E-commerce pricing optimization
Demand-based multi-pricer
Store level profit-revenue optimization model
Demand-based multi-pricer benefits
The demand-based multi-pricer is an AI pricing automation model that helps you navigate the profit revenue balance on a macro level for each product category.
How can you set revenue and profit goals for your whole store automatically?
With only several categories making revenue, you could be perceived as an affordable merchant. The rest of your goods can be priced for profit or a mix of both.
Check the 3-step process to optimizing prices in the entire store
1. Demand forecasting
Get advantage of your demand data and analyze price elasticity to:
- Utilize store orders
- Sort elastic and inelastic products
- Find price-quantity relation
2. Revenue vs. profit calculation
Decide on the desired levels of revenue and profit. Maximize revenue and profit by optimizing all retail prices at once.
- Get optimal prices drawn by the model
- Re-assess profit and revenue limits
- Take into account the strategic goals
3. Apply business rules
Get the next best prices from the model and make sure all business rules are taken into account: margin guards, competition levels, roundings, shipping costs …
- Set your boundaries and constraints
- Follow the short term price elasticity changes
- Iterate and monitor the progress
The demand-based multi-pricer AI model can help businesses optimize all retail prices simultaneously while blending revenue and profit so that companies can reach their ultimate goals and KPIs. Some of the benefits of that model include but are not limited to:
computing the revenue vs. profit tradeoff
setting targets for revenue and profit on a category level
getting fresh prices based on the demand changes
including business rules like margin guards, competition, and rounding
Demand-Based Multipricer Model FAQs
The Demand-Based Multipricer is an AI-driven pricing model designed to optimize e-commerce prices by analyzing the trade-offs between revenue and profit.It evaluates a range of possible price points to determine the optimal set of prices that align with your specific business goals, such as maximizing transaction volume or increasing margin.
The model allows you to position products strategically based on your goals. For example, you can configure the model to maximize revenue for high-traffic "traffic builders" (like TV sets or basic jeans) to acquire customers. Conversely, you can maximize profit for niche or complementary items (like accessories or gift sets) where higher margins are sustainable.
Efficient revenue-profit points form a curve representing the best possible trade-offs for a product. These points identify price scenarios where you cannot gain more revenue without sacrificing profit, and vice versa. The model constructs this by evaluating numerous price points (e.g., 20 intervals between $6 and $8) to find the optimal balance.
At a minimum, the model requires historical sales data at the variant or product level, including units sold, timestamps, and price points. Crucially, the model creates accurate predictions only for products that have historical sales data on at least three different price points.
Yes. If COGS data is unavailable, the model can still function by optimizing for Maximum Revenue only. However, to optimize for Profit or a Revenue-Profit Mix, COGS data is essential to calculate margins accurately.
For products with sparse data, the model can incorporate optional clickstream or page visit data. This acts as a proxy for customer interest (intent signals) to enhance demand modeling. Additionally, if a specific SKU lacks data, the model can approximate demand using the category average.
The model distinguishes between seasonal and non-seasonal items. For seasonal products (e.g., winter jackets), it uses historical sales data from the equivalent season in the previous year. For non-seasonal products, it uses sales data from the preceding calendar month or period.
Yes. During the demand forecasting setup, you can include additional data points such as retail-related holidays, special events, and distinctions between weekdays and weekends to refine the demand prediction.
You can implement Business Constraints such as Min-Max price boundaries. Furthermore, you can set Margin Guardrails, which ensure that any recommended price maintains a minimum profit margin over COGS, preventing the model from suggesting prices that lead to losses.
Yes. The model can import competition data points. You can use competitor minimum, maximum, or average prices to set the lower or upper bounds (business constraints) for your own pricing strategy.
You can configure the model with three primary objectives:
- Optimal Revenue: Maximizes expected gross sales (optimize 100% for revenue)
- Optimal Profit: Maximizes margin (optimize 100% for profit).
- Profit-Revenue Mix: Uses a tunable coefficient to balance both, allowing for a strategic middle ground (optimize for 50% profit and 50% revenue, or optimize for 75% profit and 25% revenue)
Yes. You can select specific segments of your portfolio—such as a Category, Brand, Vendor, or Collection—and apply unique optimization goals to them. That's the essence of the model - to strategically position segments for profit, revenue or mix.
Price updates are determined by the schedule you set. You can select a specific period for re-pricing, at which point the model will re-calculate demand based on the latest data and suggest new prices. It depends on the sales volume, the industry and buyers behaviour. Usually customers change prices with this model once per day.