A non-transitory machine-readable medium that stores instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
- determining a price curve for an item, the price curve indicating a probability of sale of the item at each of a plurality of prices;
- determining a preference for a seller account, based on completed listings of items for sale by the seller account, the preference indicating a cost of time and the determining of the preference for the seller account comprises providing the completed listings of items for sale by the seller account as training data to a machine learning algorithm, the machine learning algorithm being trained with the training data to determine preferences for the seller account;
- based on the price curve for the item and the preference for the seller account, determining by the trained machine learning algorithm, a utility curve for the item and the seller account, the utility curve indicating sales volume for a seller associated with the seller account as a function of price of the item, the utility curve including a price threshold above which a likelihood of the seller selling the item at a first price above the price threshold decreases and the likelihood of the seller selling the item at a second price below the price threshold increases;
- based on the utility curve for the item and the seller account, using the trained machine learning algorithm to present causing presentation of a first user interface on a display associated with the seller account, the first user interface comprising:
- a slider that corresponds to the utility curve;
- a control configurable to be moved by the seller along the slider;
- a price range that includes a recommended price for the item to be sold from the seller account shown on the slider such that a seller moves the controller along the price range;
- an option to generate a listing of the item for sale at the recommended price based on the seller using the control to set the item at one of the first price or the second price;
- and
- an amount of increased likelihood of selling the item at the first price in comparison to the second price, the control being configurable to move to different positions along the slider where the amount of increased likelihood of selling the item at the first price in comparison to the second price varies based on the different positions of the control along the slider and the trained machine learning algorithm is used to provide the recommended price based on the utility curve;
- receiving an input corresponding to moving the control to one of the different positions along the slider;
- and
- presenting a second user interface in response to receiving the input, the second user interface displaying an updated likelihood of the seller selling the item in response to the movement of the control to the one of the different positions along the slider.