Overview
Estimating the net worth of a property owner is fundamentally different from valuing the property itself. A $3 million home might be owned by a retired executive with $50 million in liquid assets, or by a younger professional who stretched to afford it with heavy leverage. The property value is an important signal, but it is far from the whole picture. Plotbook's wealth estimation model takes a multi-signal approach that goes well beyond property value alone. It combines assessed property values from county records, professional background and career trajectory from databases like RocketReach and Apollo, organizational context including company size, industry, and the owner's role, public records including address history and associated properties, and AI-powered synthesis that reasons about all of these signals together. The result is a net worth range estimate — not a single precise number (which would be false precision) but a bracketed range that helps wealth managers categorize prospects into actionable tiers: high-net-worth ($1M-$10M), very-high-net-worth ($10M-$30M), and ultra-high-net-worth ($30M+). This tiering is what matters for prospecting workflows — you need to know whether someone is worth pursuing, not their exact bank balance.
Data Inputs
The wealth estimation model draws on every data source available through Plotbook's AI research agent. The foundation is the property record from Regrid, which provides the assessed market value, property type, and lot size. This alone establishes a wealth floor — someone who owns a $5M property has at least $5M in real estate assets, and typically significantly more in total net worth. On top of the property data, the model incorporates the complete output of the AI owner research agent: professional title and current employer (a C-suite executive at a Fortune 500 company has a very different wealth profile than a mid-level manager), complete employment history (a long career in finance or technology suggests accumulated equity and compensation), education background (graduate degrees from top programs correlate with higher lifetime earnings), company details including industry, size, and estimated revenue, multiple property ownership (if public records show additional properties, those values are additive), and family connections that may indicate inherited or shared wealth. Each of these signals contributes evidence to the wealth estimate. The model does not rely on any single data point — it builds a composite picture from all available information.
Property Value as a Wealth Signal
Property ownership is the single most visible indicator of wealth in the United States. The county-assessed value of a person's home establishes a reliable minimum asset threshold. Research consistently shows that for most Americans, real estate represents 25-35% of total net worth. This means a rough heuristic — multiplying property value by 3 to 4 — produces a reasonable ballpark for total net worth in many cases. Plotbook uses property value as the foundation of its wealth estimate while recognizing the wide variance around this heuristic. A homeowner in a high cost-of-living market like San Francisco may have a $3M home but a relatively modest total portfolio if they bought recently with heavy financing. Conversely, a homeowner in a mid-tier market who owns a $1.5M home free and clear may have substantial additional assets in investment accounts. The model accounts for these variations by incorporating other signals. For example, if the owner's employment history shows 20+ years at executive levels in technology companies, the model adjusts the estimate upward to account for likely stock compensation and accumulated savings. If the owner is in their 30s with a recent purchase in an expensive market, the model is more conservative about assuming liquid wealth beyond the property. Multiple property ownership is a particularly strong signal. If public records show that the same owner holds properties across multiple states or owns both a primary residence and vacation properties, the model sums the assessed values and adjusts the total net worth estimate accordingly. Multi-property owners are almost always wealthier than their single-property assessed value would suggest.
Professional and Career Signals
Professional background is the second most important input to the wealth estimation model. The AI agent retrieves employment history from RocketReach and Apollo, and the model evaluates several dimensions of that career data. Current title and level carry significant weight. C-suite executives (CEO, CFO, CTO), managing directors, and partners at professional firms typically have compensation packages including equity, bonuses, and carried interest that far exceed their base salary. The model knows that a Managing Director at Goldman Sachs has a materially different wealth profile than a Marketing Manager at a mid-size company, even if they live in similarly valued homes. Career duration and trajectory matter because wealth is accumulated over time. A professional with 25 years of progressive leadership roles has had decades of above-average earnings, savings, and investment growth. The model factors in the length and upward trajectory of the career — not just the current position — because cumulative earnings are the primary driver of non-real-estate wealth for most professionals. Industry sector provides additional context. Technology, finance, healthcare, legal, and real estate professionals tend to accumulate wealth faster than average due to equity compensation (tech), bonus structures (finance), or high billing rates (legal and consulting). The model adjusts its baseline estimates based on which industry the owner has spent their career in. Entrepreneurship and business ownership are strong wealth indicators. If the AI agent identifies the owner as a founder, owner, or principal of a business — particularly one with significant revenue or employee count — the model substantially increases its estimate to account for business equity, which is often the single largest asset for entrepreneurs.
Organizational Signals
The company where someone works provides useful context even beyond their individual title. Plotbook's model considers the organization's estimated revenue and employee count, which indicate the scale of the business and the likely sophistication of its compensation structure. Large companies with substantial revenue are more likely to offer equity compensation, deferred compensation plans, and performance bonuses that accelerate wealth accumulation. The organization's industry classification helps refine estimates — a VP at a fintech startup has a different expected wealth trajectory than a VP at a nonprofit. Company stage also matters: executives at pre-IPO technology companies may hold illiquid equity worth millions or nothing, creating wider estimate ranges. Publicly traded companies are easier to model because executive compensation is disclosed in SEC proxy statements. While Plotbook does not directly query SEC filings, the AI agent's deep research capabilities may surface compensation data from news articles, proxy summaries, or industry reports that mention the owner. The model treats organizational signals as multipliers on the professional background baseline. A senior executive at a Fortune 500 company gets a higher multiplier than the same title at a 50-person startup, because the compensation structures at large, established companies are more predictable and typically more generous at senior levels.
Public Records Intelligence
Public records provide several additional signals that inform the wealth estimate. Address history shows where the owner has lived over time, and a pattern of progressively more expensive neighborhoods suggests growing wealth. If someone moved from a $500K home to a $1.5M home to a $4M home over 15 years, that trajectory reveals both wealth accumulation and lifestyle inflation that helps bracket their current net worth. Associated persons from public records can indicate family wealth. A spouse who is a named partner at a law firm or a parent who owns multiple commercial properties adds context about the household's total financial picture. While the model focuses on the individual property owner, family wealth signals are noted in the profile. Property transaction history — when visible through county records — reveals buying and selling patterns. An owner who purchased their current home ten years ago at a lower price may have significant equity appreciation, while a recent buyer may have less equity but demonstrated the financial capacity to make a large purchase. Previous property sales suggest liquidity events where the proceeds may have been invested. The number and type of properties associated with an owner in public records is a direct wealth indicator. An individual who appears as the owner of record on three residential properties and a commercial building has demonstrable real estate assets that can be summed for a more accurate floor estimate.
AI Synthesis and Reasoning
The final step in the wealth estimation process is AI synthesis, where Google Gemini evaluates all gathered signals together and produces a reasoned estimate. This is not a simple formula or lookup table — the AI model considers the full context of the gathered information and applies reasoning that accounts for the interactions between signals. For example, the model might reason: this person owns a $4M home in Greenwich, CT, has been a Managing Director at a major investment bank for 12 years, holds an MBA from Wharton, and public records show a second property in Nantucket valued at $2.5M. The combined real estate assets are $6.5M, the career trajectory suggests 12+ years of seven-figure total compensation, and the lifestyle indicators (two premium properties, senior finance career, elite education) are consistent with a net worth in the $15M-$25M range. The AI synthesis step is particularly valuable for unusual or complex cases where simple heuristics break down. An early retiree in their 50s living in a modest home might have tens of millions from a company exit. A young professional in a lavish home might be heavily leveraged. A nonprofit executive in a moderate home might come from generational wealth. The AI model considers these nuances in ways that a rigid formula cannot. The synthesis output is a net worth range (e.g., $15M-$25M) rather than a point estimate, reflecting the inherent uncertainty in wealth estimation. The range width varies based on how much information was available — more data sources confirming consistent signals produce tighter ranges, while limited or conflicting signals produce wider ranges.
Output Format and Confidence
The wealth estimate is presented as a range rather than a single number, reflecting honest uncertainty about the true value. Typical output might read '$5M-$10M' or '$15M-$25M', giving wealth managers enough precision to categorize prospects into relevant tiers without implying false exactness. The estimate is stored alongside the full prospect profile in Plotbook's saved profiles system, where it is displayed as part of the owner intelligence dashboard. It is tagged with a status indicator — pending (if the AI research has not yet completed), completed (estimate available), or failed (insufficient data to estimate) — so you know at a glance which profiles have actionable wealth data. The wealth estimate is presented in context with the supporting evidence: the property value, the professional background, the organizational affiliation, and the key data sources that informed the estimate. This transparency lets you quickly assess whether the estimate aligns with your own judgment and industry knowledge. A wealth advisor who specializes in retired executives, for example, might weight the career history more heavily than the algorithm does, and the presented evidence lets them make that judgment. The profile's overall confidence score (0-100) also applies to the wealth estimate. A profile with 90+ confidence that draws on multiple confirming sources will have a more reliable wealth estimate than a profile built from limited public records alone. The confidence score helps you prioritize which prospects to pursue first and which might need additional manual verification.
Limitations and Honest Caveats
Wealth estimation from public data sources has inherent limitations that users should understand. First, the model cannot see private financial accounts. Investment portfolios, bank balances, trust assets, retirement accounts, and other non-public financial holdings are invisible to any external estimation tool. The estimate is based on observable signals — property ownership, career trajectory, and organizational context — and extrapolates from these to total net worth. Second, assessed property values may differ from market values by 10-20% depending on the jurisdiction and when the last assessment occurred. The model uses assessed values because they are consistently available across all counties, but users should understand that the property value input is approximate. Third, the model works best for professionals with established careers and digital footprints in professional databases. Individuals who are retired, self-employed without a web presence, or who manage wealth through trusts and LLCs may have fewer observable signals, resulting in wider estimate ranges or lower confidence scores. Fourth, inherited wealth is difficult to detect from public signals alone. A person living in a modest home with a simple career profile might have inherited tens of millions, and the model would likely underestimate their net worth. Conversely, someone with a flashy lifestyle and expensive home might be heavily leveraged, and the model might overestimate their liquid wealth. Despite these limitations, Plotbook's multi-signal approach produces estimates that are significantly more accurate than property-value-only methods and dramatically faster than manual research. For the purpose of prospect prioritization and outreach planning — which is the primary use case — the estimates are more than sufficient to identify and tier high-net-worth individuals effectively.