicted probability and actual conversion rates**  
   The model's predicted likelihood to convert should directly align with the actual conversion rates. If the predictions are accurate, high-probability leads should convert at a noticeably higher rate than lower-probability ones.

2. **High-scoring leads convert at significantly higher rates**  
   The top-scoring leads should consistently perform better than those ranked lower. This demonstrates that the model is identifying truly valuable leads.

3. **Consistent performance across different customer segments**  
   The model should show reliable results across various segments—whether you're targeting different demographics, regions, or other customer profiles. If the model works well across a broad set of data, it proves its reliability and scalability.

When a model meets these criteria, it's ready to generate custom predictive data points—whether that means optimizing [lead buying](https://faraday.ai/docs/how-to/use-cases/leads), refining your [direct mail strategy](https://faraday.ai/templates/lead-suppression), or enhancing your email [personalization](https://faraday.ai/templates/personalization) approach.

## Questions you should ask us

We want you to push us on what matters most for your business. Here are key questions to evaluate if we're the right fit:

1. **How does this model's performance compare to the lift you are seeing with similar businesses in our industry?**

   While we can't share client specifics, we can provide insight into broader trends and expected performance benchmarks.

   → **Why this matters:** Understanding industry benchmarks helps you gauge whether custom predictive data is truly a competitive advantage for your business.

2. **How much historical data do we need for reliable predictions?**

   The answer varies by business model and conversion volume. We'll help determine whether you have enough data—or how to supplement it if needed.

   → **Why this matters:** Predictive models work best when trained on strong data. Knowing the threshold for reliability ensures you're making data-driven decisions with confidence.

3. **What happens if the model doesn't perform as expected?**

   We have a structured iteration process to refine models when necessary, ensuring they reach meaningful predictive power.

   → **Why this matters:** No model is perfect on the first run. Understanding the process for continuous improvement ensures long-term success.

4. **How will this integrate with our existing marketing stack?**

   We'll provide technical specs, discuss implementation timelines, and outline the resources needed to get up and running.

   → **Why this matters:** Seamless integration means faster time-to-value and less friction in your workflow.

## Why holdout evaluations work

Holdout evaluations work because they test a model against data it has never seen — making it impossible to game the results. Any predictive data vendor can show you strong in-sample performance. Holdout testing is the check that confirms those patterns hold in the real world.

At the end of the day, three things matter to us:

1. **Confidence before commitment:**  
   You see proof before you commit

2. **Higher marketing efficiency:**
   Your marketing dollars go to the leads most likely to convert

3. **Stronger sales performance:**
   Your sales team focuses on opportunities worth their time

If the validation doesn't show meaningful lift, we'll tell you. If it does, we'll quantify exactly what that means for your business.

Interested in generating value you can trust? [Talk to a Context Consultant](/contact).

## Frequently asked questions

**What's the difference between a holdout evaluation and an A/B test?**

A holdout evaluation validates that a predictive model works—it confirms that high-scoring leads actually convert at higher rates before you commit to anything. An A/B test measures the impact of a decision you've already made (like sending a campaign to scored vs. unscored leads). Faraday runs the holdout evaluation first so you have confidence in the model itself, not just the downstream campaign.

**Do I need a large dataset for holdout testing to be meaningful?**

Not necessarily. The threshold varies by industry and conversion volume. In many cases, a few months of lead and conversion history is enough to build a reliable model. We'll tell you upfront if your data isn't sufficient—and if it's borderline, we can discuss ways to supplement it with data from the [Faraday Identity Graph](https://faraday.ai/features/faraday-identity-graph).

**What happens to my data during the validation process?**

Your first-party data is used solely to train and validate the predictive model—it is never shared with other clients or used to enrich Faraday's broader data assets. The holdout set is kept completely separate from training and is only used as an independent benchmark after predictions have been made.

Faraday uses holdout testing to validate custom predictive data before you commit, ensuring it delivers real, measurable results.

All posts