Ralo growth playbook

Rewrite the two answers today: "Ralo is licensed in California, Colorado and Texas" with the licence numbers and a link to NMLS Consumer Access, and a loan-type answer that matches /loans. Then remove the cause: generate the coverage sentence on the FAQ, Beat My Estimate, the footer and the state pages from one licensing list, so the next launch (Florida) updates every page at once. Add three state buttons (California, Colorado, Texas) next to "Get Your Rate" so borrowers confirm their state in one tap, and ask Google to recrawl /faq and /beat-my-estimate. Finally, ask the main AI assistants and search "is Ralo available in California" before and after, and keep a note of what they say.

A borrower who reads that a broker only works in another state leaves without asking; nothing in the funnel recovers them. The fix costs an hour, and it unlocks the California and Colorado pages Ralo has already paid to build, in two states where it has only one published case study between them.

[ ] /faq: "Which states does Ralo operate in?" names California, Colorado and Texas
[ ] /faq: loan types match /loans (FHA, VA, USDA, DSCR, bank statement, HELOC)
[ ] /beat-my-estimate: FAQ block matches its header
[ ] FAQ structured data updated with the same text
[ ] Footer, licensing page and state pages read from one list
[ ] State buttons beside "Get Your Rate"
[ ] Recrawl requested for /faq and /beat-my-estimate
[ ] Before/after record of search snippets and AI-assistant answers

• The answer goes stale again at the Florida launch

  Generate it from the licensing list, not hand-written copy.

• AI assistants keep repeating the old answer for a while

  Repeat the coverage line on the licensing, About and state pages so the new answer appears in several places.

• Your websiteFAQ (checked twice on November 10, 2026): "Which states does Ralo operate in?" is answered "Ralo is licensed as a mortgage broker in Texas and is expanding to additional states", while the footer of the same page lists California (DRE MLO endorsement #02386277), Colorado and Texas; "What types of mortgages does Ralo support?" names only conventional, jumbo, ARM and refinance; Ralo is paid by the lender after closing at "a fraction of the standard commission"; no lender count and no answer on credit pulls.Source
• Your websiteLicensing page: "Ralo is currently authorized to conduct business in California, Colorado, and Texas"; California DRE licence issued 03/11/2026, Colorado mortgage company registration issued 02/09/2026, Texas SML mortgage company licence issued 10/03/2025.Source
• Your websiteBeat My Estimate page: borrowers upload a Loan Estimate (one PDF or up to three photos) for a side-by-side check of rate, payment, upfront costs, points, credits and origination charges; the header says Ralo serves "California, Colorado, and Texas", but the FAQ on the same page repeats the Texas-only answer.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• Your websiteFHA loan page: Ralo "brokers FHA loans in the states where it is licensed", with a "Get Your FHA Rate" button; /loans also lists VA, USDA, DSCR, bank-statement and HELOC loans.Source
• Your websiteCase studies page: 16 case studies, 15 from Texas, 1 from California and none from Colorado; 10 homebuyers and 6 refinances; rates from 4.875% to 6.875%; closings in 7 to 28 days; the page promises "what it saves them over the life of the loan", but no case study shows a dollar saving.Source

Add state, credit band and down payment to /rates and show the matching illustrative range from the weekly report's own tables, against the PMMS average, with the APR and disclaimer. Carry the ZIP from the page the visitor came from. Then offer the next step: sign up for live quotes, or upload a Loan Estimate.

A personal number before a sign-up keeps the "no login required" promise and gives borrowers a reason to hand over their details. The data already exists, so this is a page change, not a pricing engine.

• Borrowers treat the range as a quote

  Label it illustrative and date it, as the report does.

• Your websiteRates page: Ralo's 30-year fixed at 6.49% (6.53% APR, November 9, 2026) against the PMMS 7.18% (November 5, 2026); no inputs for state, ZIP code, credit score or down payment; the call to action is "Sign Up with Ralo".Source
• Your websiteZIP-code page (90210): illustrative rates by credit score and down payment (30-year fixed at 6.75% for a 740 score and 20% down), median home value "$2,000,000+"; the calls to action "See All-In Costs" and "Compare Quotes" both go to /rates; no quote form.Source
• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source
• Your websiteHomepage: headline "Rates Worth Bragging About" and "We automated mortgages. Enjoy the savings."; a sample 30-year fixed at 6.500% (6.538% APR) for a $500,000 loan at 75% LTV and 780+ FICO beside a "MARKET RATE TODAY" of 7.180% from Freddie Mac PMMS; "0 Points" and "0 Lender Fees"; "Get Your Rate" with "No login required"; five testimonials, all from Texas, showing savings of $42,710 to $111,245, one from a borrower who found Ralo through an AI assistant; "As seen" logos for Forbes, Business Insider and The New York Times.Source

Replace the Florida calls to action with "Tell me when Ralo opens in Florida": email, city, buying or refinancing, and timeline. Show the count of people waiting. On launch day, email the list with the first Florida rates and a Beat My Estimate link. Reuse the same block for any future state.

The pages already rank and bring visitors; a waitlist turns them into a first-week pipeline instead of a bounce, and tells Ralo which Florida cities to prioritise.

• The launch slips and the list goes cold

  Send one short update with Florida rate data while waiting.

• Your websiteFlorida page: "Ralo is not yet licensed in Florida and cannot take applications there today", with launch planned for November 2026; its calls to action ("Explore Florida rates") go to /rates; no waitlist, only the general newsletter.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• Your websiteSeed round post (June 18, 2026): $2.9M from Y Combinator, Manresa Ventures, Pack Ventures and angels; rates "more than half a point below the national average" and closings in about 15 days; plans to expand into new states "where we can save homebuyers the most".Source

For every case study that fits the published method, add the same-week PMMS rate, Ralo's rate, the monthly payment difference and the lifetime saving, with a link to the method. For refinances and other loans outside the method, show the monthly saving against the borrower's previous rate or quote instead, labelled clearly. Add a sortable list by state, loan purpose and saving.

A dated, checkable number is far more convincing than a testimonial line, and it is the proof Own Up leads with ($30,000 average). Ralo's Texas average of about $60,000 is a stronger story if each case shows its sum.

• Savings look overstated for borrowers who sell early

  Show the monthly saving first and the lifetime figure second, with the method's hold-to-term note.

• Your websiteCase studies page: 16 case studies, 15 from Texas, 1 from California and none from Colorado; 10 homebuyers and 6 refinances; rates from 4.875% to 6.875%; closings in 7 to 28 days; the page promises "what it saves them over the life of the loan", but no case study shows a dollar saving.Source
• Your websiteSavings methodology: each borrower's rate against the same-week Freddie Mac PMMS, for conventional conforming fixed-rate purchase loans with 740+ FICO, 75-80% LTV and zero points; jumbo, ARM, government-backed and refinance loans are excluded; the sample size is not stated; a worked example saves $70,629 on a $400,000 loan.Source
• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source
• Competitor researchRivals' own homepages: Own Up, "Shop for your mortgage and save thousands", an average $30,000 lifetime saving and 4.9 from 4,690 verified reviews, paid by lenders that advertise; Better, "Simple, Online, AI-Powered Mortgage", "3 min" to start with "No credit impact".See the research Source

Add a two-button choice to every ZIP, city, county, loan and calculator page: "I already have a quote: upload my Loan Estimate" and "I'm just starting: see my rate". Put the upload box inline on the refinance calculator, which already reads sections D and H of the Loan Estimate. Track uploads by page type.

Borrowers with a Loan Estimate are close to a decision, and the comparison shows Ralo's value in their own numbers, as the reviews describe ("matched or beat competing offers").

• More uploads than the team can answer quickly

  Automate the first comparison and queue the human check.

• Your websiteBeat My Estimate page: borrowers upload a Loan Estimate (one PDF or up to three photos) for a side-by-side check of rate, payment, upfront costs, points, credits and origination charges; the header says Ralo serves "California, Colorado, and Texas", but the FAQ on the same page repeats the Texas-only answer.Source
• Your websiteZIP-code page (90210): illustrative rates by credit score and down payment (30-year fixed at 6.75% for a 740 score and 20% down), median home value "$2,000,000+"; the calls to action "See All-In Costs" and "Compare Quotes" both go to /rates; no quote form.Source
• Your websiteRefinance calculator: break-even month and total savings from Loan Estimate sections D and H; next steps "See personalized rates" and "Book a call with a loan officer"; no rate alert, only a general newsletter signup.Source
• Your websiteReviews page: a 5.0 average from 37 Google reviews, the only platform shown; the named testimonials are all from Texas; recurring themes are responsiveness, fast closings, competitive rates and matching or beating competing offers.Source

Add "Ralo (formerly Approval AI)" to the press page entries, the About page and the "Is Ralo legit?" page, with a one-line note on the rename. Link each logo to its article. Mention the former name in the Organization structured data so search engines connect the two.

Trust is the main hurdle for a new online broker, and a visible name history turns a possible red flag into extra credibility.

• Drawing attention to an old name

  Keep it short and factual; it answers a question borrowers already ask.

• Your websitePress page: 24 items; a September 15, 2026 release on more than $8M funded for Texas buyers; the 2025 Forbes and Business Insider pieces behind the homepage logos name Approval AI, Ralo's former name; a July 2026 New York Times piece; a HousingWire column by the co-founder and CTO; no press kit beyond a wordmark and a founder photo.Source
• Your websiteHomepage: headline "Rates Worth Bragging About" and "We automated mortgages. Enjoy the savings."; a sample 30-year fixed at 6.500% (6.538% APR) for a $500,000 loan at 75% LTV and 780+ FICO beside a "MARKET RATE TODAY" of 7.180% from Freddie Mac PMMS; "0 Points" and "0 Lender Fees"; "Get Your Rate" with "No login required"; five testimonials, all from Texas, showing savings of $42,710 to $111,245, one from a borrower who found Ralo through an AI assistant; "As seen" logos for Forbes, Business Insider and The New York Times.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source

Render the hero rate on the server so it does not wait for scripts, load Tag Manager and the Facebook pixel after the page is interactive, and pick one owner for each analytics job: PostHog for product events, one tool for ad attribution. Set long cache lifetimes for static files. Retest the homepage and the ZIP-code template.

Faster first paint helps the visitors on weak connections who arrive from ads and ZIP pages, and fewer overlapping tools means one set of numbers to trust.

• Delayed pixels lose some ad attribution

  Use server-side conversion events for sign-ups.

• PageSpeedGoogle PageSpeed (November 10, 2026): mobile score 60, desktop 84; in the lab test the mobile hero (the rate section) takes 12.5s to appear, with 322KB of unused JavaScript (about 1.9s), Google Tag Manager at 300KB and Facebook at 210KB, alongside Google Analytics, Clarity and PostHog; real Chrome visitors pass Core Web Vitals (p75 LCP 1.43s, INP 70ms, CLS 0.01) and LCP is 13% faster than before.Source

Ask every California and Colorado borrower at closing for a short case study (rate, saving, days to close) and a Google review. Place them on the matching state, city and county pages and in the homepage testimonial strip by visitor state. Until there are enough, show the weekly report's California range against PMMS on those pages.

Local proof answers the unspoken question "does this work where I live?", and it supports the coverage fix in idea 1.

• Few non-Texas closings yet

  Start with one California story and the rate data; add as loans close.

• #1 Your FAQ still tells California and Colorado borrowers that Ralo only works in Texas
• #4 Put the dollar saving on each of the 16 case studies, using your own published method

• Your websiteCase studies page: 16 case studies, 15 from Texas, 1 from California and none from Colorado; 10 homebuyers and 6 refinances; rates from 4.875% to 6.875%; closings in 7 to 28 days; the page promises "what it saves them over the life of the loan", but no case study shows a dollar saving.Source
• Your websiteCalifornia page: "Ralo is currently licensed in California"; median home price $695,400; 2026 county loan limits such as $1,249,125 for Los Angeles County; a savings example of $132 a month and $47,348 over 30 years; no California testimonials.Source
• Your websiteReviews page: a 5.0 average from 37 Google reviews, the only platform shown; the named testimonials are all from Texas; recurring themes are responsiveness, fast closings, competitive rates and matching or beating competing offers.Source
• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source

After the result, offer "Alert me when a refinance would break even within X months", saving the borrower's loan details with consent. Each week, rerun every saved scenario against Ralo's new rates and email those who cross the threshold with their new break-even and a Beat My Estimate link.

It captures intent at the moment of research and contacts people exactly when the numbers work for them, which no general newsletter can do.

• Rates rise and alerts never fire

  Send a short monthly status so subscribers stay warm.

• Your websiteRefinance calculator: break-even month and total savings from Loan Estimate sections D and H; next steps "See personalized rates" and "Book a call with a loan officer"; no rate alert, only a general newsletter signup.Source
• Market dataFreddie Mac PMMS 30-year fixed at 7.18% for the week of November 5, 2026, down from 7.24% and the third straight weekly decline; 6.22% a year earlier.Source
• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source

Publish /agents: how Ralo works with agents, closing speed (about 19 days), communication promises, and a one-buyer trial. Give each agent a link that sends their buyer to Beat My Estimate and, with the buyer's permission, shares a one-page status update with the agent at each milestone. Start with agents in the Texas cities where Ralo has already closed loans.

Agents choose lenders on reliability and speed; a public page and visible updates make Ralo easy to recommend, and one agent can send many buyers a year.

• Referral rules for settlement services

  Offer service, not payment, and have counsel review the programme.

• #6 Make "Upload your Loan Estimate" the main button on the 369 ZIP-code pages

• Your websiteBlog post for agents ("Realtor Lender Blind Spot"): asks agents to "consider adding a second option to your toolkit" and to send Ralo one buyer as a trial, with a "Book a time to chat" link; no formal agent programme or page.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source

Put a licensed author's name and NMLS number on each report, add a two-line "what changed" summary, and send it every Monday to housing reporters in Texas, California and Colorado with one local number (for example, the California range against PMMS). Offer a short quote on request.

Reporters need fresh, local rate data each week; a named, licensed source becomes the person they call, and each mention links back to Ralo.

• Illustrative rates quoted as offers

  Include the disclaimer and assumptions in every pitch.

• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source
• Your websitePress page: 24 items; a September 15, 2026 release on more than $8M funded for Texas buyers; the 2025 Forbes and Business Insider pieces behind the homepage logos name Approval AI, Ralo's former name; a July 2026 New York Times piece; a HousingWire column by the co-founder and CTO; no press kit beyond a wordmark and a founder photo.Source
• PressHousingWire column by Ralo's co-founder (August 25, 2026): producing one mortgage costs the industry $11,898 (MBA data); a "loan officer engineer" is "a licensed loan originator who also improves the system that originates loans"; "Encoded judgment compounds"; the industry takes 38.2 days on average to close.Source

Publish a short, dated "Ralo at a glance" page: states and licence numbers, loan types, how compensation works, average days to close, the savings method and the current rate gap to PMMS. Link it from the footer. Add "AI assistant" as an option in the sign-up source question, and check monthly what the main assistants say when asked for a mortgage broker in Texas, California and Colorado.

Borrowers increasingly ask assistants to compare lenders; consistent, citable facts make Ralo easier to recommend correctly, and the source question shows whether the channel is growing.

• Answers lag behind site changes

  Keep facts identical across the facts page, FAQ and licensing page.

• #1 Your FAQ still tells California and Colorado borrowers that Ralo only works in Texas

• Your websiteHomepage: headline "Rates Worth Bragging About" and "We automated mortgages. Enjoy the savings."; a sample 30-year fixed at 6.500% (6.538% APR) for a $500,000 loan at 75% LTV and 780+ FICO beside a "MARKET RATE TODAY" of 7.180% from Freddie Mac PMMS; "0 Points" and "0 Lender Fees"; "Get Your Rate" with "No login required"; five testimonials, all from Texas, showing savings of $42,710 to $111,245, one from a borrower who found Ralo through an AI assistant; "As seen" logos for Forbes, Business Insider and The New York Times.Source
• Your websiteFAQ (checked twice on November 10, 2026): "Which states does Ralo operate in?" is answered "Ralo is licensed as a mortgage broker in Texas and is expanding to additional states", while the footer of the same page lists California (DRE MLO endorsement #02386277), Colorado and Texas; "What types of mortgages does Ralo support?" names only conventional, jumbo, ARM and refinance; Ralo is paid by the lender after closing at "a fraction of the standard commission"; no lender count and no answer on credit pulls.Source
• Your websiteSavings methodology: each borrower's rate against the same-week Freddie Mac PMMS, for conventional conforming fixed-rate purchase loans with 740+ FICO, 75-80% LTV and zero points; jumbo, ARM, government-backed and refinance loans are excluded; the sample size is not stated; a worked example saves $70,629 on a $400,000 loan.Source

Enrol every funded borrower in a quiet rate watch: Ralo reruns their loan weekly and emails only when a refinance would break even within an agreed number of months, with the numbers already filled in from their file. Add a yearly "mortgage check-up" covering PMI removal, escrow and recast options.

Past borrowers already trust Ralo and their data is on file, so a refinance costs far less to win than a new borrower, and timely, accurate alerts earn reviews and referrals.

• Borrowers feel sold to

  Only email when the numbers work, and say so in the first message.

• #7 Turn the refinance calculator into a rate-drop alert: "tell me when refinancing pays off"

• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source
• Your websiteCase studies page: 16 case studies, 15 from Texas, 1 from California and none from Colorado; 10 homebuyers and 6 refinances; rates from 4.875% to 6.875%; closings in 7 to 28 days; the page promises "what it saves them over the life of the loan", but no case study shows a dollar saving.Source
• Market dataFreddie Mac PMMS 30-year fixed at 7.18% for the week of November 5, 2026, down from 7.24% and the third straight weekly decline; 6.22% a year earlier.Source
• Your websiteRefinance calculator: break-even month and total savings from Loan Estimate sections D and H; next steps "See personalized rates" and "Book a call with a loan officer"; no rate alert, only a general newsletter signup.Source

Publish fair, dated pages against Rocket and Own Up, and upgrade the Better page, each with a worked example: the same borrower's rate, points and lender fees from a recent Ralo quote beside the rival's published offer or model, plus a clear note on how each company is paid. End each page with Beat My Estimate.

"X vs Y" searches come from borrowers close to choosing, and line items are Ralo's strongest argument.

• Out-of-date rival claims

  Date every figure and link the rival's own page.

• Your websiteRalo vs Better page: Ralo is a broker paid "a thin commission", Better a direct lender licensed in 50 states and DC; Ralo's products are listed as conventional, jumbo, ARM, FHA, DSCR, bank statement and HELOC; no rates, APRs or fee amounts are shown.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• Competitor researchRivals' own homepages: Own Up, "Shop for your mortgage and save thousands", an average $30,000 lifetime saving and 4.9 from 4,690 verified reviews, paid by lenders that advertise; Better, "Simple, Online, AI-Powered Mortgage", "3 min" to start with "No credit impact".See the research Source

Build the review request into the closing step: the loan officer engineer sends a one-tap Google review link with the borrower's own saving, the day the loan funds. Add a second platform popular with homebuyers, show the live count on the homepage, and reply to every review.

Borrowers are happiest on closing day; asking then, with their saving in the message, gets the highest response.

• Incentivised reviews breach platform rules

  Never pay or reward for reviews.

• Your websiteReviews page: a 5.0 average from 37 Google reviews, the only platform shown; the named testimonials are all from Texas; recurring themes are responsiveness, fast closings, competitive rates and matching or beating competing offers.Source
• Competitor researchRivals' own homepages: Own Up, "Shop for your mortgage and save thousands", an average $30,000 lifetime saving and 4.9 from 4,690 verified reviews, paid by lenders that advertise; Better, "Simple, Online, AI-Powered Mortgage", "3 min" to start with "No credit impact".See the research Source

Feed each page from the weekly report: "This week the national average is 7.18%; Ralo's illustrative rate for a 740 score with 20% down is 6.5%; so 6.5% is about average for Ralo borrowers and well under the market." Show the borrower's position by credit band, date it, and end with "Upload your Loan Estimate to check yours".

A dated, specific answer beats evergreen advice in search and AI answers, and the reader is already holding the document Beat My Estimate needs.

• Wrong data if the feed breaks

  Hide the box when the data is older than ten days.

• #6 Make "Upload your Loan Estimate" the main button on the 369 ZIP-code pages

• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source
• Your websiteBeat My Estimate page: borrowers upload a Loan Estimate (one PDF or up to three photos) for a side-by-side check of rate, payment, upfront costs, points, credits and origination charges; the header says Ralo serves "California, Colorado, and Texas", but the FAQ on the same page repeats the Texas-only answer.Source

Add a "Get pre-approved" path on the first-time buyer pages and the homepage: a short application, a licensed review and a letter the same business day, with an option to send it to the buyer's agent. When the buyer goes under contract, Ralo shops the lenders as usual.

It brings Ralo in before the buyer commits to another lender, and it gives agents (idea 8) a reason to send buyers early.

• Many pre-approvals never buy

  Keep the review light until the buyer is under contract.

• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source
• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• Your websiteBlog post for agents ("Realtor Lender Blind Spot"): asks agents to "consider adding a second option to your toolkit" and to send Ralo one buyer as a trial, with a "Book a time to chat" link; no formal agent programme or page.Source

Publish a quarterly, audited-in-spirit "cost to make a loan" figure: Ralo's fully loaded cost per funded loan, how it falls each quarter as loan officer engineers encode more decisions, and how much of the saving reaches the borrower's rate. Use it on the homepage, in recruiting and in investor updates.

It turns the reason Ralo is cheaper into a number borrowers, reporters and future hires can check, and it is very hard for a traditional lender to match publicly.

• Low volume makes the number noisy

  Publish a rolling two-quarter average.

• Reveals margin to rivals

  Publish cost, not commission rates.

• PressHousingWire column by Ralo's co-founder (August 25, 2026): producing one mortgage costs the industry $11,898 (MBA data); a "loan officer engineer" is "a licensed loan originator who also improves the system that originates loans"; "Encoded judgment compounds"; the industry takes 38.2 days on average to close.Source
• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source
• HiringCareers page: one open role, Loan Officer Engineer (New York City, full-time, in person): "Originate loans and improve the platform behind every file", listed on Y Combinator's Work at a Startup.Source

Score candidate states on three measures Ralo can see: its quoted rate gap to local averages, loan sizes (savings scale with balance), and demand from visitors and waitlists on existing pages. Build city pages and the waitlist block (idea 3) three months before each licence, and launch with the list, local rate data and a Beat My Estimate offer.

Each launch starts with a pipeline instead of from zero, and the choice of state rests on where Ralo's model wins most, which also makes a strong press story.

• Lender coverage varies by state

  Confirm wholesale lender approvals before announcing.

• #3 Collect Florida borrowers now on the Florida pages that cannot take applications yet

• Your websiteSeed round post (June 18, 2026): $2.9M from Y Combinator, Manresa Ventures, Pack Ventures and angels; rates "more than half a point below the national average" and closings in about 15 days; plans to expand into new states "where we can save homebuyers the most".Source
• Your websiteWeekly rate report (November 9, 2026): Ralo's illustrative 30-year fixed at 6.5%, "0.68 percentage points below the Freddie Mac PMMS average of 7.18%"; state ranges for a 740 score and 20% down of 6.75%-7.25% in California and 6.5%-6.875% in Texas; no named author; a newsletter signup.Source
• Your websiteFlorida page: "Ralo is not yet licensed in Florida and cannot take applications there today", with launch planned for November 2026; its calls to action ("Explore Florida rates") go to /rates; no waitlist, only the general newsletter.Source

Package a jumbo offer for California: a jumbo page per high-cost county with Ralo's jumbo pricing against local averages, a jumbo version of the savings method, a dedicated loan officer engineer, and outreach to financial advisers and relocation managers in the Bay Area and Los Angeles. Track jumbo applications as their own funnel.

Jumbo borrowers shop hard and the dollar savings are large, so a small number of loans can move revenue a lot.

• Jumbo lender appetite changes

  Keep several jumbo lenders approved.

• #1 Your FAQ still tells California and Colorado borrowers that Ralo only works in Texas
• #5 Win California and Colorado proof: one story per state a month, placed on the pages borrowers land on

• Your websiteCalifornia page: "Ralo is currently licensed in California"; median home price $695,400; 2026 county loan limits such as $1,249,125 for Los Angeles County; a savings example of $132 a month and $47,348 over 30 years; no California testimonials.Source
• Your websiteZIP-code page (90210): illustrative rates by credit score and down payment (30-year fixed at 6.75% for a 740 score and 20% down), median home value "$2,000,000+"; the calls to action "See All-In Costs" and "Compare Quotes" both go to /rates; no quote form.Source
• Your websiteSavings methodology: each borrower's rate against the same-week Freddie Mac PMMS, for conventional conforming fixed-rate purchase loans with 740+ FICO, 75-80% LTV and zero points; jumbo, ARM, government-backed and refinance loans are excluded; the sample size is not stated; a worked example saves $70,629 on a $400,000 loan.Source
• Your websiteCase studies page: 16 case studies, 15 from Texas, 1 from California and none from Colorado; 10 homebuyers and 6 refinances; rates from 4.875% to 6.875%; closings in 7 to 28 days; the page promises "what it saves them over the life of the loan", but no case study shows a dollar saving.Source

With explicit consent at upload, aggregate Loan Estimate data each quarter: median origination charges and points by state and loan type, the share of quotes Ralo beat and by how much. Publish only totals with a minimum group size, and pitch it to housing reporters alongside the weekly report.

Original data earns coverage and links, and each finding ends with the same call to action: upload yours.

• Too few uploads for meaningful totals

  Wait until each group has enough files; start with Texas only.

• Borrower trust

  Opt-in only, aggregate only, stated plainly.

• #6 Make "Upload your Loan Estimate" the main button on the 369 ZIP-code pages

• Your websiteBeat My Estimate page: borrowers upload a Loan Estimate (one PDF or up to three photos) for a side-by-side check of rate, payment, upfront costs, points, credits and origination charges; the header says Ralo serves "California, Colorado, and Texas", but the FAQ on the same page repeats the Texas-only answer.Source
• Your websiteHomepage: headline "Rates Worth Bragging About" and "We automated mortgages. Enjoy the savings."; a sample 30-year fixed at 6.500% (6.538% APR) for a $500,000 loan at 75% LTV and 780+ FICO beside a "MARKET RATE TODAY" of 7.180% from Freddie Mac PMMS; "0 Points" and "0 Lender Fees"; "Get Your Rate" with "No login required"; five testimonials, all from Texas, showing savings of $42,710 to $111,245, one from a borrower who found Ralo through an AI assistant; "As seen" logos for Forbes, Business Insider and The New York Times.Source
• PressHousingWire column by Ralo's co-founder (August 25, 2026): producing one mortgage costs the industry $11,898 (MBA data); a "loan officer engineer" is "a licensed loan originator who also improves the system that originates loans"; "Encoded judgment compounds"; the industry takes 38.2 days on average to close.Source

At funding, give each borrower (with consent) a simple card: "I got 6.5% when the market was 7.18%", with their saving and a personal link to Beat My Estimate. Make it easy to text to a friend or post. Track referred uploads. Any thank-you for referrals should be approved by counsel first.

Homebuyers talk about rates with friends who are also buying; a specific number is more persuasive than a recommendation, and it puts the brand line to work.

• Borrowers do not want to share loan details

  Let them choose rate only, saving only, or just the link.

• #14 Ask for a review at the moment of closing, and grow from 37 Google reviews toward hundreds

• Your websiteHomepage: headline "Rates Worth Bragging About" and "We automated mortgages. Enjoy the savings."; a sample 30-year fixed at 6.500% (6.538% APR) for a $500,000 loan at 75% LTV and 780+ FICO beside a "MARKET RATE TODAY" of 7.180% from Freddie Mac PMMS; "0 Points" and "0 Lender Fees"; "Get Your Rate" with "No login required"; five testimonials, all from Texas, showing savings of $42,710 to $111,245, one from a borrower who found Ralo through an AI assistant; "As seen" logos for Forbes, Business Insider and The New York Times.Source
• Your websiteReviews page: a 5.0 average from 37 Google reviews, the only platform shown; the named testimonials are all from Texas; recurring themes are responsiveness, fast closings, competitive rates and matching or beating competing offers.Source
• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source

Send a short series after funding: first payment and servicer transfer, escrow review, when the borrower can ask to remove PMI given their balance and home value, and a yearly check-up. Each email uses their own loan figures and offers a reply to a loan officer engineer.

Practical help after closing builds the trust that produces reviews, referrals and the next loan, and it costs almost nothing once written.

• Servicer rules differ

  Point borrowers to their servicer for the final PMI request.

• #12 A rate watch for every borrower Ralo has funded, so their refinance comes back to Ralo

• Your websiteSitemap (824 URLs, checked November 10, 2026): 369 ZIP-code pages; 124 /mortgages pages (state pages for California, Colorado, Florida and Texas, 48 city and 61 county pages, of which 39 are in California, 15 in Colorado, 46 in Texas and 9 in Florida, plus 10 neighbourhood pages); 84 glossary terms; 48 cost-of-living pages; 16 /compare pages where Better is the only named rival; 13 tools including state calculators for Texas, California, Colorado and Florida; a weekly mortgage-rate report; 10 "is X% a good mortgage rate" pages; 8 credit-score pages; 4 first-time-buyer pages; no page for real estate agents.Source
• PressPulse 2.0 (September 15, 2026): more than $8 million funded in six months in Texas; Texas customers save about $60,000 over the life of their loans (company figure); about 19 days to close against 30 to 45; rates up to 0.6 points below the national average; about one-quarter the cost of traditional brokers; many borrowers are first-time buyers.Source