DE102008044453A1 - Online-shop operating method for purchasing goods, involves assigning controlling unit to server, which records input entry, output entry and purchasing entry that are accumulated in frame of communication between client computer and server - Google Patents

Abstract

The method involves assigning decision logic (1) to a server. A shopping basket value (A), offering costs (B), acceptance probability (C) and probability for a purchase failure (D) are determined as input parameters, where the decision logic calculates an excepted benefit. A positive excepted benefit is provided to customer. A controlling unit is assigned to the server, which records an input entry, output entry and a purchasing entry that are accumulated in a frame of communication between a client computer and the server.

Description

The The present invention relates to a method for operating an online shop, which on a data-connected with a client computer of a customer Server is set up, the customer being from the client machine from goods offered in the online store selects the selection of the customer in a virtual shopping cart assigned to the customer is saved and the client computer signals the server, as soon as the customer indicates the completion of his selection of goods, whereupon The server automatically decides whether to offer the customer an offer should be submitted.

Such a method is already out of the US 2006/0224464 A1 previously known. This method relates to a network of online shops to which a higher-level server is assigned. As soon as a customer indexes the completion of a purchase in one of the online shops organized in the said network, ie changes from the offer pages of the online shop to the shopping cart page, the higher-ranking server within the network checks whether an alternative online shop has an alternative Offer for the goods selected by the customer. Just in the moment in which the customer wants to conclude the purchase, he will then, if available, an alternative offer from another online store, submitted and the customer can choose between the two offers.

The Problem that a customer at the last moment before graduation of a purchase contract with an online shop but still against you Buy decides has been known for some time. providers Such online stores are even increasingly watching the customers first in a long, painstaking detail work one Put your shopping cart together, then hesitate and finally without completing a purchase, leave the pages of the online store again. There are several approaches in this regard, such as created by the provider additional purchase incentives can be to just before this last moment Conclusion of a sales contract to convince the customer yet to actually complete the purchase.

in this respect are the providers of online stores over the real ones Shops at a disadvantage after the seller in a real shop on the mood of the customer and, where appropriate, as far as the proposed business for sufficiently lucrative, the customer also can quite accommodate. In that the seller watching the customer and studying his behavior, he is So compared to the one sitting almost anonymous in front of the computer Thus, online providers looking after online customers will benefit in such a late phase of business initiation to succeed with greater certainty.

The The present invention has the object, the Disadvantages of the online store in this regard, so far so to respond to customer behavior and the customer, as far as this is necessary to submit an appropriate offer.

This is solved by a method for operating an online shop according to the features of claim 1. Further useful Embodiments of this method can the subclaims be removed.

According to the invention the server on which the online shop is based, a decision logic which decides whether to make an offer to a customer will or not. Of course it would be possible every customer at the time he makes the decision the completion of the purchase falls, however, to make an offer Such behavior on the cost side would have to be quite be called nonsensical. In particular, the costs would be for the additional offers made in the Cost accounting must be included, so that the offers of the respective online shop to the appropriate additional offers would have to be more expensive. So, um to examine to what extent the submission of an additional Offer for the online store is worth the decision logic provided different input parameters, namely, in particular, the shopping cart value on which depends to what extent offers to the respective customer can be with the desired offers associated offer costs and an acceptance probability, which indicates the probability with which the customer for the additionally submitted offer at all Interested. The probability also flows into the consideration that the customer will cancel the purchase. Only if for the decision logic after consideration this parameter still yields a profit, becomes a corresponding one Offer submitted.

The profit is calculated according to how high the shopping cart value minus the costs to be used for the offer for the operator of the online shop is, the resulting value is multiplied by the probability of acceptance and the probability of termination. Then the cost of the offer is multiplied by the expected profit multiplied by the inverse of the probability of discontinuation. An offer will only be made if the resulting value is greater than zero. In this way, it is possible that an offer only is done when it is actually worthwhile for the online store, especially when it is expected that the customer will actually respond to the offer.

Around to determine if a customer might be on such last attempt to make him an offer, reacts, be the processes during the selection of the goods the customer recorded by a control unit, which insofar Collects information about how the customer is in the online store behaves. For example, a customer who is unerring select individual goods and these in short time intervals Add to cart, finally this one enters immediately and within a short time, viz within a critical period of time, the purchase completes, not animated by an additional offer to buy Need to become. On the other hand, a hesitant Customer who puts goods into the goods basket and out again takes and stays here for a long time in the online shop, rather encouraged by an additional offer in this direction become. Such an offer may consist of, for example, the customer to post the shipping costs, additional themed Goods for the already selected by the customer goods for free or to offer discounted, vouchers to provide, Etc.

The Control unit can in advance so far already an impression of it win, with the respective customer the experience of the control unit after a break in purchase must be expected.

by virtue of the data recorded by the control unit, the customer behavior reflect and allow a conclusion on how the customer behaves in the online shop, and possibly also determined from the goods selected by the customer then a probability logic the abort probability and the acceptance probability. Because of these assessments, which is provided to the decision logic is first according to the formula described above, whether the customer a Offer is made, or whether it is foreseeable if a purchase cancellation not to be expected or expected to be lost. This calculation is always executed when the customer at least one Add product to cart. In addition, the exact Time for the submission of the offer to be variable and be self-learning optimized by the system.

Around to improve the accuracy of the method described above can individually for each customer the probability of saved a purchase, so that at another Visiting the customer in the respective online store already a value present the behavior of the customer during the last Visits to the online store reflects. This makes it possible to respond better to the behavior of each individual customer can, in particular also the respectively laid on the day Behavior of the customer to better interpret. A Further improvement results from the fact that the actual Purchase decision of the customer, so whether he actually buying the carried out or canceled, especially after Submitting an offer to which probability logic is reported, so that these for the respective customer respectively value for the general public through self-learning Can improve the process. Likewise in this way the acceptance of a particular offer is checked, so that shows in this way, whether individual offers complete Unattractive or other offers may be regular to be entitled to. This acceptance probability of Offer is made when selecting the respective customer Offer considered.

After all the method can still be improved in its accuracy be that the operator of the online store may have over the Customer collected data, such as gender or about the purchase history of the respective customer with the on-line shop, to the Likelihood logic forwards, which also includes these additional Data for obtaining the required probability values evaluates.

It is not always interesting in the sense of the customer and thus not always also promotional, if the customer regularly only the same offer is made. In that regard, stand with Advantage of the decision logic several offers available which are accessible to the decision logic Parameter memory are kept. The parameter memory contains in connection with each offer, the associated offer costs, while the probability logic for each offer the Can provide acceptance probability. The one after the above Formula calculated profit will be available for each The offer is calculated and the offer is submitted to the customer. which with this formula the biggest profit promises. In that regard, the decision logic can also be a set of rules be assigned, which, for example, prevents the customer the same offer is made regularly, although it represents the best offer in terms of profit maximization. Likewise, it allows such a set of rules to the customer to make goods-specific offers, for example in the case of Purchase a pocket radio, for example, at a discounted price To offer headphones.

As already described above, the gain is particularly under Taking into account the value of the shopping cart, so that it seems indicated that the decision logic is also on a Representation of the shopping cart can access. Based on this Representation, which is an image of the compiled in the online shop Cart represents, the value of the cart for the operator of the online store determined to end up with the offer to calculate the profit to be achieved.

With Advantage is the decision logic together with its associated Modules contained in a utility, which in particular can also run separately from the server. This makes it possible that just this service, which makes the offers, at the same time be set up for several different online shops and can serve them at the same time. To the decision logic associated modules include in particular the probability logic, the parameter memory, the rules, the representation of the shopping cart and / or the control unit for recording customer behavior.

A such configuration requires that the server, on which the online store is running, the input events the customer must forward to the utility. The utility can use this control unit to determine customer behavior. Also, data such as the goods selection of the Customers sent to the utility. The customer indexes the degree of purchase, this too is signaled to the utility which then, based on the data available to him, decides whether an offer should be submitted or not. In the event that an offer is to be submitted to the server, such as this Offer should be; in the event that an offer is not is to be submitted, the server is just communicated this fact. Ultimately, the server submits after completing the purchase or in the worst case after the termination of the purchase, which is determined by a session timeout, the utility, with what success the offer was made or how the customer behaved. The session timeout then occurs if the customer no longer has a predetermined time with the online store has communicated and so far is to be expected that the connection the customer to the online store has been terminated.

Becomes offers the customer an offer, so there is a possibility in that to the product selected by the customer Widget is created, which the additional offer, For example, a discount, indexed. However, this causes administrative Problems, since in an additional administrative instance everyone Product the respective offer must be assigned. This can however, if an offer itself is treated as a good, So when accepting the offer by the customer the offer as a selected product is added to the cart.

The The invention described above will be described below with reference to a Embodiment explained in more detail.

It demonstrate

1 a schematic representation of the decision logic and the parameters provided in a block diagram,

2 a utility in which the decision logic according to 1 as well as the modules associated with this decision logic are contained in a block diagram, as well

3 a representation of the communication between a server and the utility according to 2 in a block diagram.

1 shows a decision logic 1 , which is used for the implementation of a method for operating an online shop. As part of the operation of the online shop, the decision logic 1 decide whether to make an additional offer to the customer who wishes to conclude the purchase of the goods selected by him, which is to promote the decision of the customer to conclude the purchase. This becomes interesting at the moment when the customer visits the shopping cart area within the online shop, from where the purchase can ultimately be completed. After customers regularly cancel their purchase shortly before the purchase is completed, it is the goal of the decision logic 1 to ensure that an offer is made to the customer, as far as this is likely to influence his purchase decision.

For this purpose, the decision logic 1 provided a number of parameters, namely the basket value A, the bid cost B, the acceptance probability of the bid C and the probability of the purchase termination D. The decision logic 1 will check on entering the shopping cart page by the customer, if and what offer is submitted to the customer. The decision logic 1 For this purpose several offers are available, which are based on one of the decision logic 1 associated parameter memory 4 can be removed. For each offer then the profit for the operator of the online shop is determined by the offer value B are deducted from the basket value A. The difference here from is multiplied by the acceptance probability C as well as the probability of a purchase cancellation D, which is also one of the decision logic 1 associated probability logic 2 , which is also the decision logic 1 is assigned, is determined. The offer, which shows the greatest benefit for the operator of the online shop, is therefore chosen, with no offer being made if the profit for all offers is less than zero. Finally, if the offer submitted to the customer is also accepted by the customer, then it is inserted into the shopping cart like a selected product and can thus be taken into account by the merchandise management system without any additional effort.

2 shows a utility 7 which the decision logic 1 includes. Also includes the utility 7 a probability logic 2 which for the decision logic 1 the probability of the purchase termination D, as well as the acceptance probability C for accepting a submitted offer provides. The control unit 3 of the utility 7 serves to collect the input events E of a customer and on the one hand to the probability logic 2 to forward them so that they can derive the probabilities from the customer behavior, and - as far as the input events E are additions of goods into the shopping cart - also to a representation of the shopping cart 6 in which all goods put together by the customer are also listed. This representation of the shopping cart 6 serves to determine the basket value A and this in turn the decision logic 1 to provide. Besides that also in the utility 7 contained parameter memory 4 which is already described above contains the utility 7 in addition a set of rules 5 which the decisions of the decision logic 1 can modulate. For example, the rules and regulations 5 that a customer is not offered the same offer on a regular basis, because otherwise he could get used to it and would trust in every purchase that he would receive a corresponding offer. Likewise, the rules can 5 ensure that the goods selected by the customer are dealt with in detail, so that, for example, suitable offers, in the form of vouchers, etc., are submitted for the goods. Looking now at the utility 7 as a black box, so be to the utility 7 from the server 8th on which the online store is set up, the input events E, ultimately the purchase decision G, as well as a request for an offer recommendation to the utility 7 directed. The utility 7 on the other hand returns to the server a recommendation for an offer, which is the server 8th to submit to the customer.

3 shows again the connections between the utility 7 and the server 8th , Indexed a customer of the server 8th patched online shops that he has finished his purchase, or the compilation of his goods, as reported by the server 8th via the input events E in a request H to the utility 7 , The utility 7 is due to the constantly from the server 8th to the utility 7 transmitted input events E and within the utility 7 in the parameter memory 4 held offers a recommendation I to the server 8th in which either an offer is recommended to the customer or it is recommended that no offer be made to the customer. The last one will be the server 8th the utility 7 notify the customer of the purchase decision G so that within the utility 7 this decision of the customer in the sense of a feedback of a rule loop of the probability logic 2 which learns from this for future decisions. In addition to the described data, which about the customer to the utility 7 it is also possible that the server 8th otherwise collected customer data J, in particular, for example, the gender, age and / or the purchase history of customers, for further consideration to the utility 7 transmitted.

above Thus, a method for operating an online shop is described, which is improved so that a decision logic short before a possible cancellation of a purchase or an offer directed to the customer, which is selected on the basis of empirical values is and is then submitted to the customer, if to be expected is that this actually cancel the purchase would. As a result, this process becomes efficiency significantly increase an online store.

1

decision logic

2

probability logic

3

control unit

4

parameter memory

5

regulations

6

representation of the shopping cart

7

utility

8th

server

A

Basket value

B

offering expenses

C

acceptance probability

D

probability for the purchase cancellation

e

input events

F

rule

G

buying decision

H

inquiry

I

recommendation

J

customer data

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• US 2006/0224464 A1 [0002]

Claims (16)

Method for operating an online shop which is located on a server connected to a client computer of a customer ( 8th ), wherein the customer selects goods offered in the online shop from the client computer, the customer's selection is stored in a virtual shopping cart assigned to the customer, and the client computer ( 8th ) as soon as the customer indicates the completion of his goods selection, whereupon the server ( 8th ) automatically decides whether an offer should be made to the customer, characterized in that the server ( 8th ) a decision logic ( 1 ) for which as input parameters a shopping basket value (A), offer costs (B), an acceptance probability (C) and a probability for the purchase termination (D) are determined, on the basis of which the decision logic ( 1 ) calculates the expected profit, whereby the offer is made to the customer only with a positive expected profit. Method according to claim 1, characterized in that the server ( 8th ) a control unit ( 3 ), which in the context of the communication between client computer and server ( 8th ) records occurring input events (E), selection events and purchase events (G), and uniquely assigns them to the customer. Method according to claim 2, characterized in that a probability logic ( 2 ) because of the control unit ( 3 ) recorded input events (E) the probability of a purchase cancellation (D) by the customer and the probability of acceptance (C). A method according to claim 3, characterized that indicates the probability of the purchase (D) stored on the server in connection with the customer. Method according to one of claims 3 or 4, characterized in that the server ( 8th ) the purchase decision (G) of the customer to the probability logic ( 2 ), due to which preferably the acceptance probability (C) of the offer, preferably also or only with respect to the respective customer, is modulated. Method according to one of claims 3 to 5, characterized in that the server ( 8th ) to the probability logic ( 2 ) is transmitted from this to be taken into account, customer-describing data (J). Method according to one of the preceding claims, characterized in that the decision logic ( 1 ) can choose between different offers, whereby the expected profit for each offer is calculated and in each case the offer is submitted, for which the largest expected profit is determined. Method according to claim 7, characterized in that the decision logic ( 1 ) with a parameter memory ( 4 ) in which the available bids are stored in relation to the respective bid costs (B). Method according to one of claims 7 or 8, characterized in that the decision logic ( 1 ) a set of rules ( 5 ), which, depending on the goods selected by the customer, selects particular bids through the decision logic ( 1 ) favors. Method according to one of the preceding claims, characterized in that the decision logic ( 1 ) a representation of the shopping cart ( 6 ), which contains the current selection of the customer of goods and on the basis of which the shopping cart value (A) is determined and the decision logic ( 1 ) provided. Method according to one of the preceding claims, characterized in that the decision logic ( 1 ) from the server ( 8th ) separated in a utility ( 7 ) runs on a third computer, which preferably also the probability logic ( 2 ), the parameter memory ( 4 ), the regulations ( 5 ), the representation of the shopping cart ( 6 ) and / or the control unit ( 3 ) includes. Method according to claim 11, characterized in that the server ( 8th ) the utility ( 7 ) transmits the input events (E) and the indexing of the completion of the goods selection of the customer, the utility ( 7 ) the server ( 8th ) either an offer to submit or, in the event that no offer is made, the server ( 8th ) and the server ( 8th ) the utility ( 7 ) finally announces the purchase decision (G). A method according to claim 12, characterized characterized in that the offer only after expiry of a critical Dwell time after indexing the completion of the goods selection by the customer is submitted to this, the critical length of stay preferably self-learning determined by the system. Method according to one of the preceding claims, characterized in that the indexing of the completion of the goods selection in Change to shopping cart exists. Method according to one of the preceding claims, characterized in that a purchase canceled by the customer via a session timeout is determined. Method according to one of the preceding claims, characterized in that at sub-widths of an offer this treated as a commodity and in case of acceptance by the customer the cart will be added.