Cohen Crowder — conversion rate

Gerrard's Insurance · lead → client · 26 May – 26 Aug 2026 · gerrards-crm synced 26 Aug 14:44 · generated 2026-08-26 14:44 NZ

3-month rate

13.2%

37 clients won from 281 leads

Rest of team

18.3%

213 from 1,166 · gap -5.1 pts

Rank

8 / 8

of brokers with 20+ leads

Trend

↓

16.3% Jun → 13.0% Jul → 12.1% Aug

The read

Cohen is the lowest of the eight brokers measured — 13.2% against the team's 18.3%, and he is below the field in every single month of the window. This is the one gap of the three that clears significance on the full window (z = −2.03, p = 0.042), though it softens on mature-only leads (p = 0.087).

The gap is not uniform — it is entirely in web leads. On untagged web/inbound leads he converts 8.8% against the team's 16.4%. On phone leads he converts 46.2% against 28.1%, better than anyone here. He is not a weak closer; he is losing leads before the conversation starts. That got worse exactly as his load went from 49 leads in June to 123 in July.

June's 16.3% is his best month and his least trustworthy — he made zero calls in the week of 22 June while the office made 421, so he was out for roughly 15–26 June, and Dean came online mid-month and absorbed the volume. That leaves a 49-lead month weighted to the first fortnight. The honest comparison is July's 13.0% against May's 7.3%, with August tracking below July.

The map — 4-week rolling conversion vs the rest of the team

Rolling window smooths weekly noise (his cohorts run 4–33 leads). The line never touches the team line. Dots are individual weeks, sized by lead volume.

Cohen (4-wk rolling) Rest of team (4-wk rolling) Not yet mature

By month

June is a short month for him — see the read above. His lead volume went up 2.5× between June and July while his conversion went the other way.

Month	Leads	Won	Cohen	Team	Gap
May mostly pre-window	110	8	7.3%	21.2%	-13.9
June on leave 15–26 Jun	49	8	16.3%	20.8%	-4.4
July	123	16	13.0%	19.8%	-6.8
August	91	11	12.1%	13.6%	-1.5

Where Cohen sits — all brokers, 1 Jun – 5 Aug cohort

By lead source

This is the whole finding: he beats the field on the phone and loses to it on web leads, which are 87% of his volume.

Lead source	n	Cohen	Team	Read
(no source tag)	147	8.8%	16.4%	-7.6 pts
Phone Call	13	46.2%	28.1%	+18.1 pts
Client Referral	3	66.7%	33.3%	n too small
Networking Referral	2	50.0%	50.0%	n too small
Professional Referral	2	0.0%	29.2%	n too small

What counts as one of Cohen's leads

Three systems count August differently, and only one describes leads a human actually worked. The lead-distribution service logged 171 rows for Cohen in August — but that is 150 distinct leads double-counted by the CRM tee, and 155 of the 171 were never delivered to GHL at all. Matching those 150 on email to where they actually landed: only 40 ended up with Cohen; the rest went to Kynwin (32), Dean (28), Jordan (22), Caitlin (11) and five others. Also worth knowing: only 47 of his 91 August leads appear in the digital pipeline at all — about half his book arrives by phone, referral or direct, and that is the half he converts best.

Count	August	What it is
Whole business — GHL contacts	437	Every new lead, all sources, all brokers
Cohen — GHL assigned	91	The denominator used here. Leads GHL says are his
Cohen — lead-dist rows	171	150 distinct · 155 never delivered to GHL

Method & what I would not claim

Wins are read from the client won contact tag, not opportunity status. Won opportunities are deleted from the GHL pipeline, so status understates wins by roughly 4× from May 2026.
The denominator is contacts, not opportunities. The "Lost Lead" workflow destroyed opportunity rows through May–June, so opportunity-based conversion reads 5.0% company-wide in May against a stable ~20% on the tag measure.
August is incomplete. Half of wins land within 6 days, but the slowest 10% take 23+, so leads added after roughly 27 July still have wins to come. Expect August to finish above what it shows.
Significance: full window z = -2.03 (p = 0.042); on the fully-mature subset only (168 leads, 13.7% vs 19.3%) z = -1.71 (p = 0.087).
Deal counts only — no revenue weighting. broker_earnings stopped at 22 Jun 2026 and conversions holds zero won rows, so a $500 policy counts the same as a $50k one.
No speed-to-contact. opportunities.first_contacted_at is NULL on every row in this database.
Do not use enquiries.assigned_broker_id as a lead count. It is a shadow routing decision that matches real GHL assignment about 27% of the time, and it double-counts via the CRM tee.
The August count moved while this was being written. It read 92 when verified in GHL on the morning of 26 August; a later sync reassigned one lead and this page shows 91. Both were correct at the time — the mirror is live.
The leave dates are inferred from call activity, not confirmed with anyone. Worth checking with Ethan before using them in a review.

Call activity by week (used only to sense-check presence, not performance)

Week	Cohen	All brokers
2026-06-01	22	257
2026-06-08	15	400
2026-06-15	10	395
2026-06-22	0	421
2026-06-29	25	516
2026-07-06	40	418
2026-07-13	47	518
2026-07-20	89	557
2026-07-27	60	607
2026-08-03	87	557
2026-08-10	57	611
2026-08-17	79	699
2026-08-24	24	369