Patent Application
20160155145
The invention is from the fields of marketing and mobile communication. Specifically the invention is related to coupon recommendation systems for mobile devices.
A coupon is a document provided by a retailer, manufacturer, or service provider to potential customers that can be exchanged for a discount when purchasing a product or service. The purposes of the coupon are to attract price conscious consumers to buy specific products and/or to attract new customers. Originally coupons were printed in newspapers, magazines, or on the packaging of goods and the customer would cut them out and take them to the retail store/s specified on the coupon to be redeemed. In recent years with the development of the internet, and more recently mobile communication services, coupons are more and more being distributed in electronic form.
Smart mobile devices, being small computers that are always connected to the internet and carried by consumers, are increasingly being used for context-based recommendation. Existing applications for mobile devices push messages and coupons to the device owner mainly based on his location. The mobile devices comprise sensors that provide mobile service providers with a great deal more information than just the location of the device holder. This additional information includes, degree of activity, data usage, incoming and outgoing call and messaging logs, location of friends etc. In addition information about a specific customer's social connections, e.g. “friends” from social networks, is readily available.
These applications enable a service provider to offer different business the option to promote their products by offering coupons \discounts\offers to their potential customers. In order to be attractive to the business the applications must be designed to maximize the coupons\discounts\offers consumption while minimizing the cost to the businesses. In other words, the application must propose the right offer to the right potential customer such that there is a high probability that the offer will be consumed while searching for the most valuable, i.e. profitable, offer from the point of view of the business.
It is a purpose of the present invention to allow mobile communication device service providers to make use of the information available to them to recommend to businesses coupons that can be proposed to consumers, so that overall coupon consumption is maximized in a manner that increases the revenues of the businesses participating in the program.
It is another purpose of the present invention to provide a system that learns the pattern of coupons consumption and presents a set of coupons to a customer in the right order and time so that the system will not only increase the consumption of the customers but will also avoid giving customers the feeling that they are being pushed into accepting an offer that does not interest them.
Further purposes and advantages of this invention will appear as the description proceeds.
In a first aspect the invention is a system for executing a context-based recommendation process running in a cloud based system that utilizes the capabilities of mobile communication networks and devices to provide businesses with the ability to supply sets of coupons to a potential customer in a cost effective manner. The system comprises:
In embodiments of the system of the invention the Coordination Module comprises components adapted to send the best set of coupons and the order and timing between the offers in the set to the Recommendation module, which comprises components adapted to send the offers to the customer's mobile device.
In embodiments of the system of the invention the Coordination Module comprises components adapted to send the best set of coupons and the order and timing between the offers in the set to the customer's mobile device and to the Database.
In embodiments of the system of the invention, if the coupons in the set are not all sent at the same time, the algorithms in the Coordination Module make decisions to send subsequent coupons in the set that depend on the consumption of the preceding coupon.
In embodiments of the system of the invention the decision how best to present the coupons in the set to a specific customer in a specific context is made by the algorithms in the Coordination Module on the basis of past experience with the customer and statistical analysis of the behavior of similar customers in similar contexts.
In embodiments of the system of the invention the process of deciding which coupons are to be offered to the customer is executed by the algorithms in the Coordination Module using a probabilistic (Bayesian) platform. In these embodiments the probabilistic (Bayesian) platform is K-arm Bandits.
In a second aspect the invention is a context-based recommendation process running in a cloud based system that utilizes the capabilities of mobile communication networks and devices to provide businesses with the ability to supply sets of coupons to potential customers in a cost effective manner.
The system on which the process of the second aspect runs comprises:
The process of the invention comprises:
If the coupons in the set are not all sent at the same time, the decision to send subsequent coupons in the set depends on the consumption of the preceding coupon.
In embodiments of the process of the invention the decision how best to present the coupons in the set to a specific customer in a specific context is made on the basis of past experience with the customer and statistical analysis of the behavior of similar customers in similar contexts.
In embodiments of the process of the invention the process of deciding which coupons are to be offered to the customer is executed using a probabilistic (Bayesian) platform. In these embodiments the probabilistic (Bayesian) platform can be K-arm Bandits.
All the above and other characteristics and advantages of the invention will be further understood through the following illustrative and non-limitative description of embodiments thereof, with reference to the appended drawings.
The present invention employs a context-based recommendation process running in a cloud based system that utilizes the capabilities of mobile communication networks and devices to provide businesses with the ability to supply sets of coupons to potential customers in a cost effective manner. The process uses real time and historical information such as location, number of social messages (SMS, WhatsApp . . . ), applications used, activity (moving or stationary) etc., that is provided by various sensors on a customer's smart mobile device to derive the customer's context. To make a recommendation, the system determines the current context of the customer and selects the right offer for the current context. Herein the word “context” is defined as both the external environment (time of day, day of week, weather, temperature, traffic, location of the customer, location of the customer's friends etc.) and the internal environment (mood, hunger level, current activity, the manner in which the mobile device is used etc.) in which the customer operates. Herein the term “current context” is used to denote the context at a specific time at which the system is activated to determine which, if any, coupons to offer to a customer.
In previously filed patent applications, e.g. U.S. Ser. No. 14/691,895, the inventors have described a system in which several coupons are presented at a given point in time to a customer. The customer is then able to choose one or more of these coupons (or none), thus ending the current “transaction”. The present invention differs from the previously described inventions by the key aspect that, in this case, each of the coupons presented to the customer is potentially a set of coupons that will be presented to the customer over time.
The following example illustrates the principle of the present invention. In this example a customer is presented with three coupons on her mobile device. Two of these coupons will simply offer discounts to nearby restaurants, as in the prior art. The third coupon is in fact a set of two coupons. The customer only sees the first coupon in the set, which offers a coupon to a restaurant located in a nearby mall. If the customer chooses to consume this coupon, then, approximately 40 minutes after the meal begins, she will be offered the second coupon in the set, which offers a discount to a nearby shoe store.
It is important to note that the offer of the second (and potentially third coupon, fourth, and so on) is contingent on the consumption of the preceding coupons in the set. It should also be noted that the time intervals between the different coupons in the set may vary—a pair of coupons may be an hour apart (a coupon for ice cream after a meal at a restaurant, for example) or five minutes (as soon as a customer purchases a pair of shoes, we may offer a coupon for a nearby hat store).
The ability to offer “sets” of coupons presents both an opportunity and a difficulty: the opportunity lies in the ability of the system to consider multiple sets which share the first coupon(s). The difficulty lies in the additional calculations that need to take place. This issue is best explained by another example: This example assumes that there are three sets of coupons can be presented to a customer and that currently the customer can be presented with only a single coupon. Set 1 has coupons {A, B, C}, Set 2 has the coupons {A, D, E} and Set 3 has the coupons {F, G, H}. In this example, based on the utility values of the coupon sets as described herein below, Set 3 has a slightly higher probability of being consumed, but sets 1 and 2 have a shared initial coupon (A), which has a relatively large possibility of being purchased. The purchase of coupon A “opens the way” to two possible sets of coupons—depending on which coupon/s is/are chosen to present to the customer after the purchase of A—either coupon B or coupon D or both). The large number of options makes offering coupon A likely to be more profitable than coupon F and therefore the system chooses to initially present coupon A to the customer.
In addition, both
As an illustrative but non-limiting illustration of how the method of the invention operates, assume that there are presently two coupons—c1 and c2—that could be offered to customer u based on his/her current location and context. The following six scenarios are possible:
One of the reason that the coupons in the set should be presented separately (scenarios 5-6) instead of together (scenario 4) is that for some customers, having several coupons pushed to them at one time may arouse a feeling that an attempt is being made to force them into making purchases that they are not really interested in making and thereby cause them to “rebel” by not using any of the coupons. For these customers, the best course of action in order to increase coupon consumption would be proposing the coupons in “steps”. Other customers, on the other hand, may appreciate more “comprehensive” deals and to them the system can offer the coupons as a bundle. It is the function of the algorithms in the Coordination module to decide, on the basis of past experience with the customer and statistical analysis of the behavior of similar, e.g. age, sex, marital status, customers in similar contexts, how best to present the coupons in the set to a specific customer in a specific context.
The process of deciding which coupons are to be offered to the customer can be executed using a wide variety of algorithms. The literature in the area of recommendations (on mobile devices and in general) is extensive. The present invention can be efficiently executed using a probabilistic (Bayesian) platform, for example K-arm Bandits.
The system of the invention uses the sensors of the customer's mobile device and possibly those of his/her friends or people in the vicinity to determine the context of the customer. For each context, each individual coupon and a set (combination) of coupons are assigned with a utility score. Given a customer and his/her current context, the combination of coupons with the highest utility will be offered.
The utility score is composed of two factors: exploration and exploitation. The exploitation component represents the recommendation systems estimation of the immediate gain that can be obtained by offering the coupon to the customer. One example for a basic exploitation component would be a multiplication of the probability of the user using the coupon and the monetary gain that is to be had by that consumption. The exploration component represents the amount of information that is available for a given coupon (or a set of coupons). If little information is available for a coupon, the component will assign it with a high score, indicating an interest in experimenting with the coupon further. By combining the scores from these two components, the recommendation system balances between the need to offer customers relevant and useful coupons, while also gaining knowledge about potential new offers. The weight assigned to each component may depend on multiple factors such as the maturity of the systems, the coupon type etc.
The utility value is computed also for various consumption patterns of the same coupon set. For example, assuming that a coupon set C comprises two coupons—c1 and c2. The algorithms of the system will compute the utility function within a specific context for the case of offering different ways of presenting the coupons to the customer, e.g. presenting the two coupons together and for presenting cl and then one hour later presenting c2.
The exploration factor in the calculation of the utility score ensures that a sufficient number of offers of the specific combination of coupons were previously offered in order to make a sound recommendation, i.e. to better learn the consumption behavior of customers within the context.
The exploitation factor in the calculation of the utility score ensures that the system offers the combination of coupons with the highest reward associated with the coupon set in the specific context for the businesses offering them. The reward may depend on a variety of factors, including (but not limited to) the size of the discount, whether or not the customer has visited the business before, the day of the week, rate of consumption of the coupon combination within the current context, etc.
Both the exploitation and exploration factors of the coupon set are updated when information about the actual consumption is available. This phase is also referred to as model update.
Although embodiments of the invention have been described by way of illustration, it will be understood that the invention may be carried out with many variations, modifications, and adaptations, without exceeding the scope of the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 236016 | Dec 2014 | IL | national |
