DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 5/11/2026 has been entered.
Response to Amendment
Applicant has amended claims 1, 3, 5, 7, 8, 10, 12, 14-15, 18, and 20. Claims 1-20 are pending and rejected.
The amendments to claims 3, 5, 10, 12, and 18 are sufficient to overcome the 35 U.S.C. § 112(b) rejections set forth in the previous office action.
Response to Arguments
Applicant’s arguments with respect to 101 Rejections have been fully considered but are non-persuasive.
Applicant argues the claims do not recite an abstract idea, and specifically the generating limitations use machine learned models that extend far beyond the certain methods of organizing human activity and the mental processes grouping, and further are not a mathematical relationship
Examiner respectfully disagrees. The 35 USC 101 rejection has been updated below. The rejection does not reference the mental process or mathematical relationships. The claims recite identifying casual reasons for determining appropriate treatments, specifically dealing with identifying and preventing customer churn. These limitations do reasonably fall within the abstract idea grouping of certain methods of organizing human activity, since they involve managing personal behavior or relationships or interactions between people and also commercial interactions (including advertising, marketing or sales activities or behaviors, and business relations. It is noted that the step or function of machine learning is not positively claimed; rather the claims include descriptive language about the models that they were machine learned.
Applicant further argues that even if the claim did recite an abstract idea, the claim is integrated into a practical application because, per paragraphs 24-25 of the specification, the claims provide an improvement to known technical problems in the technical field of probabilistic models and the claims are rooted in computer technology.
Examiner respectfully disagrees. These paragraphs describe capturing adequate and useful information for the purpose of making a prediction, specifically paragraph 24. As currently recited in the claim, these concepts and limitations reasonably fall within the abstract idea grouping of certain methods of organizing human activity and are part of the recited abstract idea. Per MPEP 2106.04, judicial exceptions need not be old or long-prevalent, and that even newly discovered or novel judicial exceptions are still exceptions. As to the AI modeling via machine learning discussed in paragraph 25, theses details are not reflected in the claim limitations. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims.
Applicant also argues that amended claim 1 recites "causing execution of an action to adjust service features to address each underlying causal reason to prevent the customer to churn" and modifies the system operation by adjusting service delivery based on model- derived causal analysis, which is a concrete effect in a service system.
Examiner respectfully disagrees. The claim ends by implementing an action to adjust service features to address a root cause and prevent churn. The specification at paragraphs 24, 30,41, and 63 discusses actions as part of strategic planning that aims to address the individual concerns of the customer and to effectively prevent the customer from churning. These actions address the causal reasons that have been identified. The limitation does not include aspects of the system used to cause the execution of an action, or specify what the action is or what features are adjusted. Thus, as currently claimed, the implementing of an action falls within the recited abstract idea and does not serve to integrate the recited abstract idea into a practical application or provide significantly more.
Finally, applicant argues that the has not been evaluated as an ordered combination, evaluating the claim as a whole and the additional elements in combination with the non-additional elements.
Examiner has considered the claim as a whole, and the additional elements alone and in combination. MPEP 2106.04(d)III does instruct to consider the claim as a whole and to evaluate together the limitations containing the judicial exception as well as the additional elements in the claim besides the judicial exception to determine whether the claim integrates the judicial exception into a practical application. This has been done in the 35 U.S.C. 101 rejection below. The additional elements, when viewed individually and when viewed as an ordered combination, are recited at a high level of generality and amount to implementing the abstract idea on a computer and no more than applying the abstract idea with generic computer components. This is not sufficient to integrate the recited abstract idea into a practical application or provide significantly more.
Applicant’s arguments with respect to prior art rejections have been fully considered but are non-persuasive.
Applicant argues that Han's company/spending/usage/account features do not correspond to the recited "risk features indicative of irregular activities posing a potential risk," such as "e.g., cease to make regular payment, intermittent payment records, less frequent use of devices or services, payment before the end of contract term" and "behavior suggesting likely churn behavior or indicative of risk of churn.
In response to applicant's argument that the references fail to show certain features of the invention, it is noted that some of the features upon which applicant relies (i.e., cease to make regular payment, intermittent payment records, less frequent use of devices or services, payment before the end of contract term) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). With respect to risk features indicative of irregular activities posing a potential risk, Han does disclose trends and changes in spending and usage patterns, such as in paragraphs 40-41. These are risk factors considered when predicting risk levels in Han.
Applicant argues that Han's spending over time features do not correspond to any of user's desire, plan or behavior to churn of the present application and thus do not correspond to the recited "intent features reflecting customers' intent to churn." Examiner respectfully disagrees. The claim recites intent features (data or types of information collected) that reflect (reveal or signal) customers intent to churn. The intent features themselves are not explicitly a desire or plan. See paragraph 37, 39, 50 of the instant specification, where intent features are features that show a user’s desire, plan or behavior to churn. Churning is estimated based on this collected information about user behavior.
Applicant argues that Han's risk factor is not indicative of any underlying cause of churn of the present application and, thus, Han does not disclose the recited "causal features characterizing underlying causes of churn”. Examiner respectfully disagrees. Han does teach customers churn for reasons such as them not feeling they are receiving adequate value from a product and/or feels that the renewal price is too high (paragraph 34). These are causes for why a user may churn, and are seen in reduced usage or in average spending amount. See paragraphs 40, 42, 43, 57, for example.
Applicant argues that Han does not disclose "a hyper targeted churn (HTC) model machine learned based on activities prior to past churning" and "a causal model machine learned based on causal reasons related to past churning”. Examiner respectfully disagrees. With regard to a hyper targeted churn (HTC) model machine learned based on activities prior to past churning, it is noted that the Lidstrom reference has been added in the rejection below, based on the current amendments. With regard to a causal model machine learned based on causal reasons related to past churning, Han does teach customers churn for reasons such as them not feeling they are receiving adequate value from a product (paragraph 34). Models are trained using data about other customers and renewal of products, for example (See paragraphs 43, 45, 57).
Applicant further argues that Tuckfield and Johnson do not cure the deficiency of Han with respect to the generating limitations of claim 1. Examiner respectfully disagrees. Please see response to arguments above.
Claim Objections
Claims 1-20 are objected to because of the following informalities: claims 1, 8 and 15 contain a typographical error. The first generating limitation recites “at a corresponding level of rick to churn”. It appears this should be risk to churn. See figure 3A. Claims 2-7, 9-14, and 16-20 depend from these claims. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3, 5, 10, 12, 16 and 18 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 3, 10, and 16: Claims 1, 8 and 15 recite “a hyper targeted churn (HTC) model machine learned based on activities prior to past churning”. Claims 3, 10, and 16 recite that the HTC model is previously trained via machine learning to capture characteristics of a churner based on data associated with past churning. It is unclear if the limitation of claims 3, 10 and 16 are further including data related to the historic churn event (see for example paragraph 45) with activities prior to past churning as data on which the HTC model was trained, or if the “data associated with past churning” is the same “activities prior to past churning” that has been added to the independent claims. See figure 3C and also paragraphs 27 and 42. Clarification is requested.
Claims 5, 12, 18: Claims 1, 8 and 15 recite “a causal model machine learned based on causal reasons related to past churning”. Claims 5, 12, and 18 recite that the causal model is previously trained via machine learning to capture characteristics of a customer with causal reasons based on data associated with past churning. It is unclear if the limitations of 5, 12, and 18 are bringing in additional data concerning the causal reasons on which the causal model was trained, or merely specifying that it is data that represents the causal reasons. Based on the specification, it is assumed that both the independent claims and claims 5, 12, and 18 are referring to the same information on which the model was trained. Clarification is requested.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. Specifically Claims 1-20 are directed to an abstract idea without significantly more.
Step 1 of the Alice/Mayo analysis is directed to determining whether or not the claims fall within a statutory class. Based on a facial reading of the claim elements, Claims 1-20 fall within a statutory class of process, machine, manufacture, or composition of matter.
With respect to Step 2A, Prong One, the claims recite an abstract idea. Claims 1, 8, and 15 include limitations reciting identifying casual reasons for determining appropriate treatments, specifically dealing with identifying and preventing customer churn, including steps:
Collecting information associated with services provided to a plurality of customers;
Generating, via a hyper targeted churn (HTC) model machine learned based on activities prior to past churning, HTC segments based on the information...
Generating, via a casual model machine learned based on causal reasons related to past churning, one or more casual segments based on the information...
With respect to each of some of the plurality of customers, estimating, based on the information, the HTC segments, and the casual segments, at least one casual reason...
Causing executing of an action to adjust service features to address each underlying causal reason to prevent the customer churn.
These limitations recite an abstract idea reasonably categorized as Certain methods of organizing human activity – managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions), and commercial or legal interactions (including advertising, marketing or sales activities or behaviors, and business relations).
Claim 2-7, 9-14, and 16-20 further describe making determinations and descriptive data that further narrow the abstract idea.
With respect to Step 2A Prong Two, the claims do not include additional elements that integrate the abstract idea into a practical application. Claims 1, 8, and 15 include the additional elements of machine-readable medium, machine, system, processor. Pursuant to the broadest reasonable interpretation, Examiner submits that each of the additional elements do not integrate the abstract idea into a practical application because these elements are generic computing elements performing generic computing functions and amount to mere instructions to apply the abstract idea on a computer under MPEP 2106.05(f). Further, when viewing the claim as a whole and the additional elements alone and in combination, these elements are generic computing elements performing generic computing functions and amount to mere instructions to apply the abstract idea on a computer under MPEP 2106.05(f).
To the extent the use of the (HTC) model and casual model are machine learning models, these models generally link the use of the abstract idea to a particular technological environment or field of use under MPEP 2106.05(h). Further, they are recited at a high level of generality and include the idea of a solution without the details of how such solution is accomplished. See MPEP 2106.05(f).
Claim 2-7, 9-14, and 16-20 do not include additional elements above and beyond claims 1, 8, and 15.
As a result, Claims 1-20 do not include additional elements that would integrate the abstract idea into a practical application under Step 2A Prong Two.
With respect to Step 2B, the claims do not include additional elements amounting to significantly more than the abstract idea. As discussed above, claims 1, 8, and 15 includes the additional elements include machine-readable medium, machine, system, processor. However, individually and when viewed as an ordered combination and pursuant to the broadest reasonable interpretation, Examiner submits that the additional elements do not amount to significantly more than the abstract idea because these elements are generic computing elements performing generic computing functions and amount to mere instructions to apply the abstract idea on a computer under MPEP 2106.05(f) and/or recite generic computer structure that serves to perform generic computer functions. This analysis also applies when considering the claim as a whole, and the additional elements in combination.
To the extent the use of the (HTC) model and casual model are machine learning models, these models generally link the use of the abstract idea to a particular technological environment or field of use under MPEP 2106.05(h). Further, they are recited at a high level of generality and include the idea of a solution without the details of how such solution is accomplished. See MPEP 2106.05(f).
Claim 2-7, 9-14, and 16-20 do not include additional elements above and beyond claims 1, 8, and 15 and thus do not provide significantly more to the abstract idea.
Thus, Claims 1-20 do not provide significantly more to the abstract idea.
Accordingly, Claims 1-20 are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-5, 8-12, and 15-18 are rejected under 35 U.S.C. 103 as being unpatentable over Han (2017/0061343) in view of Lidstrom et al. (WO 2010/082885).
Regarding Claim 1,
A method, comprising (0020-0022 – computer, code, storage medium)
collecting information associated with services provided to a plurality of
customers; (0055, 0087 - company features 224, spending features 226, usage features 228, and account features 230 may be obtained from a number of data sources....)
generating, via a hyper targeted churn (HTC) model machine learned based on activities, HTC segments based on the information via disengagement features characterizing disengaging behavior, risk features indicative of irregular activities posing a potential risk, and intent features reflecting customers’ intent to churn, wherein each of the multiple HTC segments corresponds to a level of risk to churn and includes one or more of the plurality of customers estimated to be at a corresponding level of risk to churn; (0038-0042, 0046, 0063, Figure 3A(304) – the customers filtered by churn risk level (e.g., all levels, high and medium-high, high, medium-high, medium, low) (HTC segments) were placed in their churn risk level by a first statistical model (HTC model) from their company/spending/usage /account features; company/spending/usage/account features may be considered “risk features” because they predict churn risk level; spending trend over time features may be interpreted as intent features: usage engagement score features may be interpreted as disengagement features;
[0039] Company features 224 may include attributes and/or metrics associated with a customer that is a company (or other type of organization). Company features 224 may include demographic attributes such as a location, an industry, a company type (e.g., corporate, staffing, etc.), an age, and/or a size (e.g., small business, medium/enterprise, global/large, etc.) of the company.....
[0040] Spending features 226 may relate to the customer's spending behavior or spending history with the product.... Spending features 226 may also include metrics such as the customer's previous spending amounts, discount rates associated with the customer's spending amount, and/or spending growth that tracks a trend in the customer's spending amounts over time....(intent features)
[0041] Usage features 228 may identify the customer's usage of the online professional network through which the product is purchased or used. For example, usage features 228 may include metrics related to the customer's level of activity on the online professional network, such as a number of searches, messages sent, profile views, profile updates, company updates, and/or visits to the online professional network by the customer. The metrics may be aggregated into an engagement score for the company that is included in usage features 228 with the metrics or as a substitute for the metrics... (disengagement features)
[0042] Account features 230 may characterize the customer from a sales perspective. For example, account features 230 may include a potential spending amount that represents the most the company can spend on the product, given the company's size and needs. Account features 230 may also identify the stage of the sales renewal cycle occupied by the customer....(intent features)
Account features 230 may also include attributes and/or metrics that are relevant to the product. For example, account features 230 for predicting the customer's churn risk 216 for a recruiting solution may include the number of recruiters, number of talent professionals (e.g., human resources staff), and/or size of the staffing department in the company.
generating, via a causal model machine learned based on casual reasons related to past churning, one or more causal segments based on the information via causal features characterizing underlying causes of churn, the risk features, and the intent features, wherein each of the one or more causal segments corresponds to an underlying causal reason to drive a customer to churn and includes at least one of the plurality of customers estimated to have the underlying causal reason to churn; (
[0077] A second statistical model (causal model) is then used to obtain one or more risk factors (causal features) associated with the churn risk (operation 412), independently of the churn risk level of the customer. For example, the statistical model may use one or more decision trees to compare a subset of the features (ie. causal, risk, intent) for the customer to a number of thresholds. When a feature does not meet a given threshold, the corresponding risk factor is identified in the customer.
Figure 3A (304), 0047–0048- the displayed risk factors (casual segments) for customers associated with churn risk level.
Customers churn for reasons such as them not feeling they are receiving adequate value from a product (paragraph 34). Models are trained using data about other customers and renewal of products, for example (See paragraphs 43, 45, 57))
with respect to each of some of the plurality of customers, estimating, based on the information, the HTC segments, and the causal segments, at least one causal reason, each of which corresponds to an underlying causal reason that potentially drives the customer to churn (Figure 3A (304) – the display of the customers associated with a filtered churn risk level (ie. high) associated to each of the risk factor causes (ie. prod usage, prod performance))
[0053] Finally, management apparatus 206 may provide one or more recommendations 240 for reducing high customer churn risk levels in GUI 204. For example, GUI 204 may identify one or more risk types associated with risk factors 232 and include information for engaging with customers to mitigate high churn risk levels based on the risk types.
Figure 3C, 0071-0072- the GUI may show a set of user-interface elements 330-336 related to risk factors associated with churn risk in customers of a product such as a recruiting solution....User-interface elements 330-332 may summarize different types of customer churn risk. For example, user-interface elements 330-332 may include names of the churn risk types, symptoms of the churn risk types, and/or prescriptions for addressing the churn risk types. User-interface elements 334-336 may provide detailed information for managing customers associated with the risk types identified in user-interface elements 330-332. For example, user-interface elements 334-336 may include slide decks that describe techniques for engaging with the customers and/or otherwise addressing issues associated with the risk types. In other words, the GUI of FIG. 3C may provide recommendations for reducing churn risk associated based on the risk types, which may be used by a sales professional to manage customer relationships and improve his/her sales performance.
causing execution of an action to address each underlying causal reasons to prevent the customer to churn. (0078-a communication containing content for reducing the churn risk is optionally transmitted to the customer (operation 416). For example, a risk factor associated with sub-optimal results experienced by the customer with the product may be mitigated by engaging the customer with marketing content that addresses a number of potential sources of the sub-optimal results. 0085 – a recommendation is provided, such as to a sales professional, such as including information on how to engage the customer and reduce churn risk)
While Han teaches generating HTC segments via a hyper targeted churn (HTC) model that was machine learned based on activities (See 43, 57, 89 and 93), Han does not explicitly disclose that the model was machine learned based on activities prior to past churning.
Lidtrom discloses that the model was machine learned based on activities prior to past churning (See figure 3, page 3, lines 21-25, and page 8, lines 24-28, where a machine learning system uses training data corresponding to the activity of churned customers before their churn to identify trends. The data was previously collected during a period of time prior to the churning of a customer(s)). It is noted that Lidstrom explicitly identifies causal reasons (see page 8, lines 20-29, page 9, lines 15-20).
Both Han and Lidstrom are directed to predicting and reducing potential customer churn using information about the customers and risks. It would have been obvious to one of ordinary skill in the art before the effective filing date to include a model trained on activities prior to past churning in order to most effectively identify patterns in data of existing customer of those customers most likely to churn based on data of activities of previous customers prior to them churning.
Further while Han discloses recommended actions to take to reduce customer churn risk, Han does not specifically disclose the action being to adjust service features. Lidstrom discloses the action being to adjust service features to address each underlying causal reasons to prevent the customer to churn (See page 12, lines 6-17 and line 31 – page 13, line 2, where churn retention measures are generated to address reasons for churn and attempt to retain the potential churning customer. See also figure 3, page 12, lines 1-5).
Both Han and Lidstrom are directed to predicting and reducing potential customer churn using information about the customers and risks. It would have been obvious to one of ordinary skill in the art before the effective filing date to include adjusting service features in the actions taken with customers at risk to churn in order to best address the cause or reason for the likely churn of the customer and most effectively retain the potential churn customer. See page 12, lines 14-17, of Lidstrom and paragraphs 53, 70, 72 and 85 of Han.
Regarding Claim 2, Han discloses: The method of claim 1, wherein the generating the HTC segments comprises:
with respect to each of the plurality of customers and based on the information relevant to the user, extracting the disengagement features of the customer,
[0041] Usage features 228 may identify the customer's usage of the online professional network through which the product is purchased or used. For example, usage features 228 may include metrics related to the customer's level of activity on the online professional network, such as a number of searches, messages sent, profile views, profile updates, company updates, and/or visits to the online professional network by the customer. The metrics may be aggregated into an engagement score for the company that is included in usage features 228 with the metrics or as a substitute for the metrics... (disengagement features)
determining an intent of the customer to churn based on the intent features,
[0042] Account features 230 may characterize the customer from a sales perspective. For example, account features 230 may include a potential spending amount that represents the most the company can spend on the product, given the company's size and needs. Account features 230 may also identify the stage of the sales renewal cycle occupied by the customer....(intent features)
estimating a level of risk to churn associated with the customer, and (0046, 0063, Figure 3A(304) - customers churn risk level (e.g., all levels, high and medium-high, high, medium-high, medium, low) (HTC segments) determined by the first statistical model (HTC model))
identifying whether the customer corresponds to a churner in accordance with the HTC model; and creating the HTC segments at different levels of risk to churn, wherein each of the HTC segments associated with a level of risk to churn includes those of the plurality of customers identified as a churner and having an estimated level of risk to churn corresponding to the associated level of risk to churn. [0044] As a result, churn risk 216 may be predicted for the customer by selecting a first statistical model (HTC model) that matches the company segment, its stage in the sales renewal cycle, and/or other features of the customer, and then inputting one or more company features 224, spending features 226, usage features 228, and/or account features 230 into the statistical model. In turn, the first statistical model may generate a prediction of churn risk 216 and one or more thresholds 218 associated with churn risk 216.
[0045] As described above, churn risk 216 may be a numeric score or value that represents the customer's propensity for fully or partially churning from the product. Because churn risk 216 may be assessed in relation to other values of churn risk 216 for other customers, thresholds 218 may represent values that indicate certain levels of churn risk 216, such as medium, medium-high, or high. For example, thresholds 218 may be set to values that represent certain percentiles of churn risk 216 for the company segment and stage in the sales renewal cycle of the customer. (corresponds to a churner)
Regarding Claim 3, Han discloses: The method of claim 2, wherein the HTC model is previously trained via machine learning to capture characteristics of a churner based on data associated with past churning. (0043- Analysis apparatus 202 and/or another component of the system may create and maintain a set of statistical models 208 that predict churn risk 216 for different subsets of customers. Each statistical model may be trained and/or updated on a periodic basis (e.g., daily) using data associated with the corresponding subset of customers from data repository 134.
[0056] Finally, statistical models 208 may be implemented using different techniques and/or used to generate churn risk 216, thresholds 218, churn risk level 234, and/or risk factors 232 in different ways. For example, churn risk 216 and/or thresholds 218 may be generated using a gradient tree boosting technique, while risk factors 232 may be identified using one or more additional decision trees. Other types of statistical models, such as artificial neural networks, Bayesian networks, support vector machines, and/or clustering techniques, may also be used with or in lieu of the gradient tree boosting technique and/or decision trees to provide the functionality of analysis apparatus 202. Alternatively, churn risk 216, thresholds 218, churn risk level 234, and/or risk factors 232 may be generated using the same statistical model instead of separate statistical models.
Regarding Claim 4, Han discloses: The method of claim 1, wherein the generating the causal segments comprises:
with respect to each of the plurality of customers and based on the information relevant to the user, extracting the causal features of the customer, determining an intent of the customer to churn based on the intent features, estimating propensity of the customer to churn, and (0055, 0087, Figure 3A (308) – after receiving company features 224, spending features 226 (contains intent to churn), usage features 228, and account features 230, filtering customers associated to risk levels)
predicting a causal reason associated with the customer in accordance with the causal model; (Figure 3A (304), 0047–0048- displayed risk factors (casual segments) for customers associated with churn risk level)
and creating the causal segments with corresponding causal reasons, wherein each of the causal segments associated with a causal reason includes those of the plurality of customers predicted to have a causal reason corresponding to the associated causal reason. (Figure 3A (304) – list of customers associated to particular risk factors)
Regarding Claim 5, Han discloses: The method of claim 4, wherein the causal model is previously trained via machine learning to capture characteristics of a customer with a causal reason based on data associated with past churning. (0043- ... Each statistical model may be trained and/or updated on a periodic basis (e.g., daily) using data associated with the corresponding subset of customers from data repository 134)
Claims 8-12 and 15-18 stand rejected based on the same citations and rationale as the Claims 1-5.
Claims 6, 13, and 19 is rejected under 35 U.S.C. 103 as being unpatentable over Han in view of Lidstrom et al. and in further view of Tuckfield (US Patent 11651314)
Regarding Claim 6, Han discloses: The method of claim 1, wherein the estimating the at least one causal reason comprises: obtaining one of the HTC segments that includes the customer; identifying one or more of the multiple causal segments that include the customer; and obtaining the at least one causal reason associated with the one or more causal segments associated with the customer.
(Figure 3A (304) – the customers associated to the filtered risk level (ie. high) (HTC segment) and risk factors (casual reasons associated with casual segment))
Han and Lidstrom do not explicitly state: Tuckfield in analogous art discloses: determining an intensity of each of the causal reasons associated with the causal segments; ranking the causal reasons according to the respective intensities associated therewith; (3(3-7) - the system can evaluate the factors contributing to the customer's attrition risk, and rank them. Put another way, the system can evaluate the matters most important in explaining the likelihood of the customer's terminating the contract or the relationship.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to associate Tuckfield’s intensities and rankings to Han’s casual reasons (and the combination of Han and Lidstrom), since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Claim 13 and 19 stand rejected based on the same citations and rationale as applied to Claim 6.
Claims 7, 14, and 20 is rejected under 35 U.S.C. 103 as being unpatentable over Han in view of Lidstrom et al. and in further view of Johnson (20100223099).
Regarding Claim 7, Han discloses: The method of claim 1. Han and Lidstrom do not explicitly state: Johnson analogous art discloses: recommending the action directed to each of the at least one causal reason by:
with respect to each of the at least one causal reason, accessing a driver/action mapping model, and mapping, via the driver/action mapping model, the causal action to a corresponding market action. (Figure 2 – V-Factors that influence churn (casual reasons) feed into (map to) a MDOO process/model determining offers;
Abstract - A Multi-Dimensional Offer Optimization. TM. (MDOO) process is provided that may be defined generally as an offer simulation engine that matches an offer most likely to be accepted to the customer most likely to accept it.
[0026] The analysis of V-Factors begins with the investigation to determine the drivers of churn and their level of influence (positively or negatively) in the churn model. If specific variables contribute significantly to churn (positively or negatively), then offers built using those variables can be used to influence churn behavior. The superset of V-Factors identified defines the characteristics that influence churn. These characteristics drive the identification of offers to be given to the subscribers who are likely to leave and terminate service.
[0028] MDOO process 20 of FIG. 2 is a computer implemented offer simulation engine which matches the offer most likely to be accepted by a customer. In the illustrated example, the purpose of MDOO process 20 is to literally ask the question, "Which offer would be most effective in saving this particular subscriber?" The effect of the offer is positive if the resulting churn score is lowered and vice versa. The same model that is used to generate predictions is used to predict the effect on churn of various interventions. In one embodiment, an aspect of the invention explores all possible interventions as determined by quantitative analysis. For each intervention, a reduction in churn probability may be computed, and then an intervention is chosen that leads to the greatest reduction in probability. If different interventions have different costs, decision-theoretic techniques may be used to choose the intervention that yields the greatest savings to the carrier.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to associate Johnson’s mapping via driver/action mapping model to Han and Lidstrom’s recommended action, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Claim 14 and 20 stand rejected based on the same citations and rationale as applied to Claim 7.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Vakil et al. (US 2024/0370898) teaches hyper-targeted offer optimization and nanosegments based on specific customer-related inputs, such as customer analytic records (“CAR”s) 210. In different embodiments, a CAR may be used as input to descriptive and predictive models to determine how consumers are likely to respond to marketing offers. The models may also be used to predict a likelihood of attrition or other behaviors.
De Knijf et al. (US 2017/0169345) discloses a churn prediction model can be provided that uses both behavioral data as well as user characteristics, where the probability of churning is based on the training dataset.
Colley (Customer Attrition: How to Define Churn When Customers Do Not Tell They’re Leaving) discusses different behaviors surrounding churn and data signaling a churn event, including information about activity prior to churning.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BETH V BOSWELL whose telephone number is (571)272-6737. The examiner can normally be reached M-F 8AM - 4:30PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Tariq Hafiz can be reached at (571) 272-5350. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/BETH V BOSWELL/Supervisory Patent Examiner, Art Unit 3625
Prosecution Insights