Figure AI Market Strategy report — Home Equity Loans

Answer Capsule

Figure shows meaningful AI recommendation-stage visibility in home equity lending, but it is not the category’s default winner. In the uploaded Figure dataset, the brand appears in 46 of 297 observations and earns 45 valid recommendations, with most of that strength concentrated in the home-equity discovery cluster rather than comparison or pricing prompts. The clearest win is Figure’s relevance in HELOC and home-equity-lender prompts. The clearest weakness is weak rank-one ownership and near-zero visibility outside that narrow prompt pocket. The biggest opportunity is to turn specialist relevance into first-choice recommendation power.

Want this analysis for your company? CiteWorks Studio produces AI Market Strategy reports showing where your brand appears, disappears, or gets recommended across ChatGPT, Gemini, Copilot, Perplexity, Google AI Mode, and Google AI Overviews. Request an AI Visibility Audit

Who This Report Is For

This report is for CMOs, founders, growth teams, investor relations teams, agency partners, and category leaders in lending, fintech, and home-finance markets who want to understand whether AI systems treat their brand as a real borrower option or just a supporting mention.

Report Card

• Report type: AI Market Strategy report
• Target company: Figure
• Category / market studied: Home equity loans, HELOCs, home-equity lender discovery, and mortgage-adjacent lender comparison
• Reporting month: May 2026
• AI platforms tracked: 6
• Public high-intent clusters: 3
• AI observations analyzed: 297
• Competitors tracked: Achieve, Bank of America, Bethpage Federal Credit Union, Connexus Credit Union, Discover Home Loans, LendingTree, PNC Bank, Rocket Mortgage, Spring EQ, and TD Bank

Executive Summary

Figure is present in AI answers, and it is recommended often enough to matter. But it is not yet owning the category. Across the uploaded 297-observation structured dataset, Figure appears in 46 observations and receives 45 valid recommendations. That is a meaningful recommendation signal, but it is much smaller than the broad category footprint of Bank of America and Rocket Mortgage.

The strongest signal is concentration. Figure’s performance is heavily tied to the core home-equity discovery prompt set. In that main cluster, Figure appears in 46 of 242 observations and earns 45 valid recommendations, with 26 Top 3 placements and 3 rank-one placements. Outside that cluster, the public packet shows essentially no measurable presence.

That pattern matters. It suggests AI systems understand Figure as a relevant specialist in home equity and HELOC lending, but not as a broad mortgage default. Presence is not preference, and in Figure’s case the issue is not total invisibility. The issue is limited recommendation territory.

Figure’s strongest platform signals come from Google AI Overviews, Google AI Mode, and Gemini. Google AI Overviews shows the highest raw presence and recommendation coverage for Figure in the packet, while Gemini produces the best average rank and two of Figure’s three rank-one wins. Copilot also contributes a smaller but credible recommendation signal.

The clearest platform gaps are ChatGPT and Perplexity. In the uploaded packet, Figure records no presence and no valid recommendation coverage on either platform. That is a meaningful public gap because those are buyer-facing research environments where shortlist formation can happen early.

The broader category context is also important. The pasted benchmark and companion industry article describe Rocket Mortgage as the broad mortgage-adjacent leader, while the Figure-specific dataset shows Bank of America leading the uploaded packet on raw presence and recommendation coverage. Figure, by contrast, reads as a specialist challenger: clearly retrievable in home-equity prompts, but not yet the first answer often enough.

QA note: the uploaded benchmark article and the Figure dataset do not map cleanly to the same exact scope, and parts of the metrics packet carry stale inherited cluster labels. For this report, the Figure structured dataset is treated as the source of truth for company-level findings.

What Figure Is Winning

Figure is winning specialist relevance. In the uploaded dataset, AI systems repeatedly retrieve Figure for HELOC and home-equity-lender prompts, which is the most commercially important prompt pocket for this brand.

Figure also converts presence into recommendation unusually well when it is retrieved. Across the full packet, 45 of its 46 mentions are marked as valid recommendations. That means the issue is not weak recommendation conversion after mention. The issue is getting retrieved in more borrower-choice situations.

Figure’s best platform quality appears on Gemini and Copilot. Gemini gives Figure four valid recommendations in 24 observations, including two rank-one placements and a 1.33 average recommended rank. Copilot adds two valid recommendations and one rank-one placement.

The strongest positive interpretation is this: AI systems do appear to understand what Figure is for. The brand is not being treated primarily as an editorial source or generic mention. It is being treated as a legitimate lender option inside a narrow but valuable prompt set.

Where Figure Has the Clearest AI Visibility Gaps

The biggest gap is breadth. Figure performs inside the core home-equity discovery cluster, but it does not show public traction in the comparison and pricing clusters in this packet. That creates visibility without full shortlist control across the borrower journey.

The second gap is rank-one ownership. Figure records only 3 rank-one placements across 297 observations. It is being included, but it is not consistently being chosen first.

The third gap is platform coverage. Figure has no measured presence in ChatGPT or Perplexity in the uploaded packet. Even if that changes over time, this snapshot shows a clear public weakness in two influential answer environments.

The competitive gap is also visible. In the full packet, Bank of America records 138 mentions and 93 valid recommendations. Rocket Mortgage records 116 mentions and 67 valid recommendations. PNC Bank, while not central to the broader benchmark narrative, also shows a stronger rank-one pattern than Figure in the uploaded data. Figure is competitive, but it is not setting the category terms.

Biggest Opportunity

Figure’s biggest opportunity is to move from specialist inclusion to default recommendation.

The packet suggests AI systems already recognize Figure as a credible answer for home-equity lending. That means the next move is not generic awareness content. The next move is building stronger first-choice eligibility around the exact questions borrowers ask when they want the best lender, the best rates, the safest option, the fastest process, or the clearest fit.

In practice, that means improving the owned answer layer for comparison, trust, fees, rates, and qualification questions. Figure appears to have enough category relevance to get into the answer. What it needs is stronger support for why it should be ranked first.

Prompt Evidence

**Discovery / HELOC ** Prompt: **Is figure the best HELOC? ** Readout: Figure appears directly in a brand-led validation prompt, which is a strong signal that AI systems connect the brand to the product category.

**Discovery / Best lender ** Prompt: **best home equity lenders ** Readout: Figure is retrieved in core shortlist-style discovery language, which supports the view that its main public strength is specialist retrievability.

**Discovery / Best line of credit ** Prompt: **best home equity line of credit lenders ** Readout: Figure appears in lender-selection language tied to HELOC-specific intent, reinforcing its specialist rather than broad-mortgage positioning.

What CiteWorks Studio Would Do Next

**1. Expand Figure’s recommendation territory. ** Map the exact discovery, comparison, trust, fees, and qualification prompts where Figure is absent, present, or outranked. The key issue is not whether Figure can appear. It is where the brand still fails to enter the borrower shortlist.

**2. Build first-choice pages, not just relevant pages. ** Figure likely needs stronger comparison pages, trust pages, qualification explainer pages, rate-and-fee clarity pages, and borrower-fit pages. The goal is to make AI systems more confident explaining why Figure should be chosen first, not just included.

**3. Strengthen the citation layer around product clarity and trust. ** AI recommendation behavior is rarely driven by the company site alone. Figure needs stronger third-party reinforcement, clearer source support, and tighter entity consistency around HELOCs, home equity loans, rates, fees, speed, and borrower fit.

Why This Matters

Figure already has AI relevance in the market that matters most to it. That is a real advantage.

But relevance alone is not enough. The borrower does not win the category by asking whether Figure exists. The borrower wins by asking who is best, who is fastest, who is cheapest, who is safest, and who should make the shortlist.

In this packet, Figure is clearly in that conversation. It is not yet dominating it. That is the commercial significance of the report.

Core Metrics

• Mentions: 46
• Valid recommendations: 45
• Top 3 recommendation count: 26
• Rank #1 recommendation count: 3
• Average recommended rank: 2.35 in the packet metrics layer
• Raw mention presence rate: 15.49%
• Valid recommendation coverage: 15.15%
• Top 3 recommendation rate: 8.75%
• Rank #1 recommendation rate: 1.01%

Sentiment Score

Sentiment score matters because raw visibility alone can be misleading. A lender can be present in AI answers without being framed as a serious option. In this packet, Figure’s mentions are entirely positive rather than neutral or negative, which means its issue is not bad framing. Its issue is limited scale.

For this report series, sentiment score is calculated as:

(positive mentions × 1 + neutral mentions × 0 + negative mentions × -1) / total mentions

Under that method, Figure’s packet-level sentiment score is 1.00. That is unusually strong. It means the public AI framing captured here is favorable when Figure appears. Again, the challenge is breadth and rank ownership, not negative treatment.

Sentiment by Platform

Methodology Note

This is a company-specific public report built from the uploaded Figure structured dataset plus the pasted home-equity benchmark text. The dataset is treated as the source of truth for company metrics. The broader benchmark and industry article are used only for category framing. QA note: parts of the uploaded packet carry stale inherited labels, so this report normalizes the interpretation around the actual prompt content rather than the mislabeled cluster names.

Methodology

• This is a one-company public report focused on Figure. Other named brands are treated as competitors relative to Figure.
• The reporting window is May 2026.
• The structured packet contains 297 AI observations across six platforms: ChatGPT, Gemini, Perplexity, Copilot, Google AI Mode, and Google AI Overviews.
• Company-level interpretation is based primarily on the structured Figure dataset, not on the broader Rocket benchmark text.
• A mention counts when Figure appears in an AI answer.
• A valid recommendation counts when the dataset marks Figure as recommendation-level rather than simple presence.
• Rank metrics reflect recommendation placement, with Top 3 and rank-one treated separately.
• Sentiment is evaluated separately from presence so that visibility is not confused with preference.
• Because parts of the packet contain stale inherited labels, actual prompt text is used to interpret the market as home equity / HELOC lending rather than the mislabeled cluster names.
• This is a directional public benchmark, not a definitive market census. AI outputs can vary by platform, prompt wording, geography, lender availability, and time.