Online live video sales management system
A sales administration system that manages online sales via overlays on a video stream. Customers can comment on or respond to the video stream, and may order products using simple phrases such as “Sold”. The system can stream video with overlays to multiple clients simultaneously, including websites and mobile apps. Comments from all clients are received, aggregated, processed for orders, and displayed to the sales administrator. The system is coupled to a merchant's operational data, providing immediate visibility to changes in inventory and orders. Video overlays may be updated automatically by the system as this operational data changes, for example highlighting products with low stock. The system may analyze customer comments and orders both during the video stream and after the sales session is complete, and may generate recommendations to improve sales effectiveness. For backorders, customers may preauthorize charges to be at the front of the queue for shipment.
1 . A sales administration system for generating a video overlay for a video stream, the sales administration system comprising:
a memory that stores a plurality of product images corresponding to a plurality of products; and
a processor configured to execute instructions stored in the memory, the instructions, when executed by the processor, configuring the processor to:
analyze a video frame from among the video stream to identify a product from among the plurality of products,
determine a location of the product within the video frame,
retrieve a product image that corresponds to the product from among the plurality of product images,
generate the video overlay having the product image, information about the product, and a textual response template for purchasing the product, and
position the video overlay proximal to or over the product image within the video frame based on the product location.
2 . The sales administration system of claim 1 , wherein the instructions, when executed by the processor, configure the processor to analyze the video frame to identify a visible product in the video frame and identify the product to be the visible product in the video frame.
3 . The sales administration system of claim 1 , wherein the instructions, when executed by the processor, configure the processor to analyze the video frame using a neural network to identify the product in the video frame.
4 . The sales administration system of claim 1 , wherein the information about the product comprises a product identifier, a price, the product image, and a list of product options.
5 . The sales administration system of claim 1 , wherein the response template comprises a keyword or key phrase, a product identifier, a quantity, or a product option of said list of product options.
6 . The sales administration system of claim 1 , wherein the instructions, when executed by the processor, further configure the processor to:
track movement of the product across subsequent video frames relative to the initial video frame; and
position the video overlay proximal to or over the product image within each subsequent video frame based on the movement of the product.
7 . The sales administration system of claim 2 , wherein the instructions, when executed by the processor, further configure the processor to analyze the video frame using a neural network to identify the visible product in the video frame.
8 . A method for generating a video overlay for a video stream, the method comprising:
storing a plurality of product images corresponding to a plurality of products in a memory;
analyzing a video frame from among the video stream to identify a product from among the plurality of products;
determining a location of the product within the video frame;
retrieving a product image that corresponds to the product from among the plurality of product images;
generating the video overlay having the product image, information about the product, and a textual response template for purchasing the product; and
positioning the video overlay proximal to or over the product image within the video frame based on the product location.
9 . The method of claim 8 , wherein the analyzing comprises:
analyzing the video frame to identify a visible product in the video frame; and
identifying the product to be the visible product in the video frame.
10 . The method of claim 8 , wherein the analyzing comprises:
analyzing the video frame using a neural network to identify the product in the video frame.
11 . The method of claim 8 , wherein the information about the product comprises a product identifier, a price, the product image, and a list of product options.
12 . The method of claim 8 , wherein the response template comprises a keyword or key phrase, a product identifier, a quantity, or a product option of the list of product options.
13 . The method of claim 8 , further comprising:
tracking movement of the product across subsequent video frames relative to the initial video frame; and
positioning the video overlay proximal to or over the product image within each subsequent video frame based on the movement of the product.
14 . The method of claim 9 , wherein the analyzing the video frame to identify the visible product comprises analyzing the video frame using a neural network to identify the visible product in the video frame.
15 . A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to:
analyze a video frame from among a video stream to identify a product from among a plurality of products;
determine a location of the product within the video frame;
retrieve a product image that corresponds to the product from among the plurality of product images;
generate a video overlay having the product image, information about the product, and a textual response template for purchasing the product; and
position the video overlay proximal to or over the product image within the video frame based on the product location.
16 . The non-transitory computer-readable medium of claim 15 , wherein the instructions cause the processor to:
analyze the video frame to identify a visible product in the video frame; and
identify the product to be the visible product in the video frame.
17 . The non-transitory computer-readable medium of claim 15 , wherein the instructions cause the processor to analyze the video frame using a neural network to identify the product in the video frame.
18 . The non-transitory computer-readable medium of claim 15 , wherein the response template comprises a keyword or key phrase, a product identifier, a quantity, or a product option of the list of product options.
19 . The non-transitory computer-readable medium of claim 15 , wherein the instructions further cause the processor to:
track movement of the product across subsequent video frames relative to the initial video frame; and
position the video overlay proximal to or over the product image within each subsequent video frame based on the movement of the product.
20 . The non-transitory computer-readable medium of claim 16 , wherein the instructions cause the processor to analyze the video frame using a neural network to identify the visible product in the video frame.