Definition
The trailing change in the ratio of full-price transactions to total transactions, measuring whether promotional AI is training customers to never pay full price by systematically shifting the purchase mix toward discounted transactions.
The Measurement Gap
The Gap
Promotional AI is designed to find the optimal moment, channel, and offer to convert customers. It gets better at this over time. More conversions happen during promotions. The conversion numbers look great. But there is a question nobody is asking: what is happening to the customers who used to buy at full price?
Full-price transaction ratio is sometimes tracked as a merchandising metric, but it is rarely connected to promotional strategy. Revenue growth during promotional periods is celebrated without checking whether non-promotional revenue is declining at the same rate. Customer acquisition cost per channel is tracked, but the cost of converting existing full-price customers into discount-dependent customers is not measured because it does not fit neatly into acquisition accounting. Promotional lift metrics reward every conversion without distinguishing between a new customer attracted by the offer and an existing customer who learned to wait for it.
The structural problem is that promotional AI optimizes for conversions, and conversions are easiest to generate by offering discounts to people who are already inclined to buy. Over time, the AI learns that the most responsive audience is existing customers -- and the most effective trigger is a discount. The result is a systematic migration of the customer base from full-price purchasing to discount-dependent purchasing. Each individual optimization looks like a success. The aggregate trend is a dependency spiral.
The Decision
We propose DDI to answer: Is our promotional AI training our customers to never pay full price? This metric would track the decline in full-price transaction share over time, measured monthly and compared year-over-year, making the dependency trend visible before it becomes structurally irreversible.
Formula
Plain Language
Count the number of transactions where no promotional discount was applied. Divide by total transactions. That is the full-price ratio. Now compare the current full-price ratio to the baseline period (12 months ago or pre-promotional-AI deployment). The decline in percentage points is DDI. If the full-price ratio was 65% a year ago and is 57% now, DDI is 8 percentage points -- meaning the business has lost 8 points of full-price purchasing behavior in 12 months.
Notation
DDI = FP_ratio_baseline - FP_ratio_current
Where:
FP_ratio = Full_price_transactions / Total_transactions x 100
Variables
| Variable | Definition | Units | Source System |
|---|---|---|---|
| Full_price_transactions | Number of transactions where no promotional discount was applied | Count | E-commerce / POS / Billing |
| Total_transactions | Total number of transactions in the measurement period | Count | E-commerce / POS / Billing |
| FP_ratio_baseline | Full-price transaction ratio in the baseline period (12 months prior or pre-AI deployment) | Percentage | Historical Data |
| FP_ratio_current | Full-price transaction ratio in the current measurement period | Percentage | Derived |
Proposed formula, pending validation pilots.