Generated 2026-06-12. Actuals loaded through 2026-06-11. Source: local stickymetrics Postgres (forecasts.predictions, raw.sales_daily).
June Baseline (locked 6/1)1
10,414 units
$81,531 revenue p50
Run 28, chronos-bolt-lifecycle
Actuals, June 1 to 112
4,333 units
$31,488 revenue
+44% units vs baseline pace
Baseline Pace, June 1 to 111
3,010 units
$23,496 revenue p50
Revenue running +34% ahead
Model Rest-of-Month3
5,270 units
$42,979 revenue p50, June 12 to 30
Model still runs light vs trend
Daily: forecast vs actual
Bars are actual orders (SP-API). Solid line is the June 1 locked baseline p50. Dashed line is the latest nightly forecast p50 for the rest of the month. The 6/23 to 6/26 hump in both forecast lines is the event-calendar lift baked into the model.1,3
Month-end scenarios
Three ways June can land, depending on how much you trust the model vs the tape.
Scenario
Units
Revenue
vs Baseline
Locked baseline (run 28)1
10,414
$81,531
baseline
Actuals + latest model forecast2,3
9,603
$74,467
-8.7% rev
Actuals + baseline pace ratio held4
14,991
$109,263
+34% rev
Team expectation set 6/45
~13,000
~$95,000
+16.5% rev
Top 10 ASINs by June revenue so far
MTD actuals June 1 to 11 (raw.sales_daily), with the latest model p50 for June 12 to 30.2,3
2. raw.sales_daily (units_ordered, ordered_revenue_usd from Amazon SP-API), June 1 to 11, 2026, queried 2026-06-12.
3. forecasts.predictions, run_id 36 (covers 6/12) and run_id 37 (covers 6/13 to 6/30), nightly runs of 2026-06-11 and 2026-06-12.
4. Derived: baseline full-month p50 scaled by the MTD actual-to-baseline ratio (units 4,333 / 3,010 = 1.44, revenue $31,488 / $23,496 = 1.34), inputs from sources 1 and 2.
5. Working expectation recorded 2026-06-04 in agent memory (reference_stickymetrics_forecasting_system.md): model runs light, real expectation roughly 13k units and $95k. Marked approximate, not a model output.
Known model caveats (same memory note): May 2026 baseline ran 46% under actual, and new SKUs cold-start at a forecast of zero.