How AI Search Is Recommending Mortgage Rates

Executive Summary

Mortgage discovery is shifting from search-driven comparison to AI-assisted shortlist creation.

Instead of beginning with lender websites, rate tables, or review portals, many borrowers now ask AI platforms which lender offers the best mortgage rates, which company they should trust, and how competing lenders compare. The result is a new recommendation layer that sits between consumer intent and lender selection.

The May 2026 Mortgage Rates benchmark found a highly concentrated recommendation environment. Rocket Mortgage emerged as the dominant recommendation-stage leader, while Better Mortgage established itself as a credible challenger, particularly in lender-selection prompts. The benchmark also revealed a growing gap between visibility and recommendation strength. Brands such as NerdWallet and LendingTree appear frequently within AI-generated answers, but that visibility does not consistently translate into lender recommendation credit.

For mortgage lenders, the competitive question is no longer whether AI mentions the brand. It is whether AI advances the brand into the buyer's shortlist.

Methodology

1. Market Studied

U.S. mortgage rates and mortgage lender discovery market.

2. Brands/Entities Included

The benchmark tracked Rocket Mortgage, Better Mortgage, New American Funding, loanDepot, NerdWallet, LendingTree, Own Up, Rate, AmeriSave Mortgage, Guaranteed Rate, Freedom Mortgage, Veterans United Home Loans, Credible, Figure, Mr. Cooper, and additional lender entities appearing within AI-generated responses.

3. Data Collection Date/Window

May 2026 benchmark dataset.

4. AI Platforms Tested

Six AI search and answer-generation surfaces were included in the benchmark. Public reporting references six AI platforms, while uploaded observation data confirms ChatGPT participation. Additional platform-specific details are not fully available in the public dataset.

5. Number of Prompts Tested

887 AI observations across tracked prompt sets.

6. Prompt Categories

Three primary high-intent buyer clusters:

• Best Mortgage Lenders
• Mortgage Lender Comparisons
• Mortgage Pricing and Costs

These clusters span consideration, evaluation, comparison, and decision-stage buyer behavior.

7. Definition of a Mention

A mention occurs when a company appears within an AI-generated response regardless of recommendation status.

8. Definition of a Valid Recommendation

A valid recommendation occurs when the AI system positively advances a lender as a recommended option or shortlist candidate. Visibility alone does not receive recommendation credit.

9. Ranking and Scoring Metrics Used

The benchmark framework evaluates:

• Recommendation coverage
• Top-three recommendation rate
• Rank-one recommendation rate
• Citation share
• Net sentiment/framing
• Modeled monthly captured recommendation value

These metrics are analyzed separately because visibility, recommendation strength, and ranking performance are not interchangeable.

10. Limitations

This benchmark represents a point-in-time snapshot of AI-generated recommendations. AI outputs change continuously. Modeled recommendation values are directional estimates and should not be interpreted as revenue or attributable business outcomes.

Key Findings

Rocket Mortgage Controls the Recommendation Layer

Across the benchmark, Rocket Mortgage consistently emerged as the strongest recommendation-stage performer. The brand demonstrated high recommendation coverage, strong shortlist inclusion, and leading top-rank capture. Public benchmark reporting identifies Rocket Mortgage as the category leader.

Better Mortgage Has Established Challenger Status

Better Mortgage appears to have secured meaningful recommendation-stage visibility, particularly in Best Mortgage Lenders prompts. However, its overall recommendation footprint remains materially smaller than Rocket Mortgage's, positioning Better as a challenger rather than the category leader.

Visibility Does Not Equal Recommendation Power

NerdWallet and LendingTree remain highly visible entities in mortgage-related AI responses. Yet the benchmark indicates that visibility frequently serves an informational or comparison role rather than translating into lender recommendation credit.

Recommendation Leadership Is Concentrating

The benchmark suggests that recommendation power is becoming concentrated among brands that combine lender authority, comparison visibility, recognizable brand equity, and strong third-party validation signals.

What Changed in the Market

Historically, mortgage discovery depended heavily on search rankings, review sites, and direct rate shopping.

AI discovery changes the process.

Instead of visiting multiple websites, borrowers increasingly ask a single AI platform questions such as:

• Who has the best mortgage rates?
• Which lender should I choose?
• Is Rocket Mortgage better than Better Mortgage?
• What lender is best for first-time buyers?

These prompts compress research, comparison, and recommendation into a single interaction.

As a result, recommendation-stage visibility now influences buyer consideration earlier than traditional search alone.

What the Benchmark Found

The benchmark reveals a clear segmentation within the mortgage-rate category.

Recommendation Leaders

Rocket Mortgage sits alone as the strongest overall recommendation leader.

Strong Challengers

Better Mortgage demonstrates meaningful recommendation strength, particularly during lender-selection moments.

Recommendation-Capable Alternatives

New American Funding and loanDepot appear capable of earning recommendation consideration but trail category leaders.

Visibility-Led Information Brands

NerdWallet and LendingTree frequently appear within the answer environment but convert less effectively into lender recommendation status.

Limited Recommendation Presence

Own Up demonstrates comparatively weaker recommendation-stage visibility across benchmark observations.

Why Visibility Is Not Enough

One of the most important findings from this benchmark is the distinction between being present and being recommended.

A lender may:

• Appear frequently in AI responses
• Be cited as a source
• Influence comparison discussions
• Contribute information used in AI answers

Yet still fail to become part of the final recommendation shortlist.

This distinction matters because buyers rarely stop at informational prompts. They move through a sequence of lender-selection, comparison, pricing, and trust-validation questions.

Brands that lose recommendation credit at any stage risk losing momentum throughout the buying journey.

The Citation Layer

The uploaded benchmark observations reveal a recurring pattern of citation sources appearing alongside mortgage recommendations.

Frequently surfaced source types include:

• Financial publishers
• Editorial review sites
• Mortgage comparison resources
• Consumer review environments
• Community discussions and forums

Examples observed within the dataset include Forbes Advisor, Finder, ConsumerAffairs, and Reddit discussion threads. These sources frequently appear in recommendation-supporting contexts.

This does not prove direct causation. However, the benchmark suggests that AI systems often synthesize recommendations from a combination of lender authority, editorial validation, review ecosystems, and public discussion sources.

The implication is clear: recommendation-stage visibility is influenced not only by what a lender publishes, but by the broader public evidence layer surrounding the brand.

What Brands Need to Fix

Mortgage lenders seeking stronger AI recommendation performance should focus on four areas:

Strengthen Third-Party Validation

Editorial rankings, review coverage, comparison inclusion, and industry recognition appear increasingly important.

Expand Comparison Presence

Many high-intent prompts involve direct lender comparisons. Brands absent from these conversations risk losing shortlist consideration.

Improve Source Consistency

AI systems synthesize across multiple sources. Conflicting or incomplete public information can weaken recommendation confidence.

Build Recommendation-Stage Content

Most lenders still optimize for search traffic. Increasingly, they must also optimize for recommendation-stage discovery and comparison prompts.

How CiteWorks Studio Helps

1. Map AI Recommendation Visibility

Track prompts, platforms, company presence, valid recommendations, top-three performance, rank-one performance, framing quality, and citation sources.

2. Identify the Sources Shaping AI Answers

Analyze the editorial, review, forum, government, directory, owned, and search-visible sources influencing lender recommendations.

3. Build the Citation Architecture Plan

Strengthen the public evidence layer so AI systems have more accurate, consistent, and persuasive source material to synthesize.

Commercial Takeaway

Mortgage lending is becoming a recommendation-driven market.

The brands winning AI-generated shortlists are not always the brands with the largest advertising budgets or the most visible rate tables. They are increasingly the brands supported by stronger recommendation signals, stronger citation footprints, and stronger public evidence layers.

For lenders, the opportunity is no longer simply to be visible.

The opportunity is to become recommendable.

See How AI Is Recommending Your Brand

AI systems are already influencing lender discovery, comparison, and shortlist formation.

An AI Visibility Audit from CiteWorks Studio can help identify:

• Where your brand appears
• Where competitors receive recommendation credit instead
• Which prompt clusters carry the highest commercial risk
• Which sources influence AI-generated answers
• What needs to change to improve recommendation-stage visibility

Understanding those gaps is the first step toward improving how AI systems compare, rank, and recommend your brand.