Every MLS platform will spit out a CMA in one click. Every seller knows this. If your price recommendation looks like the auto-generated one, they'll ignore it.
The 10-minute version
1. Pull five sold comps in the last 90 days within a half-mile.
2. Cross out the two least comparable — different sqft tier, different condition, or a corner-lot/lot-size outlier.
3. Of the remaining three, adjust for the material differences (finished basement, primary on main, backyard).
4. Anchor your range to the two most similar, not the average.
5. Write it up in three lines: range, why, what to do if the market disagrees.
Why "average of comps" fails
Averages assume every comp is equally comparable. They aren't. One comp is usually 2x more relevant than the others; two are marginal. Averaging drags your recommendation toward the marginal ones.
The write-up template
- Recommended range: $X – $Y.
- Why: [The one comp closest to subject] sold at $A after adjustments; [second closest] at $B. Both cluster there. The higher outlier had [reason]; the lower had [reason].
- If it doesn't move: trigger a $Z reduction at day 21 if we're under [showings target] with [feedback signal].
That's a CMA a seller can argue with — which means it's a CMA that gets discussed, not dismissed.
Try the [CMA Summary tool](/comparative-market-analysis) with your next pull.