How AI Search Is Recommending Mortgage Rates

Mortgage rates are becoming an AI-mediated shortlist market. Borrowers are no longer only searching rate tables or lender websites. They are asking AI systems which lender has the best mortgage rate, which mortgage company is best, which bank to use for a home loan, which lender is best for first-time buyers, and how to compare online mortgage lenders.

The 2026 LLM Authority Index benchmark shows that AI recommendation power in mortgage rates is highly concentrated. Rocket Mortgage is the clear recommendation leader in the supplied dataset, while Better Mortgage is a meaningful challenger in best-lender and rate-related moments. New American Funding, loanDepot, NerdWallet, and LendingTree appear across the category, but their visibility does not always convert into lender-style recommendation power.

Methodology

1. Market studied: Mortgage rates and mortgage lender discovery, including best mortgage lenders, best mortgage rates, home loan providers, online mortgage lenders, mortgage lender comparisons, refinancing, lender selection, and mortgage pricing / cost prompts.
2. Brands/entities included: Rocket Mortgage, AmeriSave Mortgage, Better Mortgage, Credible, Figure, Freedom Mortgage, Guaranteed Rate, LendingTree, loanDepot, Mr. Cooper, NerdWallet, New American Funding, Own Up, Rate, and Veterans United Home Loans in the uploaded company universe. The structured metrics leaderboard reports measurable tracked metrics for Rocket Mortgage, Better Mortgage, loanDepot, NerdWallet, LendingTree, New American Funding, Rate, AmeriSave Mortgage, and Own Up.
3. Data collection date/window: May 2026. The structured Rocket Mortgage dataset was loaded on May 19, 2026 and reports the benchmark month as 2026-05.
4. AI platforms tested: Six AI search surfaces were tracked in the public benchmark. The dataset includes major AI discovery environments used for mortgage-rate and lender-selection prompts.
5. Number of prompts tested: The public benchmark reports 887 AI observations across three high-intent clusters.
6. Prompt categories: Best Mortgage Lenders, Mortgage Lender Comparisons, and Mortgage Pricing and Costs. In the structured metrics, these appear as three cluster containers: C01 with 422 observations, C02 with 225 observations, and C03 with 240 observations.
7. Definition of a mention: A brand counted as mentioned when it appeared in an AI answer as a lender, comparison source, marketplace, financial publisher, mortgage company, or relevant mortgage-rate entity.
8. Definition of a valid recommendation: A valid recommendation required positive, shortlist-quality lender or provider framing. Neutral mentions, comparison-source visibility, factual rate references, extraction-fallback records, or publisher citations were not treated as recommendation credit unless the dataset marked them as valid recommendations.
9. Ranking/scoring metrics used: Raw mention presence, valid recommendation coverage, recommended Top 3 rate, recommended Rank 1 rate, average recommended rank, positive / neutral / negative visibility, net sentiment score by mentions, citation/source patterns, and modeled monthly captured recommendation value. Modeled value is a benchmark estimate, not revenue, loan volume, funded mortgages, or borrower conversion.
10. Limitations: This is a point-in-time benchmark. Mortgage rates change frequently, and AI outputs vary by platform, prompt wording, retrieval state, source freshness, geography, borrower profile, loan type, and date. This report evaluates AI discovery behavior and recommendation patterns; it does not provide mortgage, rate, credit, or lender-selection advice.

Key findings

1. Rocket Mortgage is the strongest overall AI recommendation leader.
Across 887 observations, Rocket Mortgage appeared in 42.05% of responses, earned 29.54% valid recommendation coverage, reached a 26.16% Top 3 recommendation rate, and captured a 19.95% Rank 1 recommendation rate. Its modeled monthly captured recommendation value was 261,618.1924, the highest in the measured dataset.

2. Rocket’s strength is concentrated in best-lender discovery and mortgage pricing / cost prompts.
In the Best Mortgage Lenders cluster, Rocket Mortgage had 53.08% valid recommendation coverage, 46.68% Top 3 rate, and 36.02% Rank 1 rate. In Mortgage Pricing and Costs, it still led with 15.83% valid recommendation coverage, 14.58% Top 3 rate, and 10.42% Rank 1 rate.

3. Better Mortgage is the strongest challenger, but materially behind Rocket.
Better Mortgage showed 11.95% raw mention presence, 9.47% valid recommendation coverage, 6.20% Top 3 rate, 2.59% Rank 1 rate, and 53,725.6091 in modeled monthly captured recommendation value. That makes Better a real AI shortlist participant, but not the category-control brand.

4. NerdWallet and LendingTree are visible, but often as source or comparison layers.
NerdWallet appeared in 25.70% of observations and LendingTree in 12.63%, but their valid recommendation coverage was much lower at 2.25% and 1.47%, respectively. This reflects the public benchmark’s central distinction: being visible in the AI answer environment is not the same as being recommended as the lender.

5. The most visible warning sign is conversion from visibility to lender recommendation.
The public benchmark identifies this clearly: NerdWallet and LendingTree may shape the answer environment, but they do not consistently capture lender recommendation demand. Better Mortgage is present and positively framed, but Rocket Mortgage is more often advanced into the shortlist.

What changed in the market

Mortgage shopping has always been trust-sensitive and rate-sensitive. Borrowers compare APR, monthly payment, fees, points, closing costs, loan type, preapproval speed, service quality, refinance options, first-time buyer support, VA loan fit, and whether a lender is easy to work with.

AI search changes how that comparison begins. A borrower can now ask:

“Which lender has the best mortgage rate?”
“What is the best mortgage lender?”
“What is the best bank to get a home loan?”
“Which mortgage provider is best?”
“Who is best for refinancing?”
“Which online mortgage lender has the lowest fees?”

These are not generic information queries. They are shortlist-forming prompts.

Instead of sending the borrower to ten blue links, AI systems compress the market into a ranked set of lenders, marketplaces, banks, and comparison sources. The brand that gets framed as “best overall,” “best for low fees,” “best for first-time buyers,” “best for VA loans,” or “best for rate shopping” can shape the borrower’s next step before the borrower ever reaches a lender site.

What the benchmark found

The public benchmark identifies Rocket Mortgage as the category leader and Better Mortgage as the strongest challenger. The structured Rocket Mortgage dataset supports that pattern.

Rocket Mortgage owns the broad AI shortlist.
Rocket leads the measured dataset on recommendation coverage, Top 3 rate, Rank 1 rate, average recommended rank, and modeled captured recommendation value. In raw observations, Rocket is repeatedly framed as a strong or leading option for digital experience, first-time buyers, online mortgage experience, and broad lender selection.

Better Mortgage is strongest when the buyer asks about online lenders, low fees, and rate-related lender choice.
Better is a credible challenger in the Best Mortgage Lenders cluster and appears in rate-sensitive prompts, but its recommendation footprint is much smaller than Rocket’s overall.

New American Funding has meaningful positive visibility, but lower modeled value.
New American Funding had 17.02% raw mention presence and 10.03% valid recommendation coverage, slightly above Better on recommendation coverage. But its modeled captured recommendation value was only 5,718.4667, suggesting its recommendation capture occurred in lower-value or less commercially weighted prompt contexts than Better’s.

loanDepot is a visible lender option with moderate recommendation capture.
loanDepot appeared in 10.71% of observations, earned 6.31% valid recommendation coverage, and captured 23,906.4286 in modeled monthly recommendation value. It performs as a visible lender, but below Rocket and Better on the overall modeled value leaderboard.

NerdWallet and LendingTree are part of the recommendation environment, but not always the recommendation.
NerdWallet and LendingTree are important because AI systems use comparison and financial-publisher sources when assembling mortgage answers. But as brands in the tracked universe, they often function as source-layer or comparison-layer entities rather than lender recommendations.

Why visibility is not enough

Mortgage rates are one of the clearest examples of the difference between appearing in AI answers and winning AI recommendations.

A brand can appear as:

a cited finance publisher,
a rate-comparison source,
a marketplace,
a factual rate reference,
a lender example,
or a valid lender recommendation.

Only the last category carries full recommendation-stage value.

NerdWallet illustrates the difference. It appeared in 228 of 887 observations, but received only 20 valid recommendations and 10 Top 3 placements. LendingTree appeared in 112 observations, but received only 13 valid recommendations. By contrast, Rocket Mortgage appeared in 373 observations and received 262 valid recommendations, with 232 Top 3 placements.

The same distinction applies to Better Mortgage. Better is not invisible. It has meaningful positive visibility and recommendation capture. But Rocket is more frequently advanced, ranked, and value-weighted as the chosen lender.

For mortgage brands, the strategic question is no longer only:

Are we mentioned?

It is:

Are we recommended?
Are we in the Top 3?
Are we ranked first?
Are we framed as best for rates, first-time buyers, low fees, refinancing, VA loans, online speed, or comparison shopping?
Are we the lender recommendation, or only part of the source layer?

The citation layer

Mortgage-rate AI answers are shaped by a dense source ecosystem: financial publishers, mortgage comparison sites, lender pages, rate pages, consumer review sites, and community discussions.

The structured extraction shows recurring cited environments including Forbes, The Motley Fool, Finder, ConsumerAffairs, Reddit, Reuters, mortgage-rate publishers, and other comparison or lender-review sources.

That matters because mortgage recommendations are not built from lender-owned pages alone. AI systems synthesize from third-party rankings, rate explainers, lender reviews, daily or current-rate pages, comparison tables, borrower communities, and official lender content.

This creates different roles in the AI answer layer:

Rocket Mortgage benefits from broad lender familiarity, digital mortgage authority, first-time-buyer framing, and repeated positive shortlist placement.

Better Mortgage benefits from online-lender, low-fee, and rate-sensitive positioning.

New American Funding often appears where credit flexibility, less-than-perfect borrower profiles, or broader lender fit matter.

loanDepot can appear around fast closing, refinancing, and flexible requirements.

NerdWallet and LendingTree help shape the source environment, but they do not automatically capture lender recommendation value.

For lenders, the citation layer is now part of the competitive surface. It determines whether AI systems can confidently explain why a lender belongs in the shortlist.

What brands need to fix

Mortgage lenders and mortgage-rate platforms need to manage AI discovery as a recommendation system, not just a rate table or search channel.

The first fix is prompt-stage coverage. Brands need to know whether they win broad “best lender” prompts, rate-specific prompts, refinancing prompts, first-time buyer prompts, VA loan prompts, online lender prompts, and bank-versus-lender comparison prompts.

The second fix is recommendation-stage tracking. Mentions, valid recommendations, Top 3 placement, Rank 1 capture, and modeled value need to be separated.

The third fix is rate and fee framing. Mortgage-rate prompts are sensitive to current-rate evidence, but AI systems also need credible source support around fees, points, loan types, credit requirements, preapproval, and closing experience.

The fourth fix is source-role clarity. Financial publishers and marketplaces may be highly visible, but brands need to know whether they are appearing as the source, the comparison layer, or the recommended lender.

The fifth fix is citation architecture. Mortgage brands need accurate, current, and consistent evidence across official pages, third-party reviews, rate-comparison sources, editorial rankings, borrower communities, and lender-specific explainers.

How CiteWorks Studio helps

1. Map AI recommendation visibility. Track prompts, platforms, company presence, valid recommendations, Top 3 and Rank 1 performance, framing, and citation sources.
2. Identify the sources shaping AI answers. Find the editorial, review, forum, government, directory, owned, and search-visible sources that influence brand framing.
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 rates are becoming an AI-mediated trust and comparison market.

Rocket Mortgage currently controls the strongest recommendation-stage position in the supplied benchmark. Better Mortgage is a credible challenger, especially around online lending, low-fee positioning, and lender-selection prompts, but it remains behind Rocket on overall recommendation coverage, Top 3 presence, Rank 1 capture, and modeled monthly captured recommendation value.

The category’s next competitive advantage will come from owning the evidence layer around specific borrower questions: best mortgage lender, best mortgage rate, best online mortgage lender, best lender for refinancing, best lender for first-time buyers, best VA lender, lowest fees, and best lender comparison.

For mortgage brands, awareness is not enough. AI systems need enough trusted, current, and comparison-ready evidence to recommend the lender, rank it highly, and explain why it fits the borrower’s situation.

CTA

Want to know how AI systems are recommending your mortgage brand?

Request an AI Visibility Audit from CiteWorks Studio to see where your brand appears, where competitors are recommended instead, which prompts carry the most commercial risk, and which sources are shaping AI-generated mortgage-rate answers.