How seller scores work.

A seller score is a number between 0 and 100 that answers a single question: how likely is this homeowner to list within the next 90 days? It is not a prediction of price, motivation quality, or whether they'll hire you. It is a likelihood-to-act signal, updated every night.

Scores above 70 are warm. Scores above 85 are urgent - those contacts are typically within a few weeks of a decision. Scores below 40 are background noise: in the database, getting nurtured, not ready for a personal outreach.

Every score is built from five categories of input. No single signal is disqualifying - they combine and reinforce each other.

• Equity position. Homeowners with 30–60% equity have both the ability and the incentive to sell. Below 20% and the math doesn't work. Above 70% and they're often comfortable enough to wait indefinitely. The sweet spot is the middle band - enough to profit, not so much that there's no urgency.
• Tenure. The national median tenure before selling is between 6 and 8 years. Contacts at 5+ years get a tenure bump. Contacts at 8+ years who haven't listed yet get a larger one - the probability distribution peaks and then trails off.
• Behavioral activity. Valuation page visits, frequency of MLS browsing, email engagement with market-update content - these are the real-time layer. A contact who was cold three weeks ago can spike to an 80 after a burst of browsing activity.
• Life-event signals. Job changes, divorces, new children, and deaths in the household all correlate with listing decisions. These come from opt-in data partners and are blended in when available.
• Market context. Local days-on-market, absorption rate, and price-trend data create a neighborhood multiplier. A contact who scores 65 in a seller's market scores higher than the same contact in a flat market - because the probability of acting increases when conditions are favorable.

The exact weights shift quarterly as the model retrains on closed-deal data from active Yapii workspaces. As of Q1 2026, the approximate contribution breakdown looks like this:

We publish this because black-box scores erode trust. When an agent understands why a contact ranked highly, they make a better call. The reason shown next to each score isn't decorative - it's the top contributing factor, surfaced so you can lead with it.

Scores update nightly for the full database and same-day for material events - a new valuation-page visit, a listing going active, a public-record transfer. You don't need to refresh anything. Changes appear in the morning queue automatically.

Scores decay when there's no activity. A contact who spiked to 82 after a burst of browsing and then went quiet for 30 days will drift back toward their baseline. This is intentional - the model should reflect current intent, not a snapshot from months ago.

A high score is not a guarantee and is not a replacement for relationship. It tells you that the data suggests someone is close to a decision - not that they've made one, not that they'll choose you, and not that they'll be pleasant on the phone.

The agents who get the most out of Yapii treat the score as the reason to reach out, not the script. They call with genuine curiosity. The data earns the conversation; the relationship closes the deal.