Most promotions are designed from experience and measured afterwards. This system reverses that order: it analyses shopping baskets from real purchase data to find what is picked up together, and analyses price sensitivity to find the discount level that produces the highest total sales. The result is a set of bundle and segment-specific promotions with reasoning behind them before any budget is committed. ThinkVerse AI built it as a Marketing Analysis system.
Built for
Marketing and retail teams that run promotions in cycles and want the reasoning before the budget is spent, rather than a read-out afterwards
What the shopping basket can tell you
The main dashboard breaks sales down by customer status (all customers, customers still buying, and customers who have stopped) alongside purchase-behaviour trends and the most frequently bought products. From there it drills into co-purchase rates, which ask a different question: when a customer picks up one product, what tends to go into the basket with it. A detail table lists the product pairs that most often share a basket (bread with butter and milk cartons, for instance), which is the starting point for designing bundle or combo promotions and for adjusting margins directly.
- Sales split by all customers, still-buying customers and lapsed customers
- Purchase-behaviour trends and the most frequently bought products
- Co-purchase rate for each product pair
- Table of product pairs that share a basket, for bundles or combos
Finding the best discount point before the budget is spent
The part that answers the most expensive question is the price-sensitivity analysis. The system measures how sales change across different discount levels and promotion codes, then marks the point that produces the highest total sales on the same chart as the forecast figure. It also reviews promotions already run, so the next campaign is built on what actually worked rather than starting from scratch every quarter. Discounting becomes a decision backed by numbers, not a guess at how much is enough.
Promotions split by customer segment
Once it is clear what sells together and how much to discount, the next step is choosing who to send it to. Promotions can be designed per segment: customers likely to buy again, versus new customers who need a nudge towards a first purchase, two groups that should receive different offers. The coupons and creative for each campaign are tied back to the analysis they started from, and can be connected out to other systems in the organisation, so the results do not stop at the dashboard.
- Promotions designed per customer segment
- Separate offers for existing and new customers
- Coupons and campaign creative tied back to the analysis
- Connects to other systems in the organisation