Why Two STR Revenue Tools Can Disagree on the Same Address
Run the same address through two different revenue estimation approaches and it's entirely possible to get two noticeably different numbers back. That's not evidence that one tool is broken — it's a predictable result of two models making different assumptions about what counts as a comparable property and how to weigh the data.
NightYield Editorial
STR-DSCR research & underwriting desk
Published 2026-07-19
The three usual sources of disagreement
- Different comp sets: one model's radius, bedroom filter, or amenity match might include or exclude different properties than another's, even for the exact same subject address. See /learn/what-comp-set-means-str-revenue-projection/ for how much this single variable can move a result.
- Different data windows: one estimate might weight the most recent 12 months heavily, while another averages over a longer period that includes older, less representative data.
- Different treatment of seasonality: models can differ in how they smooth or weight peak versus off-season months, which changes the annual total even from an identical underlying comp set.
None of these differences means either source is doing something wrong. They're built on different methodologies with different defaults, and reasonable methodological choices can legitimately produce different answers for the same property.
A less obvious fourth source of disagreement is simply how each tool defines the property itself. If one source has the bedroom count, amenity list, or maximum occupancy of the subject property listed slightly differently than another, that alone changes which comps get pulled in and how they get weighted — a data quality issue rather than a methodology difference, but one that produces the exact same symptom of two numbers not matching.
There's also a timing element worth noting separately from data windows generally: two sources pulled even a few weeks apart can reflect a market that's genuinely moved in that short span, particularly heading into or out of a peak season. A gap between two estimates isn't always about differing methodology at all — sometimes it's simply that the underlying market shifted between the two dates the estimates were generated.
Why lenders don't just pick whichever number is higher
A common approach is to compare the two figures, investigate the source of any material gap, and lean toward the more conservative number — or require a third source (like actual trailing operating history, once it exists) to break the tie. The practical effect is that shopping around for the single most optimistic estimate rarely helps, since it isn't the number that ultimately gets used.
This reconciliation instinct exists for a straightforward reason: if a lender simply accepted whichever of two sources produced the highest number, that would create an obvious incentive for borrowers (or anyone advising them) to keep requesting new estimates until a favorably high one appeared. Reconciling toward the more conservative figure, or requiring agreement across sources, removes that incentive and keeps the qualifying number closer to what the property is actually likely to produce.
How to use disagreement productively instead of picking a favorite
When you see two estimates diverge, the useful question isn't "which one do I get to use" — it's "why do they diverge, and which assumption is more defensible for this specific property." If one estimate uses a noticeably tighter, better-matched comp set than the other, that's a reason to lean toward it regardless of which number is higher or lower.
It's also worth checking whether the two sources actually agree on the property's basic facts — bedroom count, amenities, maximum occupancy — before assuming the disagreement is purely methodological. Sometimes the fastest way to resolve a gap between two revenue estimates is simply confirming that both tools were working from the same accurate description of the property in the first place.
If a gap between two sources persists even after checking the property's listed details and the comp sets look reasonably similar, it's worth simply accepting that some amount of disagreement is a normal feature of estimating variable income, rather than searching indefinitely for a single tiebreaker. At that point, the most productive move is usually to plan conservatively around the lower of the two figures rather than continuing to seek out a third or fourth opinion in hopes of a more favorable number.
Key takeaways
- Disagreement between two revenue estimates is normal and usually traces to comp set, data window, or seasonality treatment differences.
- A mismatch in the property's basic listed details (bedrooms, amenities) can also silently cause a gap between two tools.
- Timing differences between when two estimates were generated can also explain part of a gap, separate from methodology.
- Neither estimate disagreeing is automatically "wrong" — they can both be reasonable given different assumptions.
- Lenders typically reconcile toward the more conservative figure rather than accepting the higher one.
- Investigate the source of a gap (comp quality, recency, seasonality handling, or basic property data) rather than simply picking the number you prefer.