Five prompts to help you build a clearer view of your discount marketing strategy, from customer behavior through to margin.

We all know that shoppers love a good deal, and that's truer now than ever: research published in June 2026 found that 62% of UK shoppers are more promotionally driven in their purchasing than they were a year earlier.

This presents a key challenge for retail brands today: building a healthy discount marketing strategy. Brands need to deploy discounts carefully, or risk becoming trapped in a discount spiral where customers are inadvertently trained to only buy on promotion.

To understand how discounts are impacting your business, you need to go to the data, and MCP connectors like Ometria's Deep Insights Agent have made that much easier.

Deep Insights Agent connects your Ometria data to LLMs like Claude and Microsoft Copilot, so you can ask questions about your customers, campaigns, and strategy in plain English and get answers, visualizations, and dashboards drawn from your full customer universe.

In this blog, we'll share five prompts you can give Deep Insights Agent (or your MCP connector of choice) to build a clearer view of your discount marketing strategy, split across customer behavior and value, and product and margin mix.

Plus, for more ways to use Deep Insights Agent, we've put together seven prompts for building a smarter Black Friday strategy, as well as a comprehensive prompt library covering everything from segmentation and lifecycle analysis to campaign performance and customer behavior.

Product and margin mix

These two prompts give you the size of the picture: how much of your revenue comes from discounted product, where that sits across the year, and how much margin it's costing you. It's worth starting here, because those numbers give you the context for the customer prompts that follow.

N.B. Both prompts in this section depend on discount and markdown values being sent through to Ometria (or your equivalent platform) against your orders and order lines, so it's worth checking before you run them.

How do sale products perform against full price?

How much of your revenue comes from sale products, and whether that share sits in planned sale periods or runs across the year, tells you how much of your performance is being bought with margin. This prompt compares sale and full-price products on revenue, volume, and AOV, with a month-by-month view of how the mix shifts.

The prompt

For [brand name], please compare the performance of sale products vs full-price products for ecommerce orders over [time period, e.g. the last 12 months].

For each group, include: total revenue, total orders, average order value, and percentage share of total revenue and total orders. Then show the month-by-month trend of promotional vs full-price revenue share across the period.

Present the comparison as a side-by-side table, the monthly trend as a line chart, and add 3 to 5 bullet points summarizing how the mix has shifted.

What it returns

• Revenue, orders, and AOV for sale vs full-price products, with each group's share of the total
• A month-by-month chart of promotional revenue share, showing whether discounting sits in planned sale periods or runs throughout the year
• A written summary of how the mix has moved across the period

Variations to try

• What percentage of our revenue in [time period] came from discounted products?
• Show me promotional revenue share by store or region for [time period].
• Take our top 50 products by revenue and show how many units of each sold at full price against at a discount, so I can see whether anything selling well at full price is still being marked down.

💡 Look at the baseline between your sale periods rather than the spikes. Clearance events will always stand out on the chart, so the more telling number is where promotional revenue share sits in the months between them, and whether that floor has risen year on year.

How much of your revenue is going to markdown?

Discount applied at product level sits outside campaign reporting, so it's easy for CRM and trading to be working from different numbers. This prompt reads the discount recorded against your orders and line items and shows where it sits by month, category, and region.

The prompt

For [brand name], please analyze total discount amount and discount as a percentage of pre-discount order value for ecommerce orders over [time period, e.g. the last 24 months].

Show the trend month by month, then break the same period down by product category and by store or region, including total discount amount, pre-discount order value, and discount as a percentage of pre-discount order value for each.

Present the monthly trend as a line chart and the category and store breakdowns as sortable tables, with a brief written summary of where the discount amount is concentrated.

What it returns

• A month-by-month view of total discount amount and discount as a percentage of pre-discount order value
• A breakdown of the same figures by product category and by store or region
• A written summary identifying where the discount amount sits

Variations to try

• Compare discount as a percentage of pre-discount order value for [this year] against [last year].
• Which product categories have seen the largest increase in discount share over [time period]?
• Show me discount amount by month alongside total orders, so I can see whether heavier discounting coincided with higher volume.

Customer behavior and value

These three prompts move from what discounting costs to who it reaches. They look at how your customer base divides by promotional behavior, what a discount does to the basket, and whether customers acquired on an offer go on to buy differently from those who paid full price.

How does our customer base engage with discounts?

Some customers will only buy when something is reduced, others pay full price without hesitating, and plenty sit somewhere in between. This prompt splits your active base into those three groups and shows what each one is worth, which gives you the foundation for everything else in this list.

The prompt

For [brand name], please segment our active customer base by promotional behavior over [time period, e.g. the last 24 months].

Group customers into three sets: those whose orders were all discounted, those with a mix of discounted and full-price orders, and those who only purchased at full price. For each group, include: total customers and percentage of the base, total revenue and percentage share of revenue, average order value, average orders per customer, average lifetime revenue to date, and average time between orders.

Present as a side-by-side table and add 3 to 5 bullet points on how the groups differ in value.

What it returns

• The size of each group and how much revenue each one contributes
• AOV, order frequency, and lifetime revenue for promotion-only, mixed, and full-price customers
• A written summary of the value gap between the three groups

Variations to try

• Show me how the three groups differ by time of year, so I can see whether the promotion-only group grows around known sale periods or stays consistent.
• Which lifecycle stages contain the highest share of promotion-only customers?
• What share of revenue in [time period] came from customers who have never placed a full-price order?

💡 As a general rule, you should reserve your strongest incentives for your lowest-intent audiences and use lighter offers, early access, or product content where intent is already high. Sizing the three groups tells you how much of your base each of those approaches should cover.

How do discounts impact basket size?

A discount that gets a customer to add another item is doing something quite different from one that just takes money off the same basket. This prompt looks at what changes inside the order, comparing items per order and value per item across discounted and full-price orders, so you can see which of the two is happening in your data.

The prompt

For [brand name], please compare discounted and full-price orders over [time period].

For each group, include: total orders, average number of items per order, average value per item, and average order value. Then show the same comparison broken down by month and by product category.

Present the headline comparison as a side-by-side table, the monthly view as a line chart, and add 3 to 5 bullet points on how basket composition differs between the two groups.

What it returns

• Items per order, value per item, and order value for discounted and full-price orders
• A monthly view showing whether the gap between the two widens around sale periods
• A category-level breakdown of where discounting changes basket behavior most

Variations to try

• Compare items per order for discounted and full-price orders in [known sale period] against the rest of the year.
• Which product categories show the largest gap in basket size between discounted and full-price orders?
• Show me average value per item for discounted and full-price orders by [time period].

💡 The metric to look at first is items per order. If discounted orders contain more items than full-price ones, the promotion is increasing basket size and the margin trade may well be worth making. If the item count is flat and only the value per item has dropped, there's room for the offer to do more work, for example through a spend threshold or a bundle rather than a straight reduction.

With these prompts in hand, you'll be able to get a clear read on your discount marketing strategy in minutes, covering how your customers are responding to discounts and what those discounts are costing you in margin.

Want to put Deep Insights Agent to work on your own data? Book a demo and we'll show you what's possible.