Dwellsy IQOperator IQ

Methodology · v0.7

Portfolio Size Estimator

The estimator answers a single question: about how many units does this operator actually manage?Dwellsy IQ observes only listing activity — the subset of an operator’s portfolio that hits the open rental market in a given window. The estimator scales that observable signal up to a total-managed-units projection using cohort-calibrated multipliers.

The size-banded model

For each operator we compute an annualized URU activity figure (T12 URUs adjusted for time-on-platform when months < 12), look up the operator’s cohort by Dwellsy 7-cell × URU band, and multiply by the cohort median to get the point estimate. P25 and P75 multipliers produce the confidence band.

Cohorts are joint slices of the 7-cell taxonomy and URU activity. SFR Independent receives a 3-band split (under 100 / 100-299 / 300+) because the URU-to-units ratio drops sharply with scale in that cohort. Other cells use single-band treatments because the calibration sample didn’t justify further splits.

Cohort multipliers

Each row below shows the median + P25/P75 URU-to-total-units multipliers, the calibration sample size, and confidence tier. Point estimate = annualized URU × median; low/high = annualized URU × P25 / P75.

CohortMedianP25P75nConfidence
SFR Independent, URUs <1009.295.6911.3812Low
SFR Independent, URUs 100-2993.882.494.7429Medium
SFR Independent, URUs 300+1.881.682.406Low
SFR Institutional (all)3.462.404.184Low
Hybrid (all)3.211.355.104Low
Small MF/BTR Independent (all)1.131.012.503Low
Overall fallback4.232.538.1159Medium

Calibration sample

The multipliers were derived from a 70-pair operator-market calibration set. For each calibration pair, the operator’s observed T12 URU activity was paired with a verified or credibly-public total-managed-units figure. The ratio of total-units to URU produces the cohort multiplier; medians and quartiles were computed within each cohort to anchor the point estimate and confidence band.

The calibration analysis lives in the working file Dwellsy_IQ_Portfolio_Estimator_Calibration.xlsx (research repo, sheet “Size-Banded Model”). Multipliers above are taken verbatim from that sheet.

Confidence band interpretation

  • Medium — cohort n ≥ 20. The point estimate is anchored on a reasonably-sized sample and the P25-P75 band reflects real cohort variance.
  • Low— cohort n < 20. Treat as directional rather than exact; the band widens fast with small-sample medians.
  • None of the cohorts in v0.7 reach the High tier (n ≥ 50 with median CV under 30%). Higher confidence arrives as the calibration sample grows.

Known limitations

  • Large MF/BTR — insufficient calibration data. The 70-pair sample didn’t produce a defensible multiplier for Large MF/BTR Independent / Institutional operators. Those scorecards explicitly decline to estimate and prompt for a verified self-report via the claim flow. Faking a number for these operators would be more misleading than the explicit refusal.
  • Mega-REIT extrapolation. Operators with URU activity far beyond the calibration range (e.g. scattered-site REITs running 1000+ T12 URUs) are extrapolating off the SFR Independent 300+ band rather than interpolating within it. Treat their estimates as the roughest of the bunch.
  • Annualization assumption.Operators with < 12 months on platform receive an annualization factor on their observed URUs to project a full year. This assumes a roughly steady listing cadence; operators in the middle of a portfolio ramp will be over- or under-stated.
  • URU-to-units multipliers vary by ownership posture. Within a single cohort, scattered-site SFR operators churn more frequently than long-hold REITs of the same size. The cohort median averages across both; expect individual estimates to drift from the truth in the direction of the operator’s turnover rate vs the cohort norm.
  • Context only — not in ranking.The estimate doesn’t feed the composite ranking or star assignments. It exists to give readers a back-of-envelope scale anchor, not a precision figure.

Operator override

Operators can claim their scorecard and supply a verified self-reported portfolio size that overrides the estimate. The Layer 5 widget surfaces the claim affordance on every insufficient-data scorecard; ranked operators with claimed scorecards display the verified figure with attribution instead of the estimate.

Estimator version v0.7-portfolio-est-v0.1. Surfaces on scorecard Layer 5; pre-computed at seed time; exposed via the Ask Dwellsy IQ tools and the market brief generator. See the full methodology for the rest of the v0.7 stack.