The adjustment you choose is the one you can defend.
Load your comps. Nine methods run on every feature the data supports — regression, paired sales, grouped data, sensitivity — each shown separately with its reliability tier. You pick the number knowing exactly what stands behind it, and the proof exports with your workfile.
No credit card · Cancel anytime
Why This Matters
Unsupported adjustments are the #1 revision trigger.
Fannie Mae Selling Guide B4-1.3-09 is clear: “A statement only recognizing that an adjustment has been made is not acceptable.” Under UAD 3.6, adjustment support lives inline — making weak methodology automatically machine-detectable.
Your MLS comp export (CSV)
↓
ADJUST (derives the numbers)
↓
NARRATIVEAI (writes the narrative)
↓
Documents you attach, and a workfile
that shows how you got there.From MLS export to attachable support in minutes.
See every method, not a black box
Every indicator for a feature lands on one chart — bigger dots carry a higher reliability tier, and the shading darkens where the methods cluster. Up to three comp sets, each in its own colour, side by side.
You stay in control
Sliders, not verdicts. Exclude outliers by hand, override any figure, or enter an adjustment you derived yourself — named, with your reasoning, recorded in the workfile. Adjust will offer a starting figure; it never writes one.
Bad data gets caught before it costs you
Dropped digits, garage square footage entered in the space-count field, repeat sales of the same property, physically impossible indicators — flagged automatically, decided by you. Nothing is ever silently removed.
The proof travels with the report
Adjustment Support Report, per-feature comments, a methods reference for the reviewer, and a workfile ZIP with a SHA-256 manifest. When anyone asks how you got there — here is everything you did.
Import from any MLS export
Every MLS exports different columns. Adjust reads your file, maps the columns with you once, and remembers the layout for next time.
Market conditions, cross-checked
The time adjustment is derived from your own comps and then cross-checked against the FHFA county house-price index. Where they diverge, you write the note that explains it, and it lands in the workfile.
Adjust derives the numbers. NarrativeAI writes the words.
One click passes your adjustment values into NarrativeAI, which writes USPAP-compliant inline narrative commentary for every adjustment section — pre-populated with your actual numbers. Complete adjustment documentation in under 10 minutes.
One revision avoided pays for the month.
Adjustment support is the most common thing a reviewer sends a report back for. A revision costs you unpaid hours and a turn-time hit you cannot bill for.
$29/mo
or $290 a year — two months free
No per-analysis fee. No per-report fee. Run the same comps fifteen different ways and it costs the same as running them once.
How Adjust compares
| Appraveo Adjust | Bradford Redstone | DataMaster | |
|---|---|---|---|
| Platform | Web — any device | Windows only | Windows only |
| Price | $29/mo, unlimited | $9.90/analysis | $200+/mo |
| Reliability tier on every method | ✓ | — | — |
| Transparent, traceable derivation | ✓ | — | — |
| Time adjustment (dual method) | ✓ | — | — |
| Narrative output | ✓ (via NarrativeAI) | — | — |
| Paired sales | ✓ | ✓ | Limited |
| Regression | ✓ | ✓ | ✓ |
| Report-ready documents | ✓ | ✓ | ✓ |
$29 a month, on its own
Buy Adjust by itself if adjustments are what you need fixed. Add CRM + Sites or NarrativeAI later, or take all three for $89/mo. Start with a 14-day free trial of everything.
Frequently asked questions
I used Synapse. Is this like that?
Yes, deliberately. The reconciliation view, the interaction model, and the control philosophy will feel familiar. Load a CSV and you will be at home in minutes.
Does it work with my MLS export?
Every MLS exports different columns. Adjust reads your file, maps the columns with you once, and remembers the layout for next time. It needs a sale price, a living area, and a sale or contract date — everything else is optional.
Is it UAD 3.6-ready?
Yes. Exhibit captions are cut to the UAD 3.6 100-character field, the workfile ships a submittable image set with its manifest, and compliance checks run against whichever rail your session is on. The classic PDF path stays valid for 2.6, FHA, VA and USDA.
Do you keep my data?
It is yours. Exportable every day, never sold, never shared, and our AI never trains on it.
How many comps do I need?
Fifteen sales gets you a basic analysis, twenty or more makes the regressions reliable, and thirty or more gives feature-level confidence. Below that Adjust still runs paired sales and tells you plainly that the sample is thin rather than quietly reporting a weak number as a strong one.
Does Adjust replace my professional judgment?
No. Adjust derives market-supported ranges from your data and shows what each method found. You decide the number. The tool documents how you got there — it does not make the decision for you.