Ask two beginners to value the same house and you’ll often get two very different numbers. Ask them how they got there and you’ll usually find the same problem: they picked comps they liked instead of comps that were actually comparable.
Comps aren’t a vibe. They’re evidence. And like any evidence, they need rules.
I’m Ben Lovro. I buy houses directly from sellers in the South Carolina Midlands, and these are the comp rules my team and I run on every property before an offer goes out. They aren’t fancy. They’re strict. That’s the point.
Why comps matter so much
Every investing formula starts with value. Flip offers start with ARV. Wholesale numbers start with the end buyer’s ARV. Rental refinances depend on appraised value. Even as-is offers depend on knowing what similar unrenovated houses sell for.
Comparable sales — comps — are how you prove value. Get them wrong and every downstream number inherits the error. That’s why the deal analysis sequence starts here.
Rule 1: sold in the last three to six months
Only sold properties count. Actives show hope. Pendings show an agreement that might not close. Solds show what a buyer actually paid.
And recent. I start at three months and stretch to six only if I need to. Markets move, and a sale from a different season or rate environment can mislead you.
Rule 2: within half a mile to one mile
Proximity matters because location is the one thing about a house you can’t renovate. In a dense neighborhood, I try to stay within half a mile. In a rural or suburban area with fewer sales, I’ll stretch toward a mile. Beyond that, I’m skeptical.
Rule 3: same school zone
Families shop by school zone. Two houses a few streets apart can sell very differently if a school boundary runs between them. Check the zones — don’t assume.
Rule 4: no highway or railroad between
Physical barriers often split neighborhoods into different markets. An interstate, a major highway, a rail line, sometimes even a big commercial corridor or a river. If something like that sits between your subject and your comp, the comp probably belongs to a different market. Toss it.
Rule 5: same bed and bath count
A three-bedroom, one-bath house and a three-bedroom, two-bath house are different products to buyers. So are a two-bed and a three-bed. Match beds and baths. If you can’t, you’ll have to adjust — and that’s where errors creep in.
Rule 6: within ±20% square footage
A much bigger or much smaller house isn’t comparable even if everything else matches. I keep comps within ±20% of the subject’s heated square footage. Inside that band, you can adjust. Outside it, you’re guessing.
Rule 7: similar age, style and stories
A 1950s brick ranch, a 1990s two-story vinyl home and a new-construction craftsman can sit on the same street and attract different buyers at different prices. Match the era, the architectural style and the number of stories as closely as you can.
Rule 8: three comps minimum
One comp is an anecdote. Two is a coincidence. Three starts to be a pattern. If you can’t find three that meet these rules, widen carefully — a little more time, a little more distance — and note that the value is less certain. Then underwrite more conservatively.
Here’s the order I pull them in
- Identify the subject: beds, baths, square footage, year built, style, stories, lot, garage, condition.
- Confirm the subject’s school zone and note any nearby barriers (highways, rail lines, major roads).
- Search sold properties in the last three months within half a mile.
- Filter by beds, baths, ±20% square footage, era and style.
- If you have fewer than three, expand to six months, then toward one mile — never across a barrier or school line.
- Split results into renovated and unrenovated.
- Open every listing. Study the photos, days on market, price changes and any concessions.
- Throw out anything that doesn’t truly compare. Keep the best.
Renovated versus as-is comps
Separate your comps by condition. This is one of the most useful habits you can build.
- Renovated comps tell you the after repair value — what the house could sell for after work. That’s what you need for flips. The full process is on how to calculate ARV.
- As-is or dated comps tell you roughly what the house is worth today in its current condition. That helps you understand the seller’s alternatives and sanity-check your offer.
Mixing the two into one average gives you a number that describes no real house.
Reading a comp, not just listing it
For each comp you keep, write down:
- What finish level it had (from listing photos)
- How long it sat on the market
- Whether the price dropped before it sold
- Whether the seller paid concessions
- Anything unusual about the lot or location
- Any feature your subject lacks, or vice versa
This is the detail that lets you adjust intelligently and spot the comp that looks good on paper but isn’t.
Where to get comps
- A licensed agent with MLS access. Often the richest data, especially photos, days on market and concessions. A good agent who knows renovated product in that area is gold. If you’re a licensed agent with investor-grade listings, I’d like to know you — see the agents site.
- County records. Useful for confirming sale dates, recorded prices and property characteristics.
- Public sold data online. Convenient, but cross-check — details are sometimes missing or wrong.
No single source is perfect. Cross-check, and when something looks off, dig.
Comps in the Midlands
Here’s what this looks like in practice where I work. In-town Columbia has older character neighborhoods — Shandon, Rosewood, Earlewood, Melrose Heights — where the housing stock can vary a lot within a few blocks, so tight comps matter. Across the Congaree in West Columbia and Cayce, you’ll see a lot of smaller, older homes on small lots. In Irmo, many neighborhoods were built from the 1970s through the 1990s, which makes era-matching easier but school zones critical. If you invest locally, read flipping houses in Columbia SC and flipping houses in Irmo SC for more context on those markets.
What comes after comps
Comps feed straight into the rest of your analysis:
- Comps → ARV (how to calculate ARV)
- ARV + repairs → MAO (maximum allowable offer formula)
- Then a risk check before you sign (red flags)
If you’re flipping, the 70 percent rule page shows how your comp-based ARV turns into an offer.
Your next step
Choose one house in your area that’s for sale right now. Run all eight rules and find three comps. Write down why you kept each one and why you threw others out. Then compare your number to what the house eventually sells for. Repeat until your comps start predicting the market. If you want the one-page analyzer I use to keep this consistent, it’s in the free investor playbook.
Questions people ask
How many comps do I need?
At least three good sold comps. More is better if they’re truly comparable. If you can’t find three that meet your rules, that’s information too — the value is less certain, and you should underwrite more conservatively.
How far back should comps go?
I start with sales from the last three months and stretch to six if I need more. Older sales can reflect a different market. If you have to go further back, adjust for market movement and treat the result as less reliable.
Where do investors get comps?
Common sources include a licensed agent with MLS access, county property records and public sold data online. Each has gaps, so cross-check. A trusted agent who sells in that neighborhood is often the best source of context behind the numbers.