Comparative Market Analysis: How to Price a Home Like an Appraiser

Your next seller has already priced their home. They typed the address into Zillow, asked an AI assistant what it’s worth, and walked into the listing appointment sure they know the number. Your job now is to guide a confident client to the real one.

A comparative market analysis (CMA) is an estimate of a home’s likely market value, usually prepared by a real estate agent, built by comparing the property to similar homes nearby that recently sold, are under contract, or are currently listed. The agent adjusts each comparable for differences in size, condition, location, and features, then reconciles them into a price range. It isn’t a legally binding appraisal.

This is where the CMA earns its keep. When clients self-educate before they ever call you, an automated estimate has already handed them a number with no reasoning behind it. A CMA shows the work: real comps, documented adjustments, and a range you can defend line by line. That evidence is what quietly re-establishes you as the expert in the room, and the algorithm as a starting point rather than the authority.

This guide walks through the discipline that makes a CMA defensible: the definition, how it differs from an appraisal and an AVM, the five-step process, a worked example with the math shown, and the judgment calls (comp counts, 2026 concessions, hard-to-comp homes, and the buyer side).

What is a comparative market analysis?

Ask ten sellers what a CMA is and most will describe a Zillow-style number that appears instantly and settles the question. The real thing is slower and far more defensible.

CMA stands for comparative market analysis. It’s a data-driven opinion of a home’s market value, built by an agent from recently sold, active, and pending comparables, then adjusted for the differences between those homes and the subject property.

It carries two limits worth stating up front. It isn’t legally binding, and lenders won’t accept it for a mortgage approval. That’s the appraiser’s job.

Who can actually produce one? Typically a licensed agent or broker. And there’s a real standard behind it that most sellers never hear about. Under the NAR Code of Ethics, Standard of Practice 11-1, a REALTOR’s opinion of value has to state the basis for that opinion including the market data used, state plainly that it isn’t an appraisal, and disclose whether a physical inspection of the property happened. That disclosure requirement is part of what separates a professional CMA from an algorithm’s guess.

The reframe that matters most: a CMA is a pricing and negotiating tool, not a promise the market will agree with you. You can build a flawless analysis and still meet a seller who has a firm number in mind for reasons that have nothing to do with comps. The analysis gives you ground to stand on. It doesn’t guarantee the other side moves.

A CMA, then, is a disclosed, defensible opinion of value, and its quality lives or dies on the agent’s method.

CMA vs. appraisal vs. Zestimate: what is the difference?

Zillow reports its Zestimate is off by a median of 1.83% on homes that are actively listed. On off-market homes, that error jumps to 7.01%. And accuracy isn’t uniform across the country. In Cleveland, independent trackers found only 35.9% of Zestimates landed within 5% of the true value. Same algorithm, wildly different reliability depending on where you’re standing.

That’s the core reason a CMA, an appraisal, and an AVM aren’t interchangeable, even though sellers often treat them as the same thing. Here’s how they line up.

Real agents and consumers are blunt about where the AVM fits. In one forum thread comparing a Zestimate to a proper comp analysis, users called the automated estimates “trash” and “useless if trying to price a home for sale.” The fairer read: an AVM is a starting point, not an answer. It can’t see a renovated kitchen, a busy road, or a lake view.

For your listing conversations, the takeaway is simpler. When a seller waves a Zestimate at you, skip the argument about Zillow itself. Explain that an on-market Zestimate and an off-market one carry very different error rates, and that no algorithm adjusts for the finished basement next door that never got logged correctly. The CMA is where that local detail lives.

So: reach for an AVM when you want an instant, rough directional read. A CMA is the tool when you’re setting a listing price or shaping an offer and need something you can defend. Save a formal appraisal for when the stakes are legal, a lender is involved, or comps are too thin for a CMA to hold up.

How to do a CMA in 5 steps

Give me a blank MLS search and an hour, and these five steps turn it into a price range you can defend line by line. Run them in order every time.

Step 1: Profile the subject property

Start with the home you’re pricing. Pull everything an appraiser would want: beds and baths, gross living area in square feet, lot size, condition, upgrades, view, and exact location. The Close works from a detailed subject checklist for this, and the principle behind it is simple. Every data point you skip here is an adjustment you can’t make later. Walk the property if you can, because condition is the variable an algorithm never sees. A quick photo log of each room saves you a second trip when a seller questions your condition call three weeks in.

Step 2: Pull the comps

Lead with recently sold homes, then add active and pending listings for direction on where the market’s heading. A common default is a one-mile radius and the last three to six months, though that window flexes with your market (more on that below). Pull these from the MLS directly. As Phoenix agent Allen Studebaker points out, not every sale syndicates cleanly to Zillow or Redfin, so agent MLS data beats what a consumer can find on a public site.

Step 3: Adjust each comp for differences

Now make the comps look like your subject. You adjust the comp, not the subject: a comp that’s superior to your listing gets adjusted down, an inferior one gets adjusted up. The two adjustments that move the number most are time (market conditions since the sale closed) and square footage. I’ll show the math for both in the next section, so don’t worry about the dollar figures yet. Just know that every adjustment needs a reason you can point to.

Step 4: Reconcile to a price range, not a single number

Once each comp is adjusted, you weight them. The most similar and most recent sales carry the most weight. A six-month-old comp two miles away carries less. What comes out is a range, not a single magic figure. Pricing to a range is honest about the uncertainty in any comp set, and it gives you and your client room to position within the market instead of betting the listing on one number.

Step 5: Build and present the report

Structure the report so it tells a story: market overview, subject summary, the comp breakdown, active and pending context, and your recommended range. Agent Tracy Tutor frames presentation as the moment raw numbers become a pricing argument. This is also where your NAR 11-1 disclosures live: state your basis, note that it isn’t an appraisal, and disclose your inspection. Then comes the harder skill, which is walking a seller through that range and winning the listing. That’s its own craft, covered in how to present your CMA and win the listing. This page owns the analysis. That one owns the room.

If you’re new, run all five every time until the sequence is muscle memory. Speed comes from repetition, not shortcuts.

A comparative market analysis worked example: from comps to a price range

Here’s how to defend every adjustment dollar when a seller or a competing agent challenges it. The method appraisers use works just as well for a CMA.

Derive an adjustment with paired sales

Competitors love to assert that a bedroom is worth $5,000 or a basement $3,000, but they never show where the number comes from. Paired sales is how you derive it. Find two sales that match in nearly every respect except the one feature you’re pricing. Say one home sold for $370,000 with a finished basement, and a near-twin two streets over sold for $362,000 without one. That $8,000 gap approximates the basement’s value. Find a second and third pair and average them before you trust the figure, and keep a documented workfile (addresses, sale dates, sources, and your math) so the adjustment survives a challenge. This is the appraiser discipline Mike Berg brings to every listing CMA.

Age each comp with a time adjustment

Markets move while comps sit still, so the time adjustment usually matters most. Take the trailing twelve-month appreciation rate and divide by twelve. If your market rose 6% over the past year, that’s 0.5% a month. A comp that closed four months ago gets aged up about 2%. Sanity-check the result against current active listings before you lock it in.

Build the grid and reconcile to a range

Now apply everything across a set of comps. The subject here is a 4-bed, 2.5-bath, 2,000-square-foot home with no finished basement, in a market appreciating 6% a year.

The four adjusted prices cluster between roughly $387,500 and $392,700. Weight Comps 2 and 4 most, since they closed most recently, and you land on a recommended range of about $388,000 to $393,000. One caution: if the total adjustments on any single comp climb past roughly 15% to 25% of its sale price, many agents treat that as a sign the comp is too different to trust. Drop it and find a closer match.

A documented workfile is what separates a CMA from a guess. When a seller or a rival agent pushes back, you can walk them through the arithmetic line by line rather than defend an opinion.

How many comps does a CMA need, and how recent?

Ask five agents how many comps a CMA needs and you’ll get five answers. They’re all sort of right.

The guidance really does scatter across sources:

• At least 3 closed comps as a bare minimum.
• 4 to 6, which Luxury Presence calls the single biggest driver of CMA accuracy.
• 5 to 10 for a thorough analysis.
• 9 to 12 total under HomeLight’s “Rule of Threes,” meaning 3 to 5 each of active, pending, and sold.

You’ll also see a “3 to 6 comps per NAR guidelines” claim floating around. Treat that as commonly repeated rather than official. There’s no NAR rule specifying a comp count.

What actually sets the right number is market velocity, not a magic figure. In a fast market with heavy turnover, you want fewer comps but very recent ones, sometimes just the last one to three months, because prices move quickly. In a slow or rural market, you widen the window to six or even twelve months to gather enough real sales. True comparability beats hitting a count. Three tight, recent, genuinely similar sales tell you more than ten loose ones stretched across a year and three neighborhoods.

Let the market set both the count and the recency window. Start with the closest three to five sold comps, then add or widen only until the picture stops changing. When you have to choose, recency usually beats radius, so trade a little distance for a fresher sale.

Why 2026 sale prices can throw off your comps

Nearly half of recent sales carried a seller concession, which means the comp price you’re trusting may be several thousand dollars too high.

About 44.4% of U.S. home sales included seller concessions in the first quarter of 2025, and 39% of sellers surveyed in April 2026 expected to offer them. When concessions are that common, the recorded sale price stops telling you the true net deal.

Here’s the mechanism. A 2-1 rate buydown on a $290,000 purchase costs the seller roughly $6,500. The buyer gets a lower payment for the first two years, but the price on record stays at full ask. Pull that sale as a comp and it looks $6,500 stronger than the deal really was. Do that across three or four concession-heavy comps and you’ve quietly inflated your whole range.

The fix is the same discipline appraisers are required to apply. Pull the full transaction detail, not just the headline price, because concessions don’t always show in a basic MLS export. Flag or back out the concession value to approximate the true net price, and weight concession-heavy comps less than clean cash-to-seller deals.

The tell is usually in the closing details rather than the price: a seller credit toward a rate buydown, points, or closing costs. If a comp closed at full ask in a market where nothing else did, assume a concession until you confirm otherwise.

This matters most in a concession-heavy market, which describes most of the country right now. You can ease off where cash deals dominate and concessions are rare, but check before you assume that’s your market.

Pricing a home when there are no clean comps

The listing that keeps you up at night is the one with no real comps. Rural acreage, new construction, a one-of-a-kind luxury home, a condo with a pending assessment. “Just order an appraisal” isn’t a method. Here’s one.

The general moves are the same across these cases. Widen the radius (up to about five miles in rural areas versus roughly one mile in a dense suburb), extend the recency window, and pull four categories of comps instead of one: direct sales, functional substitutes, land-only sales, and older sales carried forward with a time adjustment.

For rural land, stop pricing by a flat per-acre rate. Break value into components: the house, the land priced by usability tier, site improvements, and location. One illustrative approach values usable acres at $15,000 an acre for the first five acres, then less for each tier above that. Treat that as one methodology, not an industry formula, and build your tiers from local land sales.

For new construction, lean on the builder’s base price plus options and on recent closings in the same or a neighboring development, since finishes and lot premiums vary too much to eyeball.

For condos, prioritize comps with similar HOA dues and assessment history. A pending special assessment can cool buyer appetite even when the unit itself is a near-perfect match on size and condition.

For luxury or genuinely unique homes, comps thin out fast. Recommend a pre-listing appraisal and be transparent with your client that CMA precision is lower here.

When clean comps drop below about four or five, widen your search methodically and tell the client you’re doing it and why. Transparency about a thin comp set protects you later.

Using a CMA to back a buyer’s offer

A CMA isn’t only a listing tool. It’s how you keep a buyer from overpaying, and how you justify going higher when the home is worth it.

The discipline is identical, just aimed at the specific home your buyer wants. Build the adjusted range for that property, then hold it up against the asking price.

If the ask sits inside your range, the home is fairly priced and you can say so with confidence. If the ask sits above your range, that gap is your data-backed case for a lower offer, and you can present the comps to the listing side instead of negotiating on feel. In a bidding war, a CMA cuts the other way: it can justify offering above ask, because you know the home supports it and you’re bidding on data rather than adrenaline.

One honest limit. A seller with a firm number in mind may ignore even a well-supported buyer CMA. It’s a negotiating tool, not a guarantee the price drops. What it does guarantee is that your buyer decides with eyes open.

The tools behind a faster comparative market analysis

Software can build a CMA in under a minute and still can’t tell you what a finished basement is worth.

Tooling helps with the parts that are mechanical. Cloud CMA and Canva templates format and brand a presentation-ready report. Comp-pulling tools organize sales and generate charts. On speed, a manual CMA runs two to four hours (30 to 90 minutes per comp), while AI-assisted products claim they can cut comp pulls to under 60 seconds. Treat that number as a vendor claim, not a verified benchmark.

What no tool does is the judgment. Comp selection, the dollar amount of each adjustment, normalizing out concessions, and the final pricing and negotiation call all stay with you. A dashboard printout still isn’t a CMA.

The one thing worth investing in early is data quality. Pricing accuracy starts with accurate, current MLS data, and a client-facing search portal lets you pressure-test a proposed price against live buyer activity instead of static comps alone. That’s one piece of a broader real estate marketing strategy, not a substitute for the method in this guide.

Lean on tools if you value your hours, but skip the pricey add-ons until you’ve run enough manual CMAs to know what the software is doing for you.

Comparative market analysis FAQ