What it does
Pricing and Packaging Analyst exists because pricing decisions tend to get made once a year, by gut feel, in a spreadsheet nobody trusts. You supply a snapshot — customer usage and current invoices, competitor prices, and CRM win/loss reasons — and it models three packaging scenarios against your actual customer base. The three are computed from your data rather than assumed. Usage-based prices per unit at a blended rate derived from your total revenue over total usage. Tiered builds three usage bands and flat-prices each at the in-band median invoice. Flat plus overage sets a base fee at a low usage percentile and an overage rate informed by the competitor prices you supplied. For each scenario you get the per-customer invoice change, the counts of winners, losers and unchanged, the winning and losing customer segments, and the net revenue impact in absolute terms and percentage. That per-customer view is the part a spreadsheet exercise usually skips and the part that decides whether a change is survivable. The recommendation is deterministic and constrained: the scenario with the best net revenue that does not shrink revenue by more than 5%. It is a rule you can read and argue with rather than a judgement call. Alongside it comes a commercial risk signal from your CRM data — what share of lost deals already cite price. If that number is high, the pricing conversation carries risk before you change anything, and the agent surfaces it rather than modelling in a vacuum. Every run carries a fixed warning that the agent is decision support, not an action: a pricing change creates customer communication that is hard to reverse, so the output is meant to be reviewed with sales and finance before anything reaches a customer. Each scenario is callable alone — `model_usage_based`, `model_tiered`, `model_flat_overage` — as are `recommend_scenario` and `assess_price_risk`. All arithmetic is deterministic; the language model only narrates it. No connectors and no credentials — nothing is queried from BigQuery, a CRM or a billing system. Built for product, product marketing and finance teams.
Example prompts
- What happens to revenue if we move to usage-based pricing?
- Model a tiered package against our current customers
- Which customers would pay more and which would pay less?
