An Honest Look at UAD 3.6 from a Working Appraisal Shop

An Honest Look at UAD 3.6 from a Working Appraisal Shop

We’re a smaller Texas appraisal firm. For years we’ve run residential work in a la mode TOTAL on UAD 2.6. With the move to 3.6, we’ve been test-driving Aivre, one of the software platforms built from the ground up for the new URAR. I want to share what we’ve found, partly about the software itself and partly about the reporting format it’s built to serve. Frankly, the second issue concerns me more than the first.

Credit first where it’s due. Aivre feels like it was designed for 3.6 rather than retrofitted to it. Dwellings, units, ADUs and outbuildings are treated as separate structures, and the counts calculate on their own. The sales grid is a genuinely good tool, with side-by-side zoom and keyboard shortcuts that copy a cell to one comp or all of them. It also has an auto-adjust feature where you set a dollar or percentage rate with a threshold and it applies across the grid. You can upload an MLS CSV export and it will build the Market section graphs, a market-area map, absorption figures and even a time adjustment. Compliance checks with Fannie and Freddie run inside the program before it will let you export the delivery file. Their published guide mapping every GSE photo requirement to where it goes in the software is excellent, whatever platform you use. The company also ships updates roughly monthly and is clearly listening to users.

That said, the AI does not do nearly as much as I expected or as much as the marketing implies. It will pull public record and MLS data and read an engagement letter and sales contract. It will try to place photos and fill empty fields. But everything it touches still has to be checked line by line, and there’s a lot it simply doesn’t carry. The appraiser is still the one keying, verifying, correcting and explaining. When the verification time is added back in, the savings shrink considerably. A tool that fills 300 fields I then have to audit one by one is not the same as a tool that saves me time.

The software also has some rough edges that come with being young. The help center is thin, and most section articles are the same generic checklist with the section name swapped in. The real instruction lives in their YouTube series and release notes. Because the product is changing so quickly, some updates have told users to delete and re-enter data they had already put in. Only one person can have an order open at a time, and the grid page bogs down in the browser once five comps and their photos are loaded. The finish is rigid. You have to sign before the compliance check will run, editing after signing clears the signatures, and un-signing takes your PIN. A late correction turns into a cycle of sign, check, un-sign, fix, re-sign and re-check. A brand-new order showed us 25 validation flags before we’d worked most of the sections. Most were legitimate 3.6 requirements rather than software bugs, but it sets the tone.

My bigger frustration is with UAD 3.6 itself. The new format asks for an enormous amount of granular detail that has few, if any, measurable metrics in the market. We are now describing properties room by room and level by level. We rate condition and quality at the exterior, interior and overall levels, and itemize features, disaster mitigation, energy efficiency, amenities and outbuildings. Every defect is logged with a recommended action. Every difference between the subject and a comp needs an entry on the grid, even if that entry is a zero because the market doesn’t measure it. The MLS doesn’t track most of these items, buyers don’t price most of them in any way we can isolate, and there are no paired sales to support adjustments for most of them. We are spending real time documenting characteristics we can’t measure, which adds length and liability to the report without adding much to the credibility of the value.

To put it in practical terms, here is our current 2.6 workflow in TOTAL. Prepping a report takes about five minutes. The site visit takes 15 to 20 minutes using the TOTAL mobile app. Analyzing and completing the report on the back end takes about 45 minutes. That’s roughly an hour and ten minutes of appraiser time per typical file, and it’s a process we’ve refined over years. Under 3.6, even with an AI-assisted platform, our files are running three to five hours depending on complexity, and that doesn’t include windshield time. That’s roughly three to four times as long for the same assignment. The inspection alone has to capture far more, and the back end has more fields, more photos placed in specific sections, more validation to clear, and a heavier finish before delivery.

The industry has been told 3.6 and the new tools will make appraisers more efficient. Based on what we’ve seen, I don’t see how. The reporting standard asks for more data than the market can support. The software providers are still building, fixing and changing their products month to month. Implementation is landing on appraisers who are expected to keep turn times and fees where they are. Efficiency comes from mature, reliable tools and from requirements that line up with what the market actually measures, and right now we have neither.

That brings me to my peers. Please take some real time and think about your plan for what comes next. There is no way to flip a switch across an entire industry and expect it to absorb more detail, more work, more liability and more underwriting conditions without consequences. And let’s be honest: there is no reason to believe underwriting is about to get easier on our end of the process. More data points mean more things to question. All of this is landing on fee structures that, in 2026, are already at least two decades out of date. By the government’s own CPI numbers, $100 in 2006 takes about $166 to match today, so a fee that hasn’t moved in twenty years is a 40% pay cut. The purchasing power of the dollar has never been lower, and our cost of doing business keeps climbing. More effort, more time and more cost justify fair compensation for the work we do as professionals.

So I’ll say it plainly. If you are not yet competent with the new workflow, please do not take 3.6 orders until you are. Competency isn’t optional, and a rushed 3.6 report hurts you, your client and the borrower. And when you do take them, price them for what they actually cost you to produce. A 3.6 report is not a 2.6 report with a new form number. Accepting the same fee you charged for 2.6 work, or a discount on it, tells the market that all this added work is free. Every appraiser sets their own fees, but know your numbers before you say yes.

Appraisers are already treated as the clog in the wheel of the lending process. We’re the step everyone else waits on, the line item they’d most like to shrink, and the easiest place to point a finger when a closing slips. So it’s fair to ask a simple question: what are we actually trying to achieve with this change? UAD 3.6 has been sold as appraisal “modernization.” Modernization should mean faster, clearer, more reliable valuations that serve the borrower and the lender better. What I’m seeing points the other way, especially in the first few years: longer reports, longer turn times, more conditions, more revisions, and appraisers learning on the fly with software that is still being built. That isn’t modernization. It’s more weight on the part of the process that was already called the bottleneck.

I believe there needs to be a real midpoint meeting between the GSEs and the professionals who rely on this work to make a living. Not another webinar or FAQ update, but a genuine back-and-forth about which data actually improves credibility and risk assessment, and which is collected simply because the form now has a field for it. The people writing the requirements and the people producing the reports need to be in the same conversation before this hardens into the permanent standard.

To my fellow appraisers: do not let yourself become a commodity for the lending industry. The licenses we carry are too damn hard to earn to be reduced to anything less than professional treatment. The education, experience hours, exams, supervision, continuing education, E&O and liability we carry are not the profile of a data-entry vendor. This change is the biggest catalyst our profession has seen in decades, and how we respond to it will shape how we’re treated for the next decade. Let’s all do our part to learn the new standard, adapt, and deliver quality work. But let’s not sacrifice our time, talent and treasure to do it.

I also invite the GSEs, regulators and everyone else shaping this rollout to step back and ask how this is supposed to sort itself out without disrupting an entire profession, and the deals that depend on it, overnight. Look at where we’re starting from. UAD 3.6 becomes mandatory for new GSE submissions on November 2, 2026, and the old 1004, 1073, 2055 and related forms are retired that same day. Then, on September 30, barely a month before the deadline, Fannie and Freddie announced a one-time exception. It lets lenders who aren’t ready keep submitting UAD 2.6 appraisals through May 19, 2027. FHA has made 3.6 optional, set no mandatory date, and won’t allow formats to be mixed on the same case number. In plain English, for at least the next seven months appraisers will be producing two different report formats for different lenders and loan programs, with different rules, on software that is still changing month to month. That isn’t a transition plan. It’s a patchwork.

Now look at who is supposed to absorb it. The number of mortgage appraisers has fallen from about 55,000 in 2015 to about 37,000 in 2025, with roughly 10,000 more projected to leave by 2030. More than half of active appraisers are over 60, and it takes 1,000 to 3,000 supervised hours to replace one. Some of them are already saying they won’t do 3.6 work at all. We’ve seen this movie before. During the 2022 capacity crunch, turn times stretched from days into weeks, fees spiked, and lenders said it limited how many loans they could close. Today, with a format every appraiser knows cold, roughly one in eight contracts already has a delayed settlement according to NAR’s monthly survey, and appraisal issues are a standing cause. Now take that same shrinking workforce and add a report that takes us three to four times as long, plus a second format running alongside it and a fresh wave of 3.6 underwriting conditions. Ask yourself what happens to turn times the first time rates drop and volume comes back.

I expect the rest of 2026 and all of 2027 to be a “shit-show” for lenders and Realtors trying to close deals if the 3.6 transition continues the way it’s being handled now. Buyers and sellers already face enough hurdles in a 2.6 world. The industry’s own response tells you something: lenders are openly leaning on appraisal waivers and automated underwriting as a hedge against a 3.6 crunch. When the answer to a broken rollout is “skip the appraiser,” that should worry everyone who cares about collateral risk. The people who will feel the result most are not the AMCs or the lenders. They are the consumers: homeowners and buyers who depend on a timely, credible appraisal they can afford. They deserve a better outcome than the one I believe is being teed up for them right now.

(We have no affiliation with Aivre. These are our experiences so far.)

Sources

•         UAD 3.6 dates and retired forms: McKissock

•         GSE 2.6 exception into 2027: HousingWire

•         FHA optional 3.6: HousingWire

•         Appraiser workforce, the 2022 crunch, and lenders hedging with waivers: National Mortgage News

•         Delayed settlements: NAR Realtors Confidence Index, March 2026

•         Inflation since 2006: CPI calculator (BLS data)