Abstract: When a user accesses content describing an item able to be purchased using a first website, and the user subsequently purchases the item using a second website within a threshold length of time of the access to the content, the purchase may be inferred to have been induced by the content, even though the content would normally direct the user to the first website. To correlate purchases across different websites, data identifying the user's device or a user account and data identifying the item is acquired when the content is accessed by the user. Subsequently, additional data is acquired when the item is purchased. If the additional data corresponds to the previous data, this may be used to determine that the same user has purchased the same item described by the content that was accessed.

Abstract: Techniques for selecting a delivery method based on sensor data of an automatic replenishment device (ARD) are described. In an example, a computer system is communicatively coupled with the ARD and receives the sensor data from the ARD. The sensor data is generated by a sensor of the ARD and indicates an amount of an item that is stored by the ARD. The computer system maintains, in a profile, a consumption rate based on the sensor data. The computer system determines that the amount of the item is less than a threshold amount based on the consumption rate and, based on this amount, identifies available delivery methods for a delivery of a replacement amount of the item. The computer system selects one of the delivery methods based on the consumption rate and causes the delivery of the replacement amount of the item based on the selected delivery method.

Abstract: The disclosed techniques relate to improving sensor data collection/reporting at an automatic replenishment device (ARD). For example, sensor data associated with the item stored at an ARD may be collected. The ARD can calculate a change in the amount of the item stored by the ARD from the sensor data. If the change is over a reporting threshold, the sensor data (or the amount of change) may be transmitted to a back-end system. Reporting may not occur if additional sensor data (e.g., accelerometer data and/or images captured at the ARD) is not commensurate with a user interaction. If the difference is under the reporting threshold or unlikely due to user interaction, the ARD may refrain from transmitting the sensor data. These techniques conserves power at the ARD, relieves the back-end system from needless processing, and enables more accurate indications of the amount of the item stored by the ARD.

Abstract: Techniques for selecting a delivery method based on sensor data of an automatic replenishment device (ARD) are described. In an example, a computer system is communicatively coupled with the ARD and receives the sensor data from the ARD. The sensor data is generated by a sensor of the ARD and indicates an amount of an item that is stored by the ARD. The computer system maintains, in a profile, a consumption rate based on the sensor data. The computer system determines that the amount of the item is less than a threshold amount based on the consumption rate and, based on this amount, identifies available delivery methods for a delivery of a replacement amount of the item. The computer system selects one of the delivery methods based on the consumption rate and causes the delivery of the replacement amount of the item based on the selected delivery method.

Abstract: A reorder device having an item tag reader and one or more sensors may facilitate the automatic reordering of an item for a customer. The item may be associated with an item tag that is placed on (or near) the reorder device. The reorder device may detect the item tag and transmit such data to an entity device, which may associate the reorder device and the item. The reorder device may periodically obtain weight measurements of the item that is placed on the reorder device, as well as a timestamp that indicates a time at which the weight measurements are obtained. Such data may be provided to the entity device, and an entity associated with the entity device may determine whether weight measurement satisfies an item threshold value. If so, additional inventory of the item may be ordered and delivered to the customer, without additional input from the customer.

Abstract: Techniques for consolidation of items based on sensor data is provided. For example, an amount of an item may be identified using an electronic sensor associated with an automatic replenishment device. When the amount of the item is greater than a first reorder threshold, the item may not be ordered on behalf of the user until the amount of the item reaches the first reorder threshold. However, when an order for a different item is placed, either by accessing a network document to proactively order the different item or through a reorder schedule based on a consumption rate of the different item, the amount of the first item may be compared with a second reorder threshold. When the amount of the first item is less than the second reorder threshold, a notification may be provided to the user for additionally ordering the second item with the first item.

Abstract: Techniques for using sensor data from an auto replenishment device (ARD) to generate recommendations for ordering other items not associated with an ARD, updating a quantity of an already ordered item, or updating a time period associated with delivery of the ordered items. In embodiments, order history information associated with a user profile for a plurality of items may be maintained. Consumption data from a sensor of an ARD may be received that indicates a current amount of the item. A subset of items may be determined based at least in part on the consumption data and the order history information. In accordance with at least one embodiment, instructions for delivering the subset of items to a location associated with the user profile may be executed.

Abstract: Techniques for using sensor data from an auto replenishment device (ARD) to generate recommendations for ordering other items not associated with an ARD, updating a quantity of an already ordered item, or updating a time period associated with delivery of the ordered items. In embodiments, order history information associated with a user profile for a plurality of items may be maintained. Consumption data from a sensor of an ARD may be received that indicates a current amount of the item. A subset of items may be determined based at least in part on the consumption data and the order history information. In accordance with at least one embodiment, instructions for delivering the subset of items to a location associated with the user profile may be executed.

Abstract: Techniques for consolidation of items based on sensor data is provided. For example, an amount of an item may be identified using an electronic sensor associated with an automatic replenishment device. When the amount of the item is greater than a first reorder threshold, the item may not be ordered on behalf of the user until the amount of the item reaches the first reorder threshold. However, when an order for a different item is placed, either by accessing a network document to proactively order the different item or through a reorder schedule based on a consumption rate of the different item, the amount of the first item may be compared with a second reorder threshold. When the amount of the first item is less than the second reorder threshold, a notification may be provided to the user for additionally ordering the second item with the first item.