Brindle & Co.

The headline this month is the lead-stage scoring model. Two-hundred MQLs ran through the v2 model for four weeks; selection is converting to SQL at 18% above the legacy scorer, and the AE team's feedback has moved from "this list is noise" to "most of these are real". We're ready for full cutover at the end of June.

The Marketo to HubSpot migration is on schedule but the lead-routing pilot surfaced a DMN case-merge bug on multi-touch leads. Angus patched the immediate entry and the pattern detector picked up the broader category — see the case-merge analysis above. I think it's right; I'd like to scope a fuzzy-match pre-pass project rather than continue patching entry-by-entry. The AE team is currently absorbing 4+ hours of manual dedup a week which is a poor use of their time.

Two slow burns. The procurement-tender pattern around creative work is now a clear category — Raadad at Celeste Labs has flagged three entries in 30 days, brand continuity is suffering, and the Q3 reel is the live test of whether we can get the policy changed in time. I'd recommend a one-meeting fix with your procurement director; happy to write the proposal. Separately, Knowledge staleness keeps biting us at the channel-partner edge — Jenny pulled the audit, partners cite articles last updated in September. The analysis above outlines a review-SLA proposal; the partner-portal project (PRJ-06) is hearing the same thing from the other direction, so it's worth bundling.

Time logged this month is up 18% from April — most of that is the scoring-model validation (Angus) and the procurement-pattern triage (Raadad). Breakdown by hand is in the workspace.