[{"data":1,"prerenderedAt":191},["ShallowReactive",2],{"blog-data-driven-adjustments-from-past-reports":3},{"post":4,"related":78,"pillar":79},{"slug":5,"title":6,"seoTitle":7,"description":8,"date":9,"updated":10,"category":11,"pillar":12,"howto":12,"keywords":13,"tags":22,"related":27,"faq":31,"words":47,"minutes":48,"parts":49,"toc":52,"steps":77},"data-driven-adjustments-from-past-reports","Your Past Reports Are a Data Set: Data-Driven Adjustments from Your Own Work","Data-Driven Appraisal Adjustments | Carat","Data-driven appraisal adjustments from your own work: paired sales, grouped data, regression, your own medians, and how a firm archive becomes support.","2026-10-20","2026-10-25","ai",false,[14,15,16,17,18,19,20,21],"appraisal adjustments","data-driven adjustments","paired sales analysis","grouped data analysis appraisal","regression appraisal adjustments","support appraisal adjustments","gla adjustment per square foot","machine learning appraisal",[23,24,25,26],"Adjustments","AI","Paired Sales","Workfile",[28,29,30],"ai-appraisal-software-2026","switch-appraisal-software-without-losing-comps","ai-catches-amc-revision-requests",[32,35,38,41,44],{"q":33,"a":34},"How do appraisers support adjustments?","With market evidence: paired sales, grouped data analysis, regression, cost and depreciation analysis, and other commonly accepted methods. Fannie Mae's Selling Guide says adjustments must reflect the market's reaction and that a statement only recognizing an adjustment was made is not acceptable.",{"q":36,"a":37},"What is paired sales analysis?","Paired sales analysis compares two sales that are alike except for one feature, so the difference in price (after accounting for time and other differences) indicates what the market paid for that feature.",{"q":39,"a":40},"Can I use my own past appraisals to support adjustments?","Your past reports hold verified sales and the adjustments you developed, which makes them a useful data set for checking consistency and finding pairs. The support in a new report still has to reflect the current market, so treat your history as evidence to test, not a rate to copy.",{"q":42,"a":43},"Does Fannie Mae limit net and gross adjustments?","No. Fannie Mae's Selling Guide says it does not have specific limitations or guidelines associated with net or gross adjustments. Adjustments should reflect the market's reaction to differences.",{"q":45,"a":46},"Is regression better than paired sales?","Neither is better in every case. Regression uses many sales and handles several variables at once, but needs enough clean data. Paired sales are easy to explain but hard to find in thin markets. Many appraisers use both and reconcile.",1683,7,[50,51],"\u003Cp>"Please provide support for your adjustments." Every appraiser has read that line in a revision request. Most of us have also written "adjustments are market derived" in commentary and hoped it would be enough. It isn't, and under UAD 3.6 it's even less likely to pass a reviewer. The good news: you're sitting on years of evidence. Every report you've delivered holds sales you verified and adjustments you defended. This guide covers the main methods for data-driven appraisal adjustments (paired sales, grouped data, regression) and how to turn your own archive into support you can cite.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Key takeaways\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Fannie Mae requires adjustments that reflect the market's reaction and accepts statistical analysis, modeling, paired sales and other commonly accepted methods.\u003C\u002Fli>\n\u003Cli>Paired sales are easy to explain but rare. Grouped data uses more sales. Regression handles several variables at once.\u003C\u002Fli>\n\u003Cli>Your own past reports are a clean, verified data set. Your medians by town and price tier are a consistency check and a starting point.\u003C\u002Fli>\n\u003Cli>History supports; it doesn't decide. The current market and the current comps still drive each adjustment.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"what-the-gses-expect\">What the GSEs expect\u003C\u002Fh2>\n\u003Cp>Fannie Mae's Selling Guide section on \u003Ca href=\"https:\u002F\u002Fselling-guide.fanniemae.com\u002Fsel\u002Fb4-1.3-09\u002Fadjustments-comparable-sales\" target=\"_blank\" rel=\"noopener\">adjustments to comparable sales\u003C\u002Fa> is short and worth reading in full. The points that matter most:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>Adjustments \u003Cstrong>must reflect the market's reaction\u003C\u002Fstrong> to differences, which the guide calls market-based adjustments.\u003C\u002Fli>\n\u003Cli>It gives an example: a rule-of-thumb adjustment of $20 per square foot is inappropriate when market analysis indicates $100.\u003C\u002Fli>\n\u003Cli>Adjustment rates can be determined \u003Cstrong>"through statistical analysis, modeling, paired sales, or other commonly accepted methods."\u003C\u002Fstrong>\u003C\u002Fli>\n\u003Cli>\u003Cstrong>"A statement only recognizing that an adjustment has been made is not acceptable."\u003C\u002Fstrong>\u003C\u002Fli>\n\u003Cli>Fannie Mae \u003Cstrong>does not have specific limitations or guidelines for net or gross adjustments.\u003C\u002Fstrong> If you learned the old 15% net \u002F 25% gross rule of thumb, it isn't in the current guide. Large adjustments are fine if they're supported.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>So the question is never "is this adjustment big?" It's "can you show where it came from?"\u003C\u002Fp>\n\u003Ch2 id=\"method-1-paired-sales\">Method 1: Paired sales\u003C\u002Fh2>\n\u003Cp>Find two sales that are alike except for one feature. The price difference, after accounting for anything else that differs, indicates what the market paid for that feature.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>A worked example (illustrative numbers):\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cdiv class=\"bl-table\">\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>\u003C\u002Fth>\u003Cth>Sale A\u003C\u002Fth>\u003Cth>Sale B\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>Subdivision\u003C\u002Ftd>\u003Ctd>Same\u003C\u002Ftd>\u003Ctd>Same\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>GLA\u003C\u002Ftd>\u003Ctd>1,850\u003C\u002Ftd>\u003Ctd>1,850\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Beds \u002F baths\u003C\u002Ftd>\u003Ctd>3 \u002F 2\u003C\u002Ftd>\u003Ctd>3 \u002F 2\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Condition\u003C\u002Ftd>\u003Ctd>C3\u003C\u002Ftd>\u003Ctd>C3\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Garage\u003C\u002Ftd>\u003Ctd>1-car\u003C\u002Ftd>\u003Ctd>2-car\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Sale date\u003C\u002Ftd>\u003Ctd>March\u003C\u002Ftd>\u003Ctd>April\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Price\u003C\u002Ftd>\u003Ctd>$342,000\u003C\u002Ftd>\u003Ctd>$356,000\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003C\u002Fdiv>\n\u003Cp>If the market was flat between March and April, the pair indicates about $14,000 for the second garage bay. One pair isn't proof. Three or four pairs pointing to $12,000 to $15,000 is support.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Strengths:\u003C\u002Fstrong> easy to explain to a reviewer, a borrower or a judge.\n\u003Cstrong>Weaknesses:\u003C\u002Fstrong> truly matched pairs are rare, especially in rural or mixed markets. Small samples are noisy.\u003C\u002Fp>\n\u003Cblockquote class=\"bl-callout watch-out\">\u003Cp>\u003Cstrong>Watch out:\u003C\u002Fstrong> A "pair" that differs in two things (garage and condition, say) doesn't isolate either one. Adjust for the other difference first or throw the pair out.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch2 id=\"method-2-grouped-data-analysis\">Method 2: Grouped data analysis\u003C\u002Fh2>\n\u003Cp>Instead of single pairs, compare groups. Take all sales in the neighborhood over a period and split them by the feature: with pool vs. without, or GLA bands. Compare median price (or median price per square foot) between groups.\u003C\u002Fp>\n\u003Cp>This uses more data than paired sales, so it's less sensitive to one odd sale. It's also easy to show in a small table in your workfile. The weakness is that groups can differ in more than one way. If homes with pools are also newer, the "pool" difference is partly an age difference. Look at the groups' other characteristics before you trust the gap.\u003C\u002Fp>\n\u003Ch2 id=\"method-3-regression\">Method 3: Regression\u003C\u002Fh2>\n\u003Cp>Regression fits a line (or a plane, with several variables) through many sales at once. Simple regression of price on GLA gives you a dollars-per-square-foot slope. Multiple regression estimates several features at once (GLA, age, lot size, garage) while holding the others constant.\u003C\u002Fp>\n\u003Cp>This is the same family of math behind many automated valuation models, and the place where "machine learning appraisal" has real meaning. Used by an appraiser on a well-chosen data set, it's a legitimate support method. Used blindly, it's a black box.\u003C\u002Fp>\n\u003Cp>Practical rules:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Choose the data set like you choose comps.\u003C\u002Fstrong> Same market area, similar property type, a sensible time window.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Check sample size.\u003C\u002Fstrong> A regression on 12 sales with five variables won't tell you much.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Look at the residuals.\u003C\u002Fstrong> If one sale is far off the line, find out why before you trust the slope.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Sanity-check the sign and size.\u003C\u002Fstrong> If regression says a bedroom is worth negative $8,000, that's probably GLA and bedrooms fighting each other in the model, not a market fact.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Keep it in the workfile.\u003C\u002Fstrong> The output, the data set and your reasoning, so you can reproduce it.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"other-commonly-accepted-methods\">Other commonly accepted methods\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>Cost and depreciation.\u003C\u002Fstrong> Useful for features with a clear replacement cost (a new roof, a detached garage), adjusted for depreciation and market acceptance.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Sensitivity analysis.\u003C\u002Fstrong> Look at how the spread of adjusted comp prices narrows as you try different rates. The rate that best brackets the comps is a supported indication.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Market interviews.\u003C\u002Fstrong> Agents, builders and other appraisers. Useful context, weaker as primary support.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>Most good reports reconcile more than one method.\u003C\u002Fp>\n","\u003Ch2 id=\"method-4-your-own-archive\">Method 4: Your own archive\u003C\u002Fh2>\n\u003Cp>Here is the method most appraisers overlook. You already have thousands of verified sales and the adjustments you made to them, in your old software's files.\u003C\u002Fp>\n\u003Ch3 id=\"why-your-archive-is-good-data\">Why your archive is good data\u003C\u002Fh3>\n\u003Cul>\n\u003Cli>\u003Cstrong>You verified it.\u003C\u002Fstrong> These aren't scraped records. You confirmed the GLA, the condition, the sale terms.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>It's local.\u003C\u002Fstrong> It's from the towns, price tiers and property types you actually appraise.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>It shows your own consistency.\u003C\u002Fstrong> If you've used $35 per square foot for GLA in one town for three years and suddenly used $55, you should know why.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>It holds pairs.\u003C\u002Fstrong> Two of your old comps in the same subdivision that differ in one feature may be exactly the paired sale you need.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3 id=\"how-an-archive-becomes-numbers\">How an archive becomes numbers\u003C\u002Fh3>\n\u003Cp>The process looks like this, whether you do it by hand in a spreadsheet or let software do it:\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>Export.\u003C\u002Fstrong> Most desktop programs can export MISMO 2.6 GSE XML. ClickFORMS files and those exports hold the comps grid, the adjustments and the subject data. (The \u003Ca href=\"\u002Fblog\u002Fswitch-appraisal-software-without-losing-comps\">switching guide\u003C\u002Fa> covers getting them out.)\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Parse.\u003C\u002Fstrong> Pull out each comp: address, sale price, date, GLA, features, and each adjustment by type.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Clean.\u003C\u002Fstrong> Throw out or flag records that are unclear: missing GLA, odd adjustment types, duplicate sales.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Group.\u003C\u002Fstrong> By town (or neighborhood) and price tier, so a $250,000 ranch and a $900,000 colonial aren't averaged together.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Summarize.\u003C\u002Fstrong> Median rate for each adjustment type (GLA per square foot, garage per bay, bath, basement finish, condition step), with the count of reports behind it.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Cp>Medians beat averages here. One odd report with a huge location adjustment won't pull a median far.\u003C\u002Fp>\n\u003Ch3 id=\"what-a-summary-looks-like\">What a summary looks like\u003C\u002Fh3>\n\u003Cp>An illustrative example for one town:\u003C\u002Fp>\n\u003Cdiv class=\"bl-table\">\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>Adjustment\u003C\u002Fth>\u003Cth>Price tier\u003C\u002Fth>\u003Cth>Reports\u003C\u002Fth>\u003Cth>Your median\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>GLA (per sq ft)\u003C\u002Ftd>\u003Ctd>$250k to $400k\u003C\u002Ftd>\u003Ctd>63\u003C\u002Ftd>\u003Ctd>$38\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Garage (per bay)\u003C\u002Ftd>\u003Ctd>$250k to $400k\u003C\u002Ftd>\u003Ctd>41\u003C\u002Ftd>\u003Ctd>$9,500\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Full bath\u003C\u002Ftd>\u003Ctd>$250k to $400k\u003C\u002Ftd>\u003Ctd>37\u003C\u002Ftd>\u003Ctd>$7,000\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Basement finish (per sq ft)\u003C\u002Ftd>\u003Ctd>$250k to $400k\u003C\u002Ftd>\u003Ctd>22\u003C\u002Ftd>\u003Ctd>$15\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003C\u002Fdiv>\n\u003Cp>That table doesn't tell you what the market is paying today. It tells you what you concluded across many reports, which is a strong consistency check and a starting point for testing against current sales.\u003C\u002Fp>\n\u003Cblockquote class=\"bl-callout tip\">\u003Cp>\u003Cstrong>Tip:\u003C\u002Fstrong> When your current analysis lands far from your own median, that's not wrong by itself. Markets move. But it's the first thing a reviewer comparing your reports will notice, so explain it in a sentence.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Ch3 id=\"using-it-in-a-report\">Using it in a report\u003C\u002Fh3>\n\u003Cp>A sentence like this in your commentary is a lot stronger than "market derived":\u003C\u002Fp>\n\u003Cp>"The GLA adjustment of $40 per square foot is supported by grouped analysis of 18 sales in the subject's subdivision over the past 12 months and is consistent with the appraiser's median of $38 across 63 prior assignments in this town and price range."\u003C\u002Fp>\n\u003Cp>That's specific, testable and yours. It also fits inside UAD 3.6's commentary limits, which reward exactly this kind of short, data-backed sentence.\u003C\u002Fp>\n\u003Ch2 id=\"firms-one-archive-many-appraisers\">Firms: one archive, many appraisers\u003C\u002Fh2>\n\u003Cp>A firm's archive is bigger and more valuable than any one appraiser's. It also raises a question: whose medians are they? A few practical points:\u003C\u002Fp>\n\u003Cul>\n\u003Cli>\u003Cstrong>Separate by appraiser when it matters.\u003C\u002Fstrong> A trainee's early reports and a 25-year certified appraiser's reports shouldn't carry the same weight without review.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Share the pairs, not just the rates.\u003C\u002Fstrong> A paired sale found in a colleague's report is evidence you can verify yourself.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Watch the dates.\u003C\u002Fstrong> A median built on 2021 reports in a market that turned in 2022 may need a time cut.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Keep it in the workfile.\u003C\u002Fstrong> If you rely on firm data, keep the summary you used with the report. USPAP's Record Keeping Rule requires the workfile to be retained at least five years after preparation, or at least two years after final disposition of a judicial proceeding in which you testified, whichever is later.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2 id=\"where-ai-fits-and-where-it-doesn-t\">Where AI fits, and where it doesn't\u003C\u002Fh2>\n\u003Cp>Software can do the boring part: reading thousands of old files, cleaning them, grouping them and computing medians in seconds. That's "AI that learns your adjustments" in the honest sense: it learns from your own work, and it shows its counts.\u003C\u002Fp>\n\u003Cp>What it shouldn't do is choose your adjustment for you. A rate that appears in your grid with no source is exactly the "statement only recognizing that an adjustment has been made" the Selling Guide rejects, just automated. The broader argument is in \u003Ca href=\"\u002Fblog\u002Fai-appraisal-software-2026\">AI appraisal software in 2026\u003C\u002Fa>. And data-backed adjustments are one of the best ways to prevent the "support your adjustments" request, as covered in \u003Ca href=\"\u002Fblog\u002Fai-catches-amc-revision-requests\">how AI catches AMC revision requests before you deliver\u003C\u002Fa>.\u003C\u002Fp>\n\u003Cp>Carat does this with a bulk legacy import: upload a zip of ClickFORMS files or MISMO 2.6 XML, and it builds a searchable comps history and median adjustment rates by town and price tier from your own reports. Anything unclear goes to a review queue, never silently into a report. If you're coming from ClickFORMS specifically, see \u003Ca href=\"\u002Fblog\u002Fmigrate-from-clickforms-uad-3-6\">migrating from ClickFORMS\u003C\u002Fa>. It's part of the \u003Ca href=\"\u002F#pricing\">$99 monthly plan\u003C\u002Fa>, and you can \u003Ca href=\"\u002Fsign-in?mode=create\">start an account here\u003C\u002Fa>.\u003C\u002Fp>\n",[53,56,59,62,65,68,71,74],{"id":54,"t":55},"what-the-gses-expect","What the GSEs expect",{"id":57,"t":58},"method-1-paired-sales","Method 1: Paired sales",{"id":60,"t":61},"method-2-grouped-data-analysis","Method 2: Grouped data analysis",{"id":63,"t":64},"method-3-regression","Method 3: Regression",{"id":66,"t":67},"other-commonly-accepted-methods","Other commonly accepted methods",{"id":69,"t":70},"method-4-your-own-archive","Method 4: Your own archive",{"id":72,"t":73},"firms-one-archive-many-appraisers","Firms: one archive, many appraisers",{"id":75,"t":76},"where-ai-fits-and-where-it-doesn-t","Where AI fits, and where it doesn't",[],[79,118,159],{"slug":28,"title":80,"seoTitle":81,"description":82,"date":83,"updated":10,"category":11,"pillar":84,"howto":12,"keywords":85,"tags":94,"related":98,"faq":100,"words":116,"minutes":117},"AI Appraisal Software in 2026: What It Does, and What It Shouldn't","AI Appraisal Software in 2026: An Honest Guide | Carat","AI appraisal software in 2026, explained for working appraisers: what automation should handle, where AVMs fit, and why your judgment and signature stay.","2026-10-14",true,[86,87,88,21,89,90,91,92,93],"ai appraisal software","ai for appraisers","appraisal automation","automated valuation model","will ai replace appraisers","ai appraisal tools","appraisal software with ai","uad 3.6 ai",[24,95,96,97],"Automation","UAD 3.6","AVMs",[99,30,5],"will-ai-replace-appraisers",[101,104,107,110,113],{"q":102,"a":103},"What is AI appraisal software?","It is report software that uses automation and statistical models to handle clerical work for the appraiser: filling public-record data, checking the report against rules, and summarizing the appraiser's own past data. The appraiser still inspects, analyzes, concludes value and signs.",{"q":105,"a":106},"Can AI write my appraisal report?","It should not. USPAP holds the appraiser responsible for the analysis and the conclusions in the report. Software can fill facts and flag errors, but the narrative, the adjustments and the value opinion need to be yours.",{"q":108,"a":109},"Is an automated valuation model the same as an appraisal?","No. An AVM is a computer model that estimates value from data without an inspection or an appraiser's analysis. An appraisal is a credentialed appraiser's opinion of value developed under USPAP.",{"q":111,"a":112},"Does the 2024 AVM quality control rule apply to appraisers?","The interagency rule, effective October 1, 2025, applies to mortgage originators and secondary market issuers that use AVMs for credit decisions. The CFPB's compliance guide says it does not cover the development of an appraisal by a certified or licensed appraiser.",{"q":114,"a":115},"What should I look for in AI tools for appraisers?","Look for automation you can check: every auto-filled value shows its source, every warning links to the field, and nothing goes into the report without your review. Avoid tools that produce a value or commentary you cannot trace.",2684,12,{"slug":29,"title":119,"seoTitle":120,"description":121,"date":122,"updated":10,"category":123,"pillar":84,"howto":84,"keywords":124,"tags":133,"related":137,"faq":141,"words":157,"minutes":158},"How to Switch Appraisal Software Without Losing a Decade of Comps","Switch Appraisal Software, Keep Your Comps | Carat","How to switch appraisal software without losing your comps: export your MISMO 2.6 archive, protect USPAP workfiles and run both programs during UAD 3.6.","2026-10-06","switching",[125,126,127,128,129,130,131,132],"switch appraisal software","change appraisal software","appraisal software migration","export appraisal reports mismo 2.6","keep comps when switching appraisal software","uad 3.6 appraisal software switch","appraisal workfile retention","import old appraisal reports",[134,96,135,136],"Switching","MISMO 2.6","Workfiles",[138,139,140],"migrate-from-clickforms-uad-3-6","leaving-aci-total-desktop-appraisal-software","best-appraisal-software-uad-3-6-2026",[142,145,148,151,154],{"q":143,"a":144},"Will I lose my old reports if I switch appraisal software?","Not if you export them first. Most desktop appraisal programs can export finished reports as MISMO 2.6 GSE XML and PDF. Keep those exports, plus your original program's data folder, for as long as USPAP and your state require.",{"q":146,"a":147},"Can old UAD 2.6 reports be converted into UAD 3.6 reports?","Not in any way you should rely on for delivery. The redesigned URAR asks for data the old forms never collected. What you can do is import the old reports as a comps and adjustment history and start new UAD 3.6 reports fresh.",{"q":149,"a":150},"How long do I have to keep appraisal workfiles after switching software?","USPAP's Record Keeping Rule requires keeping the workfile at least five years after preparation, or at least two years after final disposition of a judicial proceeding in which you testified, whichever is later. Your state may add its own rules.",{"q":152,"a":153},"Should I cancel my old appraisal software right away?","Usually not. Keep it until your exports are verified, open revision requests on UAD 2.6 reports are done, and you know you can still open the old files. Revisions to reports first submitted in UAD 2.6 stay in UAD 2.6.",{"q":155,"a":156},"What is MISMO 2.6 GSE XML?","It is the data file format used to deliver UAD 2.6 appraisals to the GSE portals. Because it is a shared standard, it is the most portable way to move your report data out of one program and into another.",2967,13,{"slug":30,"title":160,"seoTitle":161,"description":162,"date":163,"updated":10,"category":11,"pillar":12,"howto":12,"keywords":164,"tags":171,"related":174,"faq":177,"words":190,"minutes":48},"How AI Catches AMC Revision Requests Before You Deliver","How AI Catches AMC Revision Requests | Carat","How AI checks catch AMC revision requests before you deliver: the UAD 3.6 rules behind them, a pre-delivery review, and what software can't catch.","2026-10-07",[165,166,167,88,168,169,170,87],"amc revision request","appraisal revision request","ai appraisal review","uad 3.6 compliance check","appraisal qc checklist","ucdp fatal error",[24,172,96,173],"Revisions","AMCs",[28,175,176],"amc-revision-request-survival-guide","uad-3-6-compliance-rules-fatal-warning",[178,181,184,187],{"q":179,"a":180},"Why do AMCs send so many revision requests?","Most come from three places: data errors that fail the GSE compliance rules at UCDP, inconsistencies a reviewer spots between parts of the report, and client-specific requirements or judgment questions about comps and adjustments. Software can prevent most of the first kind and some of the second.",{"q":182,"a":183},"Can AI review an appraisal before I send it?","Software can check the report against the GSE Appendix H-1 compliance rules and flag missing data, math that doesn't add up and fields that contradict each other. It cannot judge whether your comps or your value are right; that review is still yours.",{"q":185,"a":186},"What is a Fatal finding under UAD 3.6?","A Fatal finding is a compliance rule failure that stops the submission at UCDP until it is fixed. Appendix H-1 lists 728 rules for the redesigned URAR: 592 Fatal and 136 Warning.",{"q":188,"a":189},"How do I reduce revision requests from AMCs?","Run a rules check before every delivery and fix every Fatal, keep a list of each client's standing requirements, make sure commentary agrees with the grid, and answer the obvious questions in the report before a reviewer asks them.",1513,1793080428133]