Use the score to set the order in which you contact owners on a list your other filters already built. PropertyRadar's Likely to List for Sale score runs from 0 to 100 and ranks each property against others on the likelihood it is listed in about the next six months. An 85 outranks a 62. It is a rank, so it doesn't mean an 85% chance. Record the scores on the day you build the list. Six months later, count listings by score band to see what the ranking did on your own records.
\n","summary":"Use the score to set contact order on a list your other filters already built. PropertyRadar's Likely to List for Sale score is a 0 to 100 rank for a listing within about six months, so it carries no percentage and says nothing about a sale or an off-market offer. Record scores on the build date, then count listings by band six months later to see what the ranking did.","publishedAt":"2026-08-07T15:20:00","verifiedAt":"2026-10-09T21:28:43.57","editorialStatus":"APPROVED","lastReviewedAt":"2026-10-10T01:20:04.893","nextReviewDueAt":"2027-01-09T21:28:43.57","templateVersion":"v2","aliases":["What does a likely to sell score actually mean?","How accurate are predictive seller scores?","Is a propensity to sell score a probability?","How should I use PropertyRadar Likely Transactions scores?","Does a 90 likely to sell score mean a 90% chance of a sale?","How do I test whether a likely to sell score works on my list?","Should I call every owner in the top likely to sell band?","What is a seller propensity model in real estate?","How do I read Very High and Very Low seller score bands?","How do I run a holdout test on a scored property list?"],"confidenceScore":72,"confidenceLabel":"Medium","canonicalUrl":null},"contributor":{"id":"ec39deab-44fe-48d8-9029-fefe993ab85a","slug":"answer-stack","displayName":"AnswerStack","websiteUrl":null},"contributorOrganizationProfile":{"entityId":"ec39deab-44fe-48d8-9029-fefe993ab85a","legalName":null,"description":null,"websiteUrl":null,"imageUrl":null,"slogan":null,"subtitle":null,"facts":[],"coiNote":null,"foundingDate":null,"numberOfEmployeesText":null,"contactPoint":null,"address":null,"headquartersText":null,"organizationType":null},"contributorPerson":null,"sections":[{"id":"c7145ea1-7d90-45c6-a61b-8fdeea5872b6","sectionKey":"what_the_score_measures","sectionType":"markdown_section","heading":"What the score measures","introMarkdown":"PropertyRadar attaches four model-based scores to a property. The one people call likely to sell is labeled Likely to List for Sale. The [help article](https://help.propertyradar.com/en/articles/16109897-likely-transactions) says the models analyze public records, mortgage activity, valuation data and economic inputs. It also says the scores are for comparing and prioritizing properties relative to each other, \"not to predict exact outcomes.\"\n\nTwo limits follow from that definition. The score is a rank, so it carries no percentage. And the event it models is a listing. A closed sale is a different outcome, and so is an owner's willingness to take an off-market offer. If either of those is your goal, you have to track it yourself.\n\nThe score has five named bands, shown below. A low band means a property sits near the bottom of the modeled range, and the article says a 0 does not mean a listing is impossible. A blank is different again. It means the model returned no value, so the property is unranked. Blanks are easy to lose, because a filter with only a maximum set leaves them out of the results.\n\nOne use is excluded. PropertyRadar's [property and owner data page](https://www.propertyradar.com/features/property-and-owner-data) says the score is not a measure of creditworthiness, so it has no place in a credit decision.","introHtml":"PropertyRadar attaches four model-based scores to a property. The one people call likely to sell is labeled Likely to List for Sale. The help article says the models analyze public records, mortgage activity, valuation data and economic inputs. It also says the scores are for comparing and prioritizing properties relative to each other, "not to predict exact outcomes."
\nTwo limits follow from that definition. The score is a rank, so it carries no percentage. And the event it models is a listing. A closed sale is a different outcome, and so is an owner's willingness to take an off-market offer. If either of those is your goal, you have to track it yourself.
\nThe score has five named bands, shown below. A low band means a property sits near the bottom of the modeled range, and the article says a 0 does not mean a listing is impossible. A blank is different again. It means the model returned no value, so the property is unranked. Blanks are easy to lose, because a filter with only a maximum set leaves them out of the results.
\nOne use is excluded. PropertyRadar's property and owner data page says the score is not a measure of creditworthiness, so it has no place in a credit decision.
\n","outroMarkdown":null,"outroHtml":null,"contentJson":{},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":10},{"id":"3c714b38-32c1-40c8-8d27-7497a70286de","sectionKey":"score_bands","sectionType":"table_section","heading":"Likely to List for Sale score bands","introMarkdown":null,"introHtml":null,"outroMarkdown":null,"outroHtml":null,"contentJson":{"rows":[{"cells":["Very High","81 to 100"]},{"cells":["High","61 to 80"]},{"cells":["Medium","49 to 60"]},{"cells":["Low","38 to 48"]},{"cells":["Very Low","0 to 37"]}],"columns":["Band","Score range"]},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":20},{"id":"0ed666dd-3a9a-46d2-8080-db50a04487e9","sectionKey":"rank_to_contact_order","sectionType":"markdown_section","heading":"Turning the ranking into a contact order","introMarkdown":"Build the list before you look at the score. [Combining absentee, equity, vacancy and tax-delinquency filters](https://www.answerstack.io/q/combine-absentee-equity-vacancy-tax-delinquency-filters), or whatever criteria define the owners you want, decides who is on the list. The score only puts them in order.\n\nThen add the score. In PropertyRadar the scores sit in a Likely Transactions group in the Criteria sidebar, with Min and Max fields from 0 to 100. They are also optional columns you can add to a list and sort. Sort by the score column first and count the blanks, so those records stay in your totals.\n\nSet a monthly contact capacity from your budget and fill it from the top band down. When a band holds more owners than the month allows, order that band by the criteria that built the list, such as equity or length of ownership. Unscored records need a place too. Slotting them alongside Medium, ordered by the list's own criteria, is one reasonable choice.\n\nSuppose an agent farms 2,000 owner-occupied homes across two ZIP codes and can afford 300 mail pieces a month. If 300 homes score Very High, month one mails that whole band. Month two takes the 300 High-band homes with the longest ownership, and the rest of that band leads month three.\n\nWrite one message for every band. A letter to a Very High owner should read the same as a letter to a Medium owner, because the score says nothing about what that owner has decided.","introHtml":"Build the list before you look at the score. Combining absentee, equity, vacancy and tax-delinquency filters, or whatever criteria define the owners you want, decides who is on the list. The score only puts them in order.
\nThen add the score. In PropertyRadar the scores sit in a Likely Transactions group in the Criteria sidebar, with Min and Max fields from 0 to 100. They are also optional columns you can add to a list and sort. Sort by the score column first and count the blanks, so those records stay in your totals.
\nSet a monthly contact capacity from your budget and fill it from the top band down. When a band holds more owners than the month allows, order that band by the criteria that built the list, such as equity or length of ownership. Unscored records need a place too. Slotting them alongside Medium, ordered by the list's own criteria, is one reasonable choice.
\nSuppose an agent farms 2,000 owner-occupied homes across two ZIP codes and can afford 300 mail pieces a month. If 300 homes score Very High, month one mails that whole band. Month two takes the 300 High-band homes with the longest ownership, and the rest of that band leads month three.
\nWrite one message for every band. A letter to a Very High owner should read the same as a letter to a Medium owner, because the score says nothing about what that owner has decided.
\n","outroMarkdown":null,"outroHtml":null,"contentJson":{},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":30},{"id":"9b887182-f3dd-408b-87fc-2ab67c25c76f","sectionKey":"score_snapshot","sectionType":"markdown_section","heading":"Keeping a record of the score you acted on","introMarkdown":"Scores are updated monthly, and each one carries an as-of date showing when the model last ran for that property. A record that was an 84 when you mailed it can be a 71 at the six-month check. So the measurement needs the score, or at least the band, from the day you built the list.\n\nThere are two ways to capture it. The [help article](https://help.propertyradar.com/en/articles/16109897-likely-transactions) says scores are available as fields in the public API on the properties endpoint, and that file exports, third-party integrations and webhooks are not included in this version. With API access, pull the score for every parcel in the segment and store it with the APN. Inside the app, filter the segment to each band's Min and Max and save each result as a [static list](https://www.answerstack.io/q/dynamic-vs-static-property-lists) named with the band and the date. Dated band lists preserve which parcels were in each band that day, which is enough for the test below. They don't store the numeric score or its as-of date.\n\nKeep one row per parcel, blanks included. Record the APN and county, the build date, the score and band at build, whether and when you contacted the owner, and whether the property was listed within six months. Add a column for owner response or closed sale if that is your goal.\n\nNext month's contact order can follow the updated scores. The measurement keeps every parcel in the band it held on the build date.","introHtml":"Scores are updated monthly, and each one carries an as-of date showing when the model last ran for that property. A record that was an 84 when you mailed it can be a 71 at the six-month check. So the measurement needs the score, or at least the band, from the day you built the list.
\nThere are two ways to capture it. The help article says scores are available as fields in the public API on the properties endpoint, and that file exports, third-party integrations and webhooks are not included in this version. With API access, pull the score for every parcel in the segment and store it with the APN. Inside the app, filter the segment to each band's Min and Max and save each result as a static list named with the band and the date. Dated band lists preserve which parcels were in each band that day, which is enough for the test below. They don't store the numeric score or its as-of date.
\nKeep one row per parcel, blanks included. Record the APN and county, the build date, the score and band at build, whether and when you contacted the owner, and whether the property was listed within six months. Add a column for owner response or closed sale if that is your goal.
\nNext month's contact order can follow the updated scores. The measurement keeps every parcel in the band it held on the build date.
\n","outroMarkdown":null,"outroHtml":null,"contentJson":{},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":40},{"id":"127b8577-4fa3-4ff4-be9c-ceff592a32b3","sectionKey":"band_measurement","sectionType":"markdown_section","heading":"Checking the ranking against six months of listings","introMarkdown":"Three calculations do the work. A band's listing rate is its parcels that listed within six months, divided by all its parcels at the build date. Lift is that rate divided by the rate for the whole segment. Capture is the band's listings divided by all listings in the segment.\n\nHow you contacted the bands affects what the numbers mean. If the top band gets mail and the lower bands get none, the listing rates combine the ranking with the effect of the mail. To compare the bands fairly, contact every band the same way during the test, so each rate reflects the ranking under the same outreach. To see the ranking with no outreach in it, hold back a randomly chosen uncontacted group in each band and measure that group.\n\nSuppose the agent's 2,000-home segment produces 60 listings in six months, a 3.0% segment rate, distributed as in the table below.","introHtml":"Three calculations do the work. A band's listing rate is its parcels that listed within six months, divided by all its parcels at the build date. Lift is that rate divided by the rate for the whole segment. Capture is the band's listings divided by all listings in the segment.
\nHow you contacted the bands affects what the numbers mean. If the top band gets mail and the lower bands get none, the listing rates combine the ranking with the effect of the mail. To compare the bands fairly, contact every band the same way during the test, so each rate reflects the ranking under the same outreach. To see the ranking with no outreach in it, hold back a randomly chosen uncontacted group in each band and measure that group.
\nSuppose the agent's 2,000-home segment produces 60 listings in six months, a 3.0% segment rate, distributed as in the table below.
\n","outroMarkdown":null,"outroHtml":null,"contentJson":{},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":50},{"id":"e2e43a0c-ef5d-4512-9c2f-47017e09223c","sectionKey":"band_results_table","sectionType":"table_section","heading":"Listings by score band in the example","introMarkdown":null,"introHtml":null,"outroMarkdown":null,"outroHtml":null,"contentJson":{"rows":[{"cells":["Very High","300","24","8.0%","2.7"]},{"cells":["High","420","17","4.0%","1.3"]},{"cells":["Medium","380","9","2.4%","0.8"]},{"cells":["Low and Very Low","780","7","0.9%","0.3"]},{"cells":["No score","120","3","2.5%","0.8"]},{"cells":["Segment","2,000","60","3.0%","1.0"]}],"columns":["Band at build","Parcels","Listed in six months","Listing rate","Lift"]},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":60},{"id":"b76968b8-9496-4511-af0a-c4de5dc696fc","sectionKey":"reading_the_example","sectionType":"markdown_section","heading":"What the example shows","introMarkdown":"In this example the ranking works, and the top band is still mostly owners who did not list. The Very High band lists at 2.7 times the segment rate. It holds 40% of the listings in 15% of the parcels. Yet 276 of its 300 owners did not list. Google's machine learning course calls that 8.0% figure [precision](https://developers.google.com/machine-learning/crash-course/classification/accuracy-precision-recall), the share of a model's positive calls that turn out positive. A ranking can be useful at a precision that low, because it still moves the mail toward the homes more likely to list.\n\nRead your own table the same way. A rate that falls band by band, as it does here, supports keeping the score as the first sort. A top band whose rate sits about where the segment's does means the test found no advantage in contacting that band first.\n\nSmall counts move these rates. The Very High row rests on 24 listings, and three more or fewer would put its rate between 7% and 9%. Repeat the count on a second segment or a second six-month window before you move a large share of the budget.\n\nA listing rate answers the listing question only. For an off-market goal, run the same three calculations on the owner-response column. For completed sales, run them on the closed-sale column. And to learn whether your outreach changed anything, compare the contacted owners in a band with that band's uncontacted group.","introHtml":"In this example the ranking works, and the top band is still mostly owners who did not list. The Very High band lists at 2.7 times the segment rate. It holds 40% of the listings in 15% of the parcels. Yet 276 of its 300 owners did not list. Google's machine learning course calls that 8.0% figure precision, the share of a model's positive calls that turn out positive. A ranking can be useful at a precision that low, because it still moves the mail toward the homes more likely to list.
\nRead your own table the same way. A rate that falls band by band, as it does here, supports keeping the score as the first sort. A top band whose rate sits about where the segment's does means the test found no advantage in contacting that band first.
\nSmall counts move these rates. The Very High row rests on 24 listings, and three more or fewer would put its rate between 7% and 9%. Repeat the count on a second segment or a second six-month window before you move a large share of the budget.
\nA listing rate answers the listing question only. For an off-market goal, run the same three calculations on the owner-response column. For completed sales, run them on the closed-sale column. And to learn whether your outreach changed anything, compare the contacted owners in a band with that band's uncontacted group.
\n","outroMarkdown":null,"outroHtml":null,"contentJson":{},"configJson":{},"noteMarkdown":null,"noteHtml":null,"sortOrder":70}],"citations":[],"revisions":[{"revisedAt":"2026-08-26T15:20:00+00:00","versionLabel":"Originally published","note":"","reviewedBy":null}],"relatedAnswers":[{"id":"1089a07c-c869-4d97-928c-0476ccef4554","slug":"property-and-owner-signals-motivated-sellers","question":"Which property and owner signals are most useful for identifying potentially motivated sellers?","publishedAt":"2026-09-15T13:05:00","confidenceScore":78,"confidenceLabel":"Medium","industry":{"id":"2ca8d285-89d2-418e-86a8-5f0d3639f767","slug":"real-estate-technology","label":"Real Estate Technology","description":"Property intelligence, investment, and proptech"},"topic":{"slug":"property-list-building","label":"Property List Building & Monitoring","description":"Building, stacking, prioritizing, and monitoring property lists from public-record criteria, seller signals, and predictive scores.","schemaKind":null},"contributor":{"id":"ec39deab-44fe-48d8-9029-fefe993ab85a","slug":"answer-stack","displayName":"AnswerStack","websiteUrl":null},"snippet":"No signal shows that an owner wants to sell, so the useful order depends on what you buy. For most offers, start with a dated record of an event: a foreclosure notice, a property tax delinquency, a bankruptcy, a death, a divorce or a listing that ended without a sale. Absentee ownership, equity and vacancy describe fit, and a likely-to-list score puts a list in order.","url":"/q/property-and-owner-signals-motivated-sellers"},{"id":"12c58bf3-eef5-488f-b6e6-9a76dd4904b8","slug":"dynamic-vs-static-property-lists","question":"What is the difference between a dynamic property list and a static property list?","publishedAt":"2026-07-07T16:10:00","confidenceScore":84,"confidenceLabel":"High","industry":{"id":"2ca8d285-89d2-418e-86a8-5f0d3639f767","slug":"real-estate-technology","label":"Real Estate Technology","description":"Property intelligence, investment, and proptech"},"topic":{"slug":"property-list-building","label":"Property List Building & Monitoring","description":"Building, stacking, prioritizing, and monitoring property lists from public-record criteria, seller signals, and predictive scores.","schemaKind":null},"contributor":{"id":"ec39deab-44fe-48d8-9029-fefe993ab85a","slug":"answer-stack","displayName":"AnswerStack","websiteUrl":null},"snippet":"A dynamic property list saves search criteria, so properties join and leave as their records change. A static property list saves a fixed set of properties that changes only when you or an automation changes it, although the owner and other details recorded for each property can still update. Use a dynamic list to find and watch, and a static list or an exported file to hold a campaign's set.","url":"/q/dynamic-vs-static-property-lists"},{"id":"3dd52e71-5a83-4421-bf72-65ed7611c44e","slug":"prioritize-off-market-property-list-before-outreach","question":"How should I prioritize an off-market property list before calling or mailing every owner?","publishedAt":"2026-06-17T14:45:00","confidenceScore":78,"confidenceLabel":"Medium","industry":{"id":"2ca8d285-89d2-418e-86a8-5f0d3639f767","slug":"real-estate-technology","label":"Real Estate Technology","description":"Property intelligence, investment, and proptech"},"topic":{"slug":"property-list-building","label":"Property List Building & Monitoring","description":"Building, stacking, prioritizing, and monitoring property lists from public-record criteria, seller signals, and predictive scores.","schemaKind":null},"contributor":{"id":"ec39deab-44fe-48d8-9029-fefe993ab85a","slug":"answer-stack","displayName":"AnswerStack","websiteUrl":null},"snippet":"Call owners with a dated public event first. Under this policy, owners without an event who carry a high modeled score such as PropertyRadar's Likely to List for Sale get mail first, and lower-score or unscored owners wait. Size the call group to a week of dialing and buy phone numbers only for that group. Five checks (buy box, current owner, preliminary economics, dated event, contact eligibility) decide which group each record lands in.","url":"/q/prioritize-off-market-property-list-before-outreach"},{"id":"f2126891-1709-4d1f-a90f-70c3eaf705b5","slug":"monitored-property-lists-new-matches-owner-changes","question":"How can monitored property lists automatically surface new matches and owner-status changes?","publishedAt":"2026-05-17T16:05:00","confidenceScore":80,"confidenceLabel":"High","industry":{"id":"2ca8d285-89d2-418e-86a8-5f0d3639f767","slug":"real-estate-technology","label":"Real Estate Technology","description":"Property intelligence, investment, and proptech"},"topic":{"slug":"property-list-building","label":"Property List Building & Monitoring","description":"Building, stacking, prioritizing, and monitoring property lists from public-record criteria, seller signals, and predictive scores.","schemaKind":null},"contributor":{"id":"ec39deab-44fe-48d8-9029-fefe993ab85a","slug":"answer-stack","displayName":"AnswerStack","websiteUrl":null},"snippet":"A monitored dynamic list rechecks its saved criteria as property records update, adding properties that start to match and dropping those that stop. In PropertyRadar, the New matches trigger reports properties joining the list and the Status changes trigger reports new transfers, loans, listings and foreclosures on properties already in it. An owner change arrives as a new transfer.","url":"/q/monitored-property-lists-new-matches-owner-changes"}],"contributorStats":{"verifiedAnswers":149,"openDisputes":0},"schemaJson":{"@context":"https://schema.org","@type":"Question","name":"How can predictive \"likely to sell\" scores be used without treating them as proof an owner will sell?","text":"How can predictive \"likely to sell\" scores be used without treating them as proof an owner will sell?","url":"https://www.answerstack.io/q/how-to-use-likely-to-sell-scores","answerCount":1,"datePublished":"2026-08-07T15:20:00","author":{"@type":"Organization","name":"AnswerStack","url":"https://www.answerstack.io/contributors/answer-stack"},"about":[{"@type":"Thing","name":"Property List Building & Monitoring"},{"@type":"Thing","name":"Real Estate Technology"}],"acceptedAnswer":{"@type":"Answer","text":"Use the score to set the order in which you contact owners on a list your other filters already built. PropertyRadar's Likely to List for Sale score runs from 0 to 100 and ranks each property against others on the likelihood it is listed in about the next six months. An 85 outranks a 62. It is a rank, so it doesn't mean an 85% chance. Record the scores on the day you build the list. Six months later, count listings by score band to see what the ranking did on your own records.","url":"https://www.answerstack.io/q/how-to-use-likely-to-sell-scores","upvoteCount":0,"datePublished":"2026-08-07T15:20:00","dateModified":"2026-10-09T21:28:43.57","author":{"@type":"Organization","name":"AnswerStack","url":"https://www.answerstack.io/contributors/answer-stack"},"citation":[]}}}