Summit Health Nutrition: macro signals predicted consumer trade-down 6 weeks before POS moved.

The problem: a tiered protein catalog exposed to macro-driven trade-down

Summit Health Nutrition sells a tiered protein product catalog: premium whey isolate products at $45–$55 price points, mid-tier blended protein products at $28–$38, and budget-adjacent options at $18–$24. During inflationary periods, this tier structure becomes a macro risk: sustained CPI pressure systematically shifts consumer purchasing from premium to mid-tier across the protein category.

The predictability of this trade-down dynamic is well-established in CPG economics — but the timing is the challenge. The POS data shift lags behind the macro pressure by 6–10 weeks. Consumers don't immediately switch the day CPI accelerates — it takes sustained pressure across 2–3 months of grocery basket increases before the behavioral shift manifests in protein purchase behavior. By the time POS scan data showed Summit's premium tier losing velocity, they were already 6 weeks into a demand shift that their production commitments hadn't anticipated.

The operational consequence was expensive: overproduction of premium SKUs (which had lower-than-expected demand) combined with under-production of mid-tier SKUs (which were experiencing higher-than-ERP-predicted demand). They were simultaneously sitting on premium inventory write-offs and stock-out on mid-tier — the worst possible combination for a supply planning team during an S&OP review.

The signal fusion deployment

Heatvelo's deployment for Summit prioritized macro economic indicators as the primary signal category for their premium and mid-tier SKU tiers. The specific macro signals calibrated for protein category trade-down: CPI (all items, food at home sub-index), University of Michigan consumer sentiment index (12-month outlook component), real wage growth rate, and the sports nutrition category price sensitivity index derived from competitive shelf pricing data.

The baseline calibration established Summit's historical trade-down coefficient: for each 1% sustained CPI acceleration above 3.5%, Summit's premium protein tier historically lost approximately 4.2% of velocity to mid-tier over a 6–8 week lag period. This coefficient was calibrated from 18 months of co-movement data between macro indicators and Summit's own POS history.

During the pilot period, CPI food at home had been accelerating for 8 consecutive weeks at the time Heatvelo began running its signal fusion model. In week 3 of the pilot, Heatvelo's forecast output included a macro signal flag: "CPI sustained pressure flag active. Premium protein tier demand: -8% vs ERP baseline expected by weeks 7–10. Mid-tier protein: +12% vs ERP baseline expected by weeks 7–10. Recommend production rebalancing before week 6 production commitment window closes."

Summit's supply planning manager took the signal to their VP Supply Chain with a production rebalancing recommendation in week 4. The decision to reduce premium production runs by 9% and increase mid-tier production by 11% was made by week 5 — 4 weeks before the demand shift was visible in POS data.

The outcome: write-offs reduced, mid-tier stock-outs avoided

By week 9, POS data confirmed the trade-down shift: premium tier velocity was down 11% and mid-tier was up 14% versus the ERP baseline forecast. Summit had pre-positioned for this shift 4 weeks earlier. Their premium inventory build was 9% lower than it would have been under ERP-only forecasting, and their mid-tier inventory was 11% higher. The result: no mid-tier stock-outs during the demand peak, and premium write-offs were reduced by approximately 14% compared to what the ERP-only production plan would have produced.

Overall pilot accuracy: 93% MAPE across Summit's 520-SKU catalog (vs. 71% ERP baseline). The macro signal provided the most value on the tiered SKU structure — the 22-point MAPE improvement was concentrated in the 140 SKUs that straddled the premium/mid-tier price boundary, where the trade-down dynamics were most pronounced.