Channels — J. David Tax Law
Where leads actually come from — segmented by source, last 30 days
Channel matrix
| Channel | Last 30d | Prior 30d | Δ | → Customer | Close rate | → Opportunity | % of total | w/ GCLID |
|---|---|---|---|---|---|---|---|---|
| OFFLINE (manual / pre-bridge) | 528 | 47 | +481 | 30 | 6% | 35 | 50% | — |
| VENDOR — Plum | 407 | 174 | +233 | 102 | 25% | 61 | 39% | — |
| VENDOR — Squash | 52 | 1 | +51 | 0 | 0% | 0 | 5% | — |
| PAID (true GCLID-attributed) | 33 | 0 | +33 | 1 | 3% | 2 | 3% | 33 |
| VENDOR — Pecan | 14 | 6 | +8 | 1 | 7% | 2 | 1% | — |
| DIRECT | 12 | 0 | +12 | 0 | 0% | 0 | 1% | — |
| ORGANIC_SEARCH | 6 | 0 | +6 | 0 | 0% | 0 | 1% | — |
Insight: vendor channels punch above their weight
Plum / Squash / Pecan are external lead-gen vendors (not Pinnacle's Google Ads). Their leads land in HubSpot via API/Zapier with no UTM/GCLID, so HubSpot tags them OFFLINE. The lead_source field reveals the true vendor. Some of these vendors close at 50-80% — way higher than typical. Investigation needed: are these warm referrals, qualified phone leads, or already-paying customers being re-tagged?
The attribution gap
Google Ads reports 125 conversions but HubSpot has only 33 contacts with a GCLID attached. That gap is what the bridge rollout is closing. NY market is fully wired. 5 markets pending (Austin, Charlotte, Dallas, Houston, San Antonio). Each takes ~5 minutes via Growform admin. Once complete, the "Real CPL" hero KPI on the Pulse page should converge to within ~30% of Google's reported CPA.