How to Analyze Product Pricing Using AI
Review average selling price, discount rates, and product mix across governed revenue metrics with an AI agent using your Cube deployment.
- Compare weighted discount rates within each product
- Separate product mix from price differences
- Check the metric definitions behind each result
What is product pricing analysis with AI?
Product pricing analysis reviews list value, net revenue, units sold, and discounts across a defined time window. Finance, revenue operations, and data teams can use an AI agent to compare those governed metrics by product and customer segment, then check how the totals were calculated.
This workflow describes historical results from an existing Cube deployment. It does not set a price, estimate demand elasticity, or establish why a discount was offered. An overall average can move when the mix of products changes, so compare the same products across segments before interpreting a difference.
Choose the pricing definition and reporting window
Set one currency, a clear eligible-order rule, and the metric definitions before comparing products. The worked example uses July 1–September 30, 2026; all of its figures are fictional.
| Metric | Illustrative example |
|---|---|
| Reporting window | July 1–September 30, 2026 inclusive, in one reporting time zone. |
| Eligible units | 1,400 fulfilled units across Starter and Pro; canceled lines are excluded in this fictional example. |
| List value | $180,000 across eligible lines, calculated from list price and quantity. |
| Net revenue | $157,500 across the same eligible lines; the difference from list value is $22,500. |
| Weighted average selling price | $157,500 net revenue / 1,400 units = $112.50 per unit. |
| Weighted discount rate | $22,500 discount value / $180,000 list value = 12.5%. |
Copy. Paste. Run your pricing review.
Select a Cube deployment, then run this read-only review in your preferred agent.
Start with a Cube deployment connected to your data and a model containing the metrics you want to analyze.
Connect over MCP
Connect your agent to your existing Cube deployment and sign in through OAuth.
https://cubecloud.dev/mcp
Complete OAuth in your agent to connect your Cube account.
Use your default deployment or select another one available to your account.
Connect with CLI
Install the Cube CLI and sign in from a terminal-capable agent.
curl -fsSL https://raw.githubusercontent.com/cube-js/cube/master/install-cli.sh | sh
cube login --url https://TENANT.cubecloud.dev
cube spec --json
02Run this prompt in your agent
Choose your agent and copy the prompt below.
Use the Cube MCP server connected in Codex and select the intended deployment. Run read-only analytics queries. Run a historical product pricing review in my connected Cube deployment. Use read-only analytics requests and keep customer and contract identifiers private. DEFINITIONSInspect the model for its approved list price or list amount, net price or net revenue, eligible quantity, product, customer segment, transaction date, currency, reporting time zone, and transaction status. Report the modeled definitions and the line-item grain. Follow the model's rules for cancellations, returns, refunds, credits, taxes, and free items. If a required definition is missing or ambiguous, report the gap instead of inventing a result. For line-item amounts, use the governed extended list and net amounts when available. If only unit prices are modeled, multiply each unit price by its eligible quantity once; do not multiply an already extended amount again. Total list value is the sum of eligible list amounts. Net revenue is the sum of eligible net amounts. Discount value is list value minus net revenue. Weighted average selling price (ASP) is net revenue divided by eligible units. Weighted discount rate is discount value divided by list value. Calculate rollups from the summed numerators and denominators; do not average row-level prices or segment percentages. Report a negative discount as a premium instead of flooring it at zero. ANALYSIS1. Review the last three complete calendar months available in the model at run time. State each month's exact inclusive start and end dates and the reporting time zone. Exclude the partial current month. If the model does not contain three complete months, say which periods are available.2. Compare product and customer segment using segment values effective on the transaction date. For each product-by-segment group, show line or order counts when available, eligible units, list value, net revenue, discount value, weighted discount rate, and weighted ASP. Keep currency consistent; do not combine currencies without an approved conversion rule.3. Reconcile groups to the overall totals using summed amounts and units. Compare the same product across segments, and describe product-mix differences separately from within-product price or discount differences. Do not attribute a change to pricing when the product mix also changed.4. Report missing prices, quantities, dates, product or segment values, duplicate-grain risks, and excluded-row counts with the model's exclusion rule. Do not silently drop incomplete rows or make a comparison from mismatched definitions. OUTPUTReturn the metric definitions, time zone and exact date boundaries, a product-by-segment table with each numerator and denominator, reconciled totals, and a concise interpretation. Keep observed patterns separate from hypotheses. Do not recommend a price, estimate elasticity or willingness to pay, forecast results, or claim that a discount caused a sales change.
How it works
Confirm the price and net-sales basis
Check the modeled list and net amounts, quantity, currency, transaction date, and line-item eligibility rules. A price comparison is only useful when returns, cancellations, and contract adjustments follow the same definition across groups.
Calculate rates from totals
Calculate ASP as eligible net revenue divided by eligible units, and weighted discount as discount value divided by list value. Reconcile each product and segment to the overall totals instead of averaging percentages.
Compare like with like
Compare the same product across customer segments, then review how many units of each product each segment bought. A change in the overall ASP can come from product mix as well as from within-product price differences.
Read discounts alongside price realization
The chart compares fictional Q3 2026 weighted discount rates for each product across SMB and mid-market segments. The roll-up and worked example use fictional values, not executed queries or customer data.
| Product / segment | Units | List value | Net revenue | Weighted discount | Weighted ASP |
|---|---|---|---|---|---|
| Starter / SMB | 600 | $60,000 | $54,000 | 10% | $90.00 |
| Starter / Mid-market | 400 | $40,000 | $34,000 | 15% | $85.00 |
| Pro / SMB | 100 | $20,000 | $18,500 | 7.5% | $185.00 |
| Pro / Mid-market | 300 | $60,000 | $51,000 | 15% | $170.00 |
| Total | 1,400 | $180,000 | $157,500 | 12.5% | $112.50 |
- The total discount value is $22,500 ($180,000 list value minus $157,500 net revenue), so the weighted discount is 12.5%. Weighted ASP is $157,500 divided by 1,400 units, or $112.50.
- Within Starter, the Mid-market discount rate is 15% versus 10% for SMB, a 5 percentage-point difference. Within Pro, it is 15% versus 7.5%, a 7.5-point difference.
- Product mix differs by segment: SMB has 600 Starter and 100 Pro units, while Mid-market has 400 Starter and 300 Pro units. Compare the same products before interpreting any difference in an overall segment average.
The table and image use fictional values; they are not executed queries, customer results, benchmarks, or price recommendations. They do not establish willingness to pay, price elasticity, margin impact, or why a discount was offered. Actual results depend on the modeled currency, contract, return, and net-price definitions.