.jpeg)
You've been told your entire career: use comps from the last 3-6 months. Stay recent. Don't go back too far.
But what happens when there are only two sales in your neighborhood in the last six months, and neither one is remotely similar to your listing?
You either pick mediocre comps from the recent timeframe, or you ignore a nearly identical home that sold nine months ago because it's "too old."
Here's the thing: if you know how to apply time adjustments correctly (which we covered last week), you can expand your comp pool to a full 12 months without sacrificing accuracy. But before we get to time adjustments, you need to know how to pick good comps in the first place.
Let's start with the foundation.
The main principle in comp selection: think like the buyer.
If the property you're listing - the subject property - were to show up in your buyer's home search, what other properties would likely appear alongside it?
Let's say you're listing a 3,500 square foot, five-bedroom, four-bathroom, two-story home with a three-car garage. What kind of search criteria would you have set up for a buyer for that home to show up as a possibility?
That's the exact mindset you need when selecting comps.
It's common to hear "you can only go a mile in distance" when selecting comps. But I'd encourage you to think less about the mile and more about the locations that the typical buyer of a property in that area would actually consider.
Sometimes this means the buyer would look slightly more than a mile north or south. But it might mean they'd only look a quarter mile east or west because of certain boundaries - maybe a highway, a school district line, or a clear shift in neighborhood character.
Rather than drawing a one-mile radius on your comp map, I recommend drawing a polygon where you outline the boundaries of the market area that the typical buyer of that property would look within.
This is your search area. It should reflect buyer behavior, not arbitrary distance rules.
Now you need parameters to identify similar properties without getting too narrow.
Looking at that 3,500 square foot example, you wouldn't want to set your search at exactly those criteria - you'd get almost no results.
Instead, bracket each factor a bit:
The goal is to cast a net that's wide enough to get results but narrow enough to stay comparable.
How Many Results Should You Get?
This matters more than you might think.
If you run your search and get four results, that's like sending four listings to a buyer - they're going to ask you to expand the search so they have more options.
If you get 200 results, that's overwhelming - like sending a buyer 200 listings. They'll ask you to narrow it down.
In a good market, aim for 50-100 results. In today's limited market, 25-50 is reasonable.
If you have to choose between too narrow and too broad, go slightly too narrow. Those properties will at least be very similar, even if there aren't many of them. Going too broad can bring in properties that a typical buyer wouldn't consider - and that an appraiser won't consider when your listing goes under contract.
From your search results, you want to select:
Why under contract/pending over active? Because active listings can give a false picture of value. A home can be listed at any price - it doesn't mean it's worth that. Under contract and pending listings show you what buyers are actually willing to pay.
Pick comps that bracket the subject property in all the main categories: square footage, age, lot size, garage count, quality, and condition.
"Bracket" means some comps should be inferior in each category, and some should be superior.
For example:
If you find comps that are all the same quality or condition, that's actually ideal - but make sure you're not picking only comps that are on one side of the subject.
Here's why this matters: If you pick all superior comps - say, all in better condition than your listing - it's hard to validate your adjustments. Are your condition adjustments correct, or do they need to be larger or smaller? Without comps on both sides, you can't tell.
The perfect scenario would be finding comps all the same square footage, age, and condition. But in reality, you'll need to pick some that are slightly bigger and some slightly smaller, some newer and some older, some in similar condition and some slightly different.
When I'm selecting comps, I work in layers:
Layer 1: Nearby + Recent + Very SimilarStart with homes that sold in the neighborhood recently and are very similar in design, size, and condition. If your subject is a 3,500 sq ft five-bed, four-bath two-story, look for other two-stories of similar size and condition that sold nearby in the last few months.
Layer 2: Expand Distance OR Time + Similar HomesNext, expand either your distance or your time to find similar homes. Maybe a nearly identical home that sold four months ago, or a very similar home in an adjacent neighborhood.
Layer 3: Nearby + Recent + Overall SimilarGo back to your recent, nearby sales and look for homes that are overall similar, even if they weren't exactly the same. Maybe a ranch instead of a two-story, but similar size and quality.
Layer 4: Similar Homes in Nearby Neighborhoods + Further BackFinally, look for similar homes in adjacent neighborhoods that sold further back in time.
This layering ensures you're prioritizing the most relevant comps while still having enough to bracket the subject properly.
Here's where last week's trendline analysis becomes critical.
What if you find a home with the exact same floor plan, in the same neighborhood, in very similar condition, with very similar amenities - but it sold nine months ago?
Without time adjustments, most agents would skip it because it's "too old." But this would be a fantastic comp as long as you can factor in the market price change since it sold.
This is where your 12-month, 9-month, 6-month, and 3-month trendline data comes in. You can adjust that nine-month-old sale to today's market and use what might be your single best comparable.
By properly using time adjustments, you can:
Once you've applied time adjustments to your sold comps (but before adjusting for size, condition, etc.), your price range should:
If your time-adjusted sold comps don't align with what's currently going under contract, your time adjustment needs recalibration. Go back to your trendline analysis.
This alignment is your reality check that you've correctly identified the market movement.
Good comp selection isn't about following arbitrary rules like "stay within one mile" or "only use sales from the last six months."
It's about thinking like the buyer who would purchase your listing.
Define your market area based on buyer behavior. Bracket your search parameters to get 25-50 results in today's market. Select 3-5 sold comps that bracket your subject in the key categories, prioritizing nearby, recent, and similar - but willing to expand distance or time to find the best matches.
And then - this is key - use proper time adjustments to unlock access to highly similar comps from further back in the 12-month window.
That nine-month-old sale with the same floor plan? It might be your best comp. Don't ignore it just because it's not from the last three months.
