Three Realtors gave me three different prices for my house. Who's right?

It’s one of the most common — and most consequential — moments in selling a home: three interviews, three professionals, three numbers spread by hundreds of thousands of dollars. The natural conclusions are all wrong: that someone is lying (usually no one is), that the truth is in the middle (it often isn’t), or that the highest number is the most confident (it’s sometimes the least). The spread is information. Here’s how to read it.

Where the spread actually comes from

1. Different comparable sales. The largest driver. One agent anchored on the renovated sale around the corner; another on the original-condition sale two streets over; the third included a pending sale the others excluded. None of these choices is dishonest — but each embeds an assumption about which past sale your home most resembles. Make each agent name the two or three comps doing the most work in their number. Then look at those homes yourself. You know better than anyone whether your kitchen is the renovated one or the 2003 one.

2. Different condition and preparation assumptions. Agent B’s $1.95M assumed whole-home staging and $30K of paint and floors; Agent A’s $1.8M assumed selling as-is. Those aren’t different valuations of the same plan — they’re different plans. Restate every number on the same basis: expected sale price, minus assumed prep cost, equals comparable net. Spreads routinely halve when you do this.

3. Different reads of right now. How many competing listings, how fast the last three sold, whether rates just moved. This is the most legitimately debatable input — and the one where an agent’s recent activity in your micro-market matters most. An agent with three active buyers in the neighborhood is reading the room; an agent extrapolating from another zip code is guessing.

4. Different launch strategies. Under-list to drive multiple offers; list at value; list high “to leave room.” Each can be defended in the right market — but it means the three list prices aren’t even measuring the same thing. Always extract the second number: “Regardless of list price, what do you expect it to actually sell for, and in how many days?” Compare those.

5. And yes — sometimes the number is bait. “Buying the listing”: quoting an optimistic price to win the agreement, then engineering reductions. It’s a minority practice with a known signature: the high number arrives with thin comps, no checkpoint plan, and pressure to sign quickly. The defense isn’t cynicism about every high number — it’s the decomposition above, which the bait number uniquely cannot survive.

The reconciliation, in one table

Ask all three agents for the same six items and grid them:

Item	Agent A	Agent B	Agent C
Recommended list price
Expected sale price
Top 3 comps relied on
Condition/prep assumed (and its cost)
Expected days on market
“What would change my mind” — the two-week checkpoint

Read the grid, not the headline numbers. Convergent expected-sale-prices with divergent list prices = a strategy choice, yours to make. A divergent expected-sale-price = one agent knows (or assumes) something the others don’t — find out what. A number that can’t produce its comps = disqualified, however pleasant the meeting was.

The deeper point

The three-prices moment feels like a pricing problem. It’s actually the best agent-evaluation evidence you’ll ever get for free: you’ve just watched three professionals reason under uncertainty about your largest asset. The one whose reasoning is transparent, evidence-anchored, and falsifiable — who tells you what would prove them wrong — is showing you exactly how they’ll negotiate, advise, and course-correct after you hire them. If you’d like a referee with no listing to win putting all three sets of work through the same normalization — or you’re staring at three proposals right now — that’s the exact job of the Evaluation Report, and you can read a complete sample first.

Questions people ask

Shouldn't I just hire the agent who says my house is worth the most?

That instinct is exactly what a minority of agents exploit — quoting high to win the listing, then walking the price down once you're locked in. The tell isn't the high number itself (sometimes the high number is right); it's whether evidence supports it. Ask the high agent: which sold comparables support this, what must be true for it to work, and what happens to our plan if it doesn't? A confident answer names comps and a checkpoint date. A weak one repeats adjectives.

What's the difference between list price and expected sale price?

The list price is a marketing decision; the sale price is the outcome. Some agents recommend listing under expected value to drive multiple offers; others list at value; a few list above to 'leave room to negotiate' (which mostly manufactures days on market). Two agents can propose different LIST prices while expecting the identical SALE price — which is why you must ask every agent for both numbers. Compare expected sale prices, net of their assumed prep costs. That's the real comparison.

How do I ask agents to show their work without offending them?

Ask all three identically: 'Walk me through the three comps doing the most work in your number, what condition/prep you assumed, your recommended list price versus the sale price you actually expect, and what result in the first two weeks would tell you the price is wrong.' Good agents enjoy this conversation — it's their craft. An agent who can't or won't decompose their own number has told you how they'll handle the harder conversations later.

What if the spread is still wide after they show their work?

Then the spread has located a genuine uncertainty — usually condition (how buyers will discount your deferred maintenance) or a thin comp set. That's solvable with data: a pre-listing inspection to bound the condition question, or widening the comp window. And it's precisely the situation where a neutral second read earns its keep — a referee with no listing to win has no reason to flatter the number.