Reconciliation: this table adds up to exactly 519 — the full row count in your updated sheet. Two of those rows are duplicate header text pasted into the data (every cell just repeats "Platform," "STAGE," "REMARKS," etc.) rather than real leads; they're kept here as a separate "Data Entry Error" line rather than silently dropped, so the total always matches your file.

Two tracking systems in one column: 103 rows use lead-quality tags (Inquiry, Prospect, Fake Lead, Not Our TG, Not Serviceable, Casual, No Response, Low Budget, Not Interested) and have no date. The other 414 rows use pipeline outcomes (Lost, Contacted, Proposal Shared, Enquiry, Qualified, plus 12 Untagged) and mostly do have dates. That's why the monthly chart only shows pipeline categories.

231 of 519 rows (44.5%) have no usable date, and 147 (28.3%) have no platform filled in — both limit how far month-over-month and platform-level analysis can go until they're made mandatory at entry.

Key finding: over a third of lost leads (35%) were never actually reached — no response, switched off, wrong number, or unreachable after repeated attempts. That's a follow-up process issue, not a marketing or product issue, and is the most fixable category here.

Second finding: Not Interested (21%) and Budget Mismatch (15%) together make up over a third of losses — these leads were reached and had a real conversation, but weren't a fit. That points to targeting/qualification at the ad level.

Classified from free-text remarks (e.g. "RNR," "switch off," "no budget," "not our scope") via keyword matching, not a clean dropdown field — 15 of 308 (4.9%) didn't clearly match any category. Recommend adding a mandatory "Lost Reason" dropdown going forward.