Your second brain
for closing deals.
Speak after a showing. Forward an email. Pull up a client. Lumi captures the soft signals, fills the brief, and feeds Claude — automatically.
- Voice → CRM, auto. No forms.
- Works offline. Syncs when you're back.
- Free for agents in EU · LatAm · MENA.
9-min read · Updated April 2026
Lumi · Wednesday
Good morning, Niki.
Two showings · three leads need a nudge.
Showing · Passeig de Gràcia 84
30-min quarterly review.
AI does the math. I make the call.
Most agents skip quarterly reviews — they feel like corporate ceremony, they take hours, and the conclusions are usually “keep doing everything”. The protocol replaces this with a 30-minute pass: AI computes ROI per channel, surfaces the data, and the agent makes exactly two stops and one double-down. Compounded over four quarters, the allocation shifts dramatically toward what works.
Five rules. 30 minutes a quarter.
The discipline that makes quarterly review actually shift the agent's behaviour. Each rule prevents one of the failure modes that turns review from useful into ceremony.
- 01
ROI is per hour, not per deal.
Most quarterly reviews look at GCI per channel and conclude 'Channel X produced the most money'. But Channel X may have consumed 200 hours; another channel produced 60% of that GCI in 40 hours. Per-hour ROI is the only honest comparison. The protocol's first job is computing this honestly — and the comparison usually flips the agent's intuition about which channels work.
- 02
Always 2 stops. Always 1 double-down. Constraints force decisions.
The 2-stop / 1-double-down rule is the protocol's core constraint. Without it, agents review and conclude 'this all went pretty well, let's keep doing everything'. With it, the agent is forced to drop something — and to focus on something. The constraint is uncomfortable; the constraint is the value.
- 03
The double-down must be specific.
Vague recommendations don't ship. 'Double down on past-client referrals' is a sentiment; 'Allocate 20 hours/quarter to past-client check-in calls and 1 quarterly handwritten card per past client' is a plan. The protocol's prompt actively forbids vague double-downs and produces specific allocations the agent can implement on day 1 of the next quarter.
- 04
Calibration: last quarter's recommendations get reviewed.
Every quarterly review starts by reporting on the previous quarter's stops and double-down. Did the stops actually free up time? Did the double-down produce the expected lift? This honesty creates the only feedback loop that makes the protocol get smarter over time — without it, the recommendations are unmoored from reality.
- 05
The compound effect: 4 quarters of disciplined stops + double-downs.
One quarter of the protocol shifts the agent's allocation by ~15%. Four consecutive quarters of disciplined application shifts it by 50-70% — toward the channels that actually produce. The compound effect is what separates agents who run quarterly reviews honestly from agents who feel like they should but don't. The lift is in the discipline of the cadence.
Three reviews that change nothing.
The shapes the review defaults to without strict discipline. Each one is what 90% of agent quarterly reviews look like — and why those reviews don't compound into shifted allocations.
“What worked: open houses + IG outreach. What didn't: nothing major. The call: keep doing what's working, optimise on the margins.”
No stops named, no double-down specified. The review concludes 'keep doing everything', which is the same allocation as last quarter — and the same as the quarter before. The protocol's discipline forces a stop and a double-down precisely because 'keep doing everything' is the default that produces flat performance.
“Best channel: open houses (€220k GCI). Worst channel: handwritten cards (€18k GCI). Stop: handwritten cards.”
Ranked by total GCI without considering hours invested. Open houses may have consumed 300 hours for €220k (€733/hr); handwritten cards 8 hours for €18k (€2,250/hr). The 'worst' channel by total GCI is actually the highest ROI channel by hour. Per-hour ROI flips the call.
“Double down on referrals — they're our best channel.”
Sentiment, not plan. The agent reads this and... keeps doing what they were doing. Specific double-downs name hours, activities, allocations: 'Add 4 hours/week to past-client outreach via the radar protocol' is implementable. 'Double down on referrals' is not.
What to feed Claude.
Run on the last working day of each quarter. Inputs: closed transactions (CRM), self-reported hours per channel (a 5-min Google Form completed across the quarter), previous review's recommendations.
You are a senior real-estate agent's
quarterly business-review analyst.
INPUT
You receive: every closed transaction
of the quarter with source channel
(referral / IG / open-house / cold
outreach / past client / other), GCI
per deal, agent hours invested in
each channel during the quarter
(self-reported), and the previous
quarter's review notes.
OUTPUT
A 1-page review with three sections:
WHAT WORKED — the 1 channel with
the highest ROI
(GCI/hour). Names it,
quantifies it, and
says why this quarter.
WHAT DIDN'T — the 1-2 channels
with the lowest ROI
(or unmeasurable ROI).
Names them honestly.
THE CALL — exactly:
- 2 STOPS: which 2 channels to
pause or drop next quarter.
Specific reasoning per stop.
- 1 DOUBLE-DOWN: which 1 channel
to invest more in next quarter.
Specific allocation suggestion.
The review ends with a 1-line
calibration note: how accurate were
last quarter's stops and double-down?
RULES (non-negotiable)
1. Always 2 stops, always 1 double-
down. Never more, never fewer.
The discipline is in the constraint.
2. ROI is GCI per hour, not GCI
total. A channel that produced
€40k from 200 hours is worse
than a channel that produced
€30k from 60 hours.
3. Don't recommend stopping a channel
the agent is emotionally attached
to without strong data. Surface
the data; let the agent decide.
4. The double-down recommendation
must be specific (hours/budget/
activities), not vague ('do more
of this').
5. Calibration: each quarterly
review reports on the previous
quarter's recommendations
(what was stopped, what was
doubled down, what changed).
ANTI-PATTERNS (never produce these)
- More than 2 stops (paralysis)
- More than 1 double-down (no focus)
- Vague recommendations ('focus on
what works')
- ROI based on GCI total alone
- Hiding the calibration deltaRun on the last working day of each quarter. The agent reads the review (5 min), considers (15 min), commits to the call (10 min).
Running the review is step one.
Honouring 2 stops + 1 double-down is step two.
Lumi is the app that runs this workflow for you. You speak after a showing — Lumi captures the soft signals. You forward an email — Lumi updates the constraints. You open the app at 8am — the brief is already there, ready to feed Claude.
- Voice → structured CRM, automatically
- No forms. No data entry. No copy-paste.
- Free for agents in EU · LatAm · MENA
Lumi · Wednesday
Good morning, Niki.
Two showings · three leads need a nudge.
Showing · Passeig de Gràcia 84
Pipeline
Active
8
Warm
4
Cold
2
Clara Ruiz
Active€1.8M · 3BR
Passeig de Gràcia showing · 11:30
Andreas Moreno
Active€2.4M · 4BR
Send comps by 18:00
Dimitri Schneider
Warm€900K · 2BR
Contract review today
Silent 3d · last 3 days ago
Sarah Mitchell
Cold€1.2M · 3BR
Draft re-engagement
Silent 9d · last 9 days ago
A real-estate adaptation of the quarterly-review discipline from startup operations — the simplest version that actually changes behaviour quarter over quarter. Our slice: the 2-stops + 1-double-down rule applied to the agent's channel allocation.
More guides like this on @lumi.estate. Follow if any of this was useful — it's how we know to keep writing.