A forty-agent desk drowning in portal imports
Situation. A Calgary brokerage had accumulated roughly forty thousand contacts from portal feeds, open-house tablets and legacy spreadsheets. Agents spent Monday mornings manually sorting "who might be warm" with no shared definition. Hot leads from Saturday open houses often waited until Wednesday for a call.
Approach. We audited their CRM fields, defined a scoring rubric with the ops manager and wired engagement events — email opens, showing requests, repeat portal views — into a nightly refresh. A weekly review meeting calibrated weights when the market shifted. Agents received a daily shortlist of fifteen contacts, not an overwhelming sortable grid.
Human review & compliance. Scoring rules excluded proxies for protected characteristics. When sample bias appeared in certain quadrants, we widened geography rather than over-fitting. PIPEDA retention schedules were applied to stale contacts — archive, not silent delete.
Outcome framing. Triage time dropped for teams that adopted the shortlist habit. Conversion rates varied by agent and season — we reported that honestly in the close-loop review rather than publishing a universal "3× leads" claim.
Scoring review · anonymised engagement