Patent Application
20080059281
Reference will now be made in detail to various and alternative exemplary embodiments and to the accompanying drawings, with like numerals representing substantially identical structural elements. Each example is provided by way of explanation, and not as a limitation. In fact, it will be apparent to those skilled in the art that modifications and variations can be made without departing from the scope or spirit of the disclosure and claims. For instance, features illustrated or described as part of one embodiment may be used on another embodiment to yield a still further embodiment. Thus, it is intended that the instant disclosure includes modifications and variations as come within the scope of the appended claims and their equivalents.
At step 50, information that has been gathered and analyzed may be presented in a human-readable form. One of skill in the art will recognize that data may be presented long after it is gathered and/or analyzed. Furthermore, not all persons providing data will necessarily view the results, nor will all persons reviewing results necessarily have provided data. As will be discussed in further detail below, the information may be provided to entities including the consumer who provides qualitative data, to other consumers, to product designers, to marketing, advertising, and sales personnel, and to other persons in a wide variety of contexts, depending upon particular data needs and applications.
The presently-disclosed subject matter may be implemented by any suitably configured computer system or systems running a survey application. For example, consumer interaction such as presentation of images to a consumer and collecting data may be implemented using a web-based survey application provided from a server with supporting scripts such as Javascript. Alternatively, some or all of the consumer interaction may take place using a standalone survey application, such as a standalone executable file. The survey application may include one or more component or metafiles that direct the operation of other applications running on a computing device.
The survey application may include components downloaded to a computing device over the Internet or another network, components provided via media, such as a CD or DVD-ROM for example, or components that operate over a network environment. The survey application may be implemented entirely on a remote computer, entirely on a consumer computer, or partially on one or more computers.
The consumer may use any suitable computing device, including, but not limited to, desktop, laptop, tablet, and network PCs, cellular telephones, and/or personal digital assistants (PDAs), for example.
Data collected from the consumer can include data defining region selection areas associated with each image presented to the consumer. Furthermore, qualitative data is collected. Qualitative data includes any subjective information provided by the consumer, such as intensity rankings, descriptive text, multiple-choice selections, and freehand drawings, for example. Other useful data that may also be collected may include consumer-specific data such as consumer identification, location or demographic data, and time of the survey. Survey metadata such as order of selection, number and order of images and/or products, the amount of time each image or region was considered, and other information describing the survey process may also be collected.
The survey application may provide all such data to one or more product quality database(s), with the qualitative data associated to particular region(s) (if any) of each particular product selected by the consumer. For example, the data may be associated with coordinates corresponding to portions of the product, such as vector coordinates. In alternative embodiments, qualitative data may be associated with particular coordinates of one or more images of the product. In such embodiments, regions of the product may be defined as certain areas within one or more images of the product. The qualitative data itself may be stored partially or entirely in graphical form, such as in the form of an image.
Data collected from consumer may be stored in one or more product quality databases. Such databases may be implemented using one or more computers, such as servers, running any suitable database program and configured to receive, either directly or indirectly, the consumer data from survey applications. For instance, the product quality database may be supported by a first server configured to receive data from a second server, with the second server configured to interact with consumers via one or more survey applications. Alternatively, a single server could be used. The collected data may be stored in a form so that qualitative data may be accessed based on input specifying a region of a product. For instance, qualitative data may be associated with a particular region or particular regions of one or more images depicting a product. Alternatively, qualitative data may be associated with other data defining portions of a product, for example, physical coordinates of the product itself. In such a case, the physical coordinates may be determined through analysis of region selections in the images, for example. Of course, one of skill in the art will recognize that the database may further include data not associated with particular portions of images or regions of products.
Analysis of the collected data may be performed on the same server(s) housing the database or supporting consumer interaction, or may be implemented using further computing devices. For instance, the product quality data may be downloaded to a computer running appropriate analysis software; alternatively, some or all analysis may be provided as part of the database functionality.
The selection data may be obtained in any suitable manner. For instance, the consumer may select an area by freehand drawing, highlighting, or clicking on areas of the image with a mouse, tablet, or touchscreen interface, for example. The regions may be predefined or may be defined by the consumer. For example, the image could be divided into regular or irregular shapes, with each shape defining a region, with consumer input cross-referenced to the predefined regions via a grid or other coordinate system. Alternatively or additionally, the areas could be defined based on pinpointing actual areas selected by the consumer. As a further alternative, predefined regions could be explicitly presented to the consumer for selection or non-selection, for example, by highlighting each of a plurality of regions in sequence and prompting the consumer for a response.
For instance, the survey software may be configured to generate a qualitative data input window 101 every time a consumer graphically selects an area 140 in the image 110. As another example, the consumer may first select one or more regions 140, with the software configured to present a plurality of qualitative input windows 101 simultaneously or in succession while indicating which window is associated with which graphical selection. Furthermore, qualitative data may be received even if a consumer does not select any regions of an image.
Although exemplary window 101 includes text area 160 and attribute intensity selection rating menu 150, other types of qualitative data may be obtained by other means. For instance, rather than text input, the software could be configured to recognize speech or consumer handwriting and convert the same to text or other machine-recognizable form. Qualitative data could be input in freeform or may be obtained by providing one or more choices of, for example, key words, discomfort intensity levels (or other numerically-indicated attributes), or by providing graphics that could be manipulated to indicate intensity, such as a clickable thermometer or a sliding level indicator.
Although a single image and graphical selection are shown, the survey software may be configured to provide a plurality of different images, including images of the same product and/or user in different views, images of multiple different users using the same product, and images of multiple different users using different products. Each image may depict one or more users and one or more products. Once data has been gathered, the graphical selection data and qualitative data are compiled using a variety of statistical analysis techniques. These techniques may provide a wealth of usable data for a number of different applications.
An example of such data is provided in
The compiled data may be of great use to a variety of personnel involved with the manufacture and sale of the consumer product 130. For instance, regions 170A and 170B could be further analyzed and could become the subject of a product redesign.
Image processing or morphological operations may be employed to enhance or otherwise alter the image(s) before, during, and/or after any part of the survey process. For example, images may be merged, cleaned, blurred, or otherwise enhanced. Image processing operations may be used to provide additional or more useful data from which to extract features and make product recommendations. For example, this may include binary or grayscale analysis, frequency analysis, and more complicated densitometry. Similar techniques may be used to alter, analyze, and process qualitative data for instances in which the qualitative data itself is stored in graphical form. For example, region selections may be stored as images and the images accumulated to determine regions selected by multiple consumers.
The various images 210-1910 may be presented to one or more consumers in the same manner as discussed above in conjunction with image 110. For example, each of a plurality of consumer may be directed to graphically select one or more areas of interest and provide comments and a discomfort rating pertaining to that area of interest. The graphical selections and qualitative data provided by each consumer may be cross-referenced to determine areas that were selected by a plurality of consumers and to indicate a relative measure of intensity. For example, a metric may be applied to the intensity rating provided by each consumer for a particular area to weigh or normalize the intensity data.
Regions of interest may be defined, for example, as areas for which the compiled data exceeds a threshold value. Alternatively, regions of interest may be explicitly defined and data corresponding to those regions may be correlated and displayed. For instance, in
For example,
Results such as those shown in
Qualitative data may be obtained for any suitable product attribute. For instance, qualitative data may describe the consumer's perceived feelings, perceptions, impressions, or other thoughts regarding the product. Such attributes may include perceived comfort, discomfort, softness, roughness, fit, tightness, looseness, linearity, symmetry, sags/droops, gaps, physical attributes of the product, sturdiness, resiliency, smoothness/wrinkles, agreeability/disagreeability of colors, graphics, images, shapes, product layout, appearance, etc. The attribute or attributes may be measured, for example, by a numerical value indicating perceived intensity. However, other suitable metrics may be employed.
The compiled data may be analyzed in other ways and presented in non-graphical formats to provide still further advantages to consumer product manufacturers. For instance,
For example, the chart indicates that a match was found in panelist no. 1's textual comment for the key words “saggy” and “full.” One of skill in the art will note that the actual text of “sagging” was cross-referenced to “saggy” by use of, for example, software analysis routines. In this example, panelist no. 5 provided an attribute rating but no text. Furthermore, this example shows that panelist no. 7 viewed two different images (front and back) of the product.
Correlation and analysis of the qualitative data provided by consumers may include a key word search, which may be based, for example, on a list of key words provided by the survey takers. Furthermore, the key word list may be generated dynamically by analyzing comments for a particular region by a plurality of consumers and extracting words that are used at or above a given frequency as is known in the art.
Alternatively, the same images provided to consumers may be presented to internal personnel, such as product designers, engineers, sales personnel, or others, and such internal personnel may be prompted to provide descriptive words using internal terminology. For example, consumers may describe a particular region of a garment as “wrinkly,” while a product design engineer may use a different term, such as “creped” for the same region. Other internal data may include internal terminology for product regions, component names, part numbers, product and/or part measurements, and material attributes, such as composition, for example. Qualitative consumer data from the database may be correlated with internal data to provide more closely-tailored suggestions or comments to designers. Furthermore, the internal data may be used as the basis for sorting and analysis of the data provided by consumers. The internal data may be gathered in a manner the same as or similar to the data gathered from consumers.
Correlation may include analysis and processing of graphical data contained within the images themselves. In one embodiment, graphical data could be evaluated using image analysis and processing and then correlated with consumer-provided data. For example, consumer selections could be scored against light and dark areas of an image to determine the influence the contrast or composition of the image has on responses.
As another example, consumer selections indicating degrees of one or more attributes, “wrinkliness,” for example, could be cross-referenced with areas having a particular pattern of dark and light pixels. Varying degrees of “wrinkliness” could then be scored throughout the image (and in other images) based on identifying pixel attributes with consumer perceptions. As a further example, quantitative measurements, such as sizes of gaps, product areas, etc. could be correlated to consumer perceptions.
For instance, for a wearable product, images showing varying gaps between the product and the user could be evaluated to determine how big the gap would have to grow before consumers perceived a problem. The gap size could be based on quantitative measurements provided to the system and cross-referenced to the image. Alternatively, the gap size may be measured by analyzing the image itself.
Accordingly, the system may include a toolset for extracting qualitative data about the product(s), user(s), or other subjects depicted in the image, and such qualitative data could then be analyzed and correlated to qualitative data provided by system users. For instance, such measurements may be based on 2D or 3D analysis of one or more images of the product. As noted above, the system may additionally or alternatively provide for the input of measured physical parameters, such as physical measurements of the product taken at the time images of the product are produced.
Correlation may include analyzing qualitative data including suggested changes or improvements to the product. For example, consumers may be presented with various styling choices or feature combinations for an automobile and may be prompted to choose the most desirable. Alternatively, the consumers may be prompted to provide suggested changes in color schemes, for instance.
A product quality database may, as noted above, provide a wide variety of avenues for improvement of product design features as well as improvements in tailoring marketing and advertising strategy. Furthermore, such a database may be useful as a component in a computer-based system that allows for both product marketing as well as collection of qualitative data while also providing purchase guidance to consumers.
For instance, a product quality database could be assembled in accordance with the subject matter discussed above, such that the product quality database includes data identifying at least two products and qualitative data about each product, with at least some of the qualitative data associated with particular regions or parts of each product. As noted above, the qualitative data could be associated with particular parts of each product by way of particular areas of an image or images of such products, or by way of other means such as vector coordinates. The product quality database could include data pertaining to a wide variety of products across multiple fields and multiple manufacturers. Furthermore, the database could include one or more images of each product. The qualitative data in the product quality database may be obtained from consumers. However, the data may be obtained in whole or in part from other sources, such as from personnel associated with providing the product.
A purchase guidance system may include one or more computers configured to prompt a consumer to provide product attribute data. The product attribute data may comprise any suitable identification of a product, such as the product name, the product brand name, a brand family which includes the product, or an inventory or other identification number, for example. Furthermore, product identification data could include user-specific data, such as sizes or other user measurements. The product attribute data may be only a rough indicator of desired product traits, which could be especially useful if a name or other designation is unknown. Based on the product attribute data, the system could be configured to correlate the attribute data with data associated with products and stored in the product quality database to determine one or more purchase candidate products.
For example, a user could select “diapers” and provide a size or range of sizes. Based on the data provided by the consumer, the database may return one or more products matching the criteria, e.g., a variety of different types of diapers in matching or close-to-matching sizes. The system can provide further information about these returned products (referred to as “purchase candidate products” herein), such as one or more images of each purchase candidate product, and other purchase guidance data associated with the product. In addition to the images, the purchase guidance data may include, for example, product price, use and care instructions, or other information about the product. For example, the purchase guidance data may include information obtained by correlating qualitative data provided by other consumers in association with particular regions of the product.
Furthermore, the purchase guidance system may also prompt the consumers to select a region of a purchase candidate product and provide associated qualitative data. Such data may then be added to the product quality database for further analysis as discussed herein.
The system may be configured to recognize particular consumers and identify qualitative data with the particular consumer providing the data. For instance, the system may prompt the consumer for identification data prior to providing purchase guidance data. Providing purchase guidance data can include accessing qualitative data provided by the particular consumer in the past and using that data as a basis for providing purchase recommendations or purchasing guidance data.
For example, if the consumer had indicated a certain area on a diaper of first size as bulky, tight, or otherwise undesirable, such data may be added to the product quality database associated with that particular diaper and stored for later use. If the same consumer later requests a recommendation for a diaper of a second size, the system could consider that particular consumer's dislike for aspects of the diaper of the first size when making the purchase recommendation. For instance, assume the consumer indicated the leg area of a certain style of diaper to appear tight when shopping for a diaper of a first size. Later, if the consumer requests a diaper of a larger size (for instance, to accommodate a growing infant), the system may exclude diapers of the non-preferred style.
One of ordinary skill in the art will recognize that many statistical techniques, including discriminant analysis, clustering, supervised learning algorithms, and the like exist to classify consumers on the basis of their qualitative data and to aid in the recommendation process.
The system may even be configured extrapolate preferences from one style of product to another based on the qualitative data. Using the above example, if the non-preferred region of a particular style of diaper is correlated to a certain component or material of the diaper, such as a particular liner type, the system may exclude or provide lower rankings to diapers of other styles using that same component or material.
As a further example, the product quality database may include data pertaining to clothing products. Such data could include qualitative data indicating certain preferred styles in casual wear clothing as provided by a particular consumer. Such data could be used when the same consumer is selecting business clothing or swimwear, for example, such as preferred color combinations and the like. The system may include factors that weigh how “close” to product categories are. For instance, casual wear and business clothing may be considered closer than business clothing and swimwear, while clothing and any sort of tool or personal care product would be considered not to be close. Nonetheless, even seemingly-disparate products may share attributes for which consumer preferences may be considered; for example, a consumer may prefer certain color schemes in home furnishings that complement his or her clothing selections.
The purchase guidance system may be configured to create profiles of consumers based on recommended products, previously-provided purchase guidance data, and other consumer data. For example, as noted above, a purchase guidance system may be configured to track consumer preferences as to clothing style by generating a profile that includes various preferred clothing styles and combinations. The system may further be configured to track the consumer's style over time and cross-reference it to other consumer profiles and demographic data. The system may also be configured to aggregate profiles for a plurality of consumers to track trends across demographic groups, such as ages, income levels, locations, and the like. The system may also recommend additional items based on selection or other feedback provided during the recommendation process.
Furthermore, after one or more products have been recommended, the system may prompt the consumer for feedback as to whether the guidance data is accurate. Using the example noted above, after recommending a particular diaper, the system may inquire as to whether the consumer plans to purchase the recommended diaper. The feedback may be obtained at a later time. For example, assuming the consumer purchases diapers a week later, the system may inquire as to whether the recommended diaper was a good buy or request input as to where the recommendation was inaccurate. By “remembering” consumers, the system may further be capable of obtaining long-range data about product use and changing consumer perceptions of the product as it is used.
Consumer feedback may be used to fine-tune software routines, algorithms, and other components used in implementing the purchase guidance system. For example, the feedback may be used to train neural networks or expert systems used in generating the purchase guidance. The feedback may be used to customize routines for individual consumers or groups of consumers. For example, if feedback across a wide variety of system users indicates bad recommendations, the algorithms may be altered and/or the problem may be brought to the attention of human personnel.
Consumer profiles may be accessed by the system as part of making purchase recommendations or otherwise providing purchase guidance data. The profiles may also be analyzed individually and/or in aggregate to extrapolate consumer trends. For instance, if the profiles include location data, the profiles may be sorted and otherwise analyzed on the basis of location. As noted below, the purchase guidance system may be implemented to operate in a retail environment. The profiles could be analyzed and the data provided to an entity or entities responsible to the retail environment, and may be correlated with data collected from the retail entities, for example, purchase data.
The product selection guidance system may be implemented using one or more computer systems and databases, for example, using one or more computer servers with access to the product quality database or databases. The consumer who desires product selection guidance could then access the server by way of a client computing device. The client device may comprise a desktop computer, a laptop computer, a tablet computer, a network computer, a personnel digital assistant (PDA), a mobile telephone, or any other capable device. For example, the purchase guidance system may be incorporated into an e-commerce site accessed using the client device via the internet.
Alternatively, the consumer may access the purchase guidance system by way of a client device located, for example, at a retail location. For example, the client device may be implemented as a kiosk including an appropriate network connection to the purchase guidance server and/or other suitable connections for accessing the purchase guidance database. The kiosk may be located at the point of purchase or in an area or areas of the retail location at which consumers are confronted with a choice in products. The kiosk may be configured to obtain data from consumers, for example, by keyboard, mouse, touchpad, or other input means. For instance, the kiosk may include a barcode and/or RFID or other scanning device to obtain information from indicia on an actual product located in the store. The kiosk may then access the product quality database and provide purchase guidance data based on the indicia.
For example, the consumer could choose a roll of paper towels and scan the barcode or RFID tag associated with the roll of paper towels. The purchase guidance system could then provide information about the particular type of paper towels indicated, as well as competing types indicated to have similar characteristics. The consumer could identify himself to the system, for example by scanning a shopping loyalty card or other identification, and the system could access stored preference attributes and/or stored profiles for that consumer and further refine the recommendation.
The client device may be part of a vending machine or other delivery system configured to physically present products to the consumer upon purchase. For example, a vending machine may include a touch-screen panel interfaced to a server running a purchase recommendation application. The consumer may interact with the purchase recommendation system to determine which product best suits his or her needs, and upon receiving a recommendation, may complete the purchase transaction with the vending machine.
The server(s) and client device(s) may be implemented as part of an e-commerce system. For example, an online store may be maintained using one or more servers to present an online storefront to consumers using client devices such as PCs. The online store may be further configured to access the purchase guidance system as part of the purchase process. Alternatively, the purchase guidance system may be accessed by client devices and provide links to the online store to purchase the recommended item(s).
Thus far, the present disclosure has provided examples of use of a system in conjunction with consumer products, such as diapers, paper towels, and the like. However, one of ordinary skill in the art will recognize that the system is equally applicable for use with other consumer products not discussed herein. For example, vending machines are now available to sell consumer electronics, such as music players. Consumer electronics in any context (including vending machines) often confront consumers with a wide variety of possible choices and configurations, and accordingly a product recommendation system could be advantageous in the sale and marketing of consumer electronics.
By way of non-limiting example, the methods and systems discussed herein may be utilized to gather, analyze, and process data and/or make recommendations for products including, but not limited to: apparel, appliances, accessories, baby products, cleaning products, collectables, computers, cosmetics, decorative items, electronics, fitness equipment, food and food products, footwear, fixtures, furnishings, hardware including tools, home and garden products, household supplies, jewelry, personal care products, sporting goods and equipment, telephones and other communications equipment, toys, and vehicles of all sorts.
Furthermore, the system is suitable for use in determining consumer desires and preferences with regard to non-consumer products. For example, purchasers and users of industrial and commercial-grade equipment may have needs and desires with regard to product attributes that may be ascertained using the present subject matter.
The methods and systems discussed herein may also be useful in assessing consumer perceptions in contexts other than those involving perception of products. For example, the system could be used to evaluate consumer perceptions of certain areas or aspects of advertising images, for example.
The technology discussed herein makes reference to servers, databases, software applications, and other computer-based systems, as well as actions taken and information sent to and from such systems. One of ordinary skill in the art will recognize the inherent flexibility of computer-based systems allows for a great variety of possible configurations, combinations, and divisions of tasks and functionality between and among components. For instance, server processes discussed herein may be implemented using a single server or multiple servers working in combination. Databases and applications may be implemented on a single system or distributed across multiple systems. Distributed components may operate sequentially or in parallel. When data is obtained or accessed between a first and second computer system or component thereof, the actual data may travel between the systems directly or indirectly. For example, if a first computer accesses a file from a second computer, the access may involve one or more intermediary computers, proxies, and the like. The actual file may move between the computers, or one computer may provide a pointer or metafile that the second computer uses to access the actual data from a computer other than the first computer, for instance.
The technology referenced herein also makes reference to the relay of communicated data over a network such as the internet. It should be appreciated that such network communications may also occur over alternative networks such as a dial-in network, a local area network (LAN), wide area network (WAN), public switched telephone network (PSTN), the Internet, intranet or Ethernet type networks and others over any combination of hard-wired or wireless communication links.
These and other modifications and variations to the present invention may be practiced by those of ordinary skill in the art, without departing from the spirit and scope of the present invention, which is more particularly set forth in the appended claims. In addition, it should be understood that aspects of the various embodiments may be interchanged both in whole or in part. Furthermore, those of ordinary skill in the art will appreciate that the foregoing description is by way of example only, and is not intended to limit the invention so further described in such appended claims.
