Techniques for determining behavioral pairing in task distribution system

Abstract

The invention relates to techniques for determining behavioral pairing in a task distribution system. In one particular embodiment, the technique may be implemented as a method for determining behavioral pairing in a task distribution system, comprising: determining, by at least one computer processor communicatively coupled with and configured to operate in the task distribution system, behavioral pairing in the task distribution system; determining a plurality of possible task-agent pairs among the at least one task waiting for allocation and the at least one agent available for allocation; and selecting, by the at least one computer processor, a first task-agent pair of the plurality of possible task-agent pairs for allocation in the task allocation system based at least in part on a first set of offers to be proposed by the agent or a first compensation to be received by the agent.

Description

[0001] This application is a divisional application of a patent application filed on December 25, 2020, with an application date of September 15, 2020, application number 202080003629.5 (PCT / US2020 / 050850), and invention name “Technology for determining behavior pairing in task assignment systems”.

[0002] CROSS-REFERENCE TO RELATED APPLICATIONS

[0003] This international patent application claims priority to U.S. patent application No. 16 / 576,434, filed on September 19, 2019, the entire contents of which are incorporated herein by reference. Technical Field

[0004] The present disclosure relates generally to behavior pairing, and more particularly to techniques for deciding behavior pairing in a task allocation system. Background Art

[0005] Typical task distribution systems algorithmically distribute tasks arriving at a task distribution center to agents available to handle them. Sometimes, a task distribution center may be in an "L1 state" with agents available and waiting to be assigned tasks. Other times, a task distribution center may be in an "L2 state" with tasks waiting in one or more queues for agents to become available for assignment.

[0006] In some typical task distribution centers, tasks are assigned to agents in order based on arrival time, and agents receive tasks in order based on when they become available. This strategy may be referred to as a "first-in, first-out," "FIFO," or "round-robin" strategy. For example, in an L2 environment, when an agent becomes available, the task at the head of the queue is selected for assignment.

[0007] In other typical task dispatch centers, performance-based routing (PBR) strategies can be implemented to prioritize task dispatch to higher-performing agents. Under PBR, for example, the highest-performing agent among the available agents receives the next available task. Other PBR and PBR-like strategies may use specific information about agents for dispatch.

[0008] The "behavioral pairing" or "BP" strategy for assigning tasks to agents improves upon traditional assignment methods. BP aims to balance agent utilization while improving overall task assignment center performance, potentially exceeding what FIFO or PBR approaches can achieve in practice.

[0009] In some task allocation systems, it may be advantageous to consider suboptimal actions for tasks assigned to specific agents. Therefore, it can be appreciated that a decision-making BP strategy may be needed that considers suboptimal actions for task-agent pairings to optimize the overall performance of the task allocation system. Summary of the Invention

[0010] A technique for determining behavioral pairings in a task assignment system is disclosed. In one specific embodiment, the technique can be implemented as a method for determining behavioral pairings in a task assignment system, comprising: determining, by at least one computer processor communicatively coupled to the task assignment system and configured to operate in the task assignment system, a plurality of possible task-agent pairings among at least one task waiting to be assigned and at least one agent available for assignment; and selecting, by the at least one computer processor, a first task-agent pairing from the plurality of possible task-agent pairings for assignment in the task assignment system based at least in part on a first offer set to be provided by the agent or a first compensation to be received by the agent.

[0011] According to other aspects of this particular embodiment, the task allocation system may be a contact center system.

[0012] According to other aspects of this particular embodiment, selecting the first task-agent pairing may be based at least in part on both the first set of bids and the first compensation.

[0013] In accordance with other aspects of this particular embodiment, the method may further include selecting, by at least one computer processor, a first set of bids from a plurality of sets of potential bids.

[0014] In accordance with other aspects of this particular embodiment, the method may further include selecting, by at least one computer processor, a first compensation from a plurality of potential compensations.

[0015] According to other aspects of this particular embodiment, selecting the first task-agent pairing may be based on at least one of a first ranking of the plurality of tasks and a second ranking of the plurality of agents, and wherein at least one of the first ranking and the second ranking is expressed as a percentile or a range of percentiles.

[0016] According to other aspects of this particular embodiment, the method may further include adjusting, by at least one computer processor, the first set of bids or the first offset to adjust the first ranking or the second ranking.

[0017] In another particular embodiment, the technology can be implemented as a system for determining behavior pairings in a task assignment system, the task assignment system comprising at least one computer processor communicatively coupled to the task assignment system and configured to operate in the task assignment system, wherein the at least one computer processor is further configured to perform the steps in the above method.

[0018] In another specific embodiment, the technology can be implemented as an article for determining behavior pairing in a task allocation system, which task allocation system includes a non-transitory processor-readable medium and instructions stored on the medium; wherein the instructions are configured to be readable from the medium by at least one computer processor, and the at least one computer processor is communicatively coupled to the task allocation system and is configured to operate in the task allocation system, thereby causing the at least one computer processor to operate so as to perform the steps in the above-mentioned method.

[0019] The present invention will now be described in more detail with reference to specific embodiments of the invention as illustrated in the accompanying drawings. Although the present disclosure is described below with reference to specific embodiments, it should be understood that the present disclosure is not limited thereto. Those of ordinary skill in the art having access to the teachings herein will recognize additional embodiments, modifications, and embodiments within the scope of the present disclosure described herein, as well as other areas of use, and with respect to which the present disclosure may have significant utility. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to facilitate a more complete understanding of the present disclosure, reference is now made to the accompanying drawings, in which like elements are represented by like numerals. These drawings should not be construed as limiting the present disclosure, but are intended to be used for illustration only.

[0021] Figure 1 A block diagram of a task distribution center according to an embodiment of the present disclosure is shown.

[0022] Figure 2 A block diagram of a task allocation system according to an embodiment of the present disclosure is shown.

[0023] Figure 3 A flowchart of a task allocation method according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0024] A typical task distribution system algorithmically distributes tasks arriving at a task distribution center to agents available to handle them. At times, the task distribution system may be in an "L1 state" with available agents waiting to be assigned tasks. At other times, the task distribution center may be in an "L2 state" with tasks waiting in one or more queues for agents to become available for assignment. Still other times, the task distribution system may be in an "L3" state with multiple available agents and multiple tasks awaiting assignment. An example of a task distribution system is a contact center system that receives contacts (e.g., phone calls, internet chat sessions, emails, etc.) to be assigned to agents.

[0025] In some traditional task distribution centers, tasks (e.g., callers) are assigned to agents in order based on arrival time, and agents receive tasks in a ranking based on when they become available. This strategy may be referred to as a "first-in, first-out," "FIFO," or "round-robin" strategy. For example, in an L2 environment, when an agent becomes available, the task at the head of the queue is selected for assignment. In other traditional task distribution centers, a performance-based routing (PBR) strategy may be implemented to prioritize task assignments to higher-performing agents. Under PBR, for example, the highest-performing agent among the available agents receives the next available task.

[0026] The present disclosure relates to optimized strategies, such as "behavior pairing" or "BP" strategies, for assigning tasks to agents that improve upon traditional assignment methods. The goal of BP is to balance agent utilization while improving overall task assignment center performance that has the potential to exceed what FIFO or PBR methods can achieve in practice. This is a remarkable achievement because BP operates on the same tasks and the same agents as FIFO or PBR methods, approximately balancing agent utilization as provided by FIFO, while improving overall task assignment center performance beyond what FIFO or PBR can provide in practice. BP improves performance by assigning agent and task pairs in a way that takes into account potential subsequent assignments of agent and task pairs, such that when the benefits of all assignments are aggregated, they are likely to exceed the benefits of FIFO and PBR strategies.

[0027] Various BP strategies can be used, such as diagonal model BP strategies or network flow (or "off-diagonal") BP strategies. These and other task allocation strategies are described in detail in the contact center context, for example, in U.S. Patent Nos. 9,300,802, 9,781,269, 9,787,841, and 9,930,180, all of which are incorporated herein by reference. BP strategies can be applied in L1 environments (agent remaining, one task; selection among multiple available / idle agents), L2 environments (task remaining, one available / idle agent; selection among multiple tasks in a queue), and L3 environments (multiple agents and multiple tasks; selection among paired permutations).

[0028] The various BP strategies discussed above can be viewed as two-dimensional (2-D), where one dimension relates to agents and the second dimension relates to tasks (e.g., callers), and each BP strategy considers information about agents and tasks to pair them. As described in detail below, embodiments of the present disclosure relate to determining BP strategies that consider higher-dimensional assignments. For a three-dimensional (3-D) example, a BP strategy might assign an agent to a task and a set of offers or actions the agent can take during task assignment. For another 3-D example, a BP strategy might assign an agent to a task and a reward (monetary or non-monetary) to be given to the agent for a given task. For a four-dimensional (4-D) example, a BP strategy might assign an agent to a task, a set of offers, and a reward.

[0029] These decision-making BP strategies may also consider historical outcome data, such as agent-task-quote, agent-task-reward, or agent-task-quote-reward pairings, to build BP models and apply BP strategies to “paired” tasks and agents and specific quote sets and / or agent compensation (throughout this specification, the noun and verb “pairing” and other forms such as “behavior pairing” may be used to describe triplets and higher-dimensional groupings).

[0030] Figure 1A block diagram of a task allocation center 100 according to an embodiment of the present disclosure is shown. The specification herein describes network elements, computers and / or components for a system and method for pairing strategies in a task allocation system that may include one or more modules. As used herein, the term "module" may be understood to refer to computing software, firmware, hardware and / or various combinations thereof. However, a module should not be interpreted as software that is not implemented on hardware, firmware or recorded on a non-transient processor-readable recordable storage medium (i.e., a module is not the software itself). Note that modules are exemplary. Modules may be combined, integrated, separated and / or replicated to support various applications. In addition, instead of or in addition to the functions performed at a specific module, the functions described herein as being performed at a specific module may be performed at one or more other modules and / or by one or more other devices. In addition, modules may be implemented across multiple devices and / or other components that are local or remote to each other. In addition, a module may be removed from one device and added to another device, and / or may be included in two devices.

[0031] like Figure 1 As shown, the task distribution center 100 may include a central switch 210. The central switch 110 may receive inbound tasks (e.g., telephone calls, internet chat sessions, emails, etc.) or support outbound connections to contacts via a dialer, telecommunications network, or other module (not shown). The central switch 110 may include routing hardware and software to facilitate routing tasks between one or more sub-centers or to one or more private branch exchanges ("PBXs") or automatic call distribution (ACD) routing components or other queuing or switching components within the task distribution center 100. If there is only one sub-center, or if there is only one PBX or ACD routing component in the task distribution center 100, then the central switch 110 may not be required.

[0032] If more than one sub-center is part of the task distribution center 100, each sub-center may include at least one switch (e.g., switches 120A and 120B). Switches 120A and 120B may be communicatively coupled to the central switch 110. Each switch in each sub-center may be communicatively coupled to a plurality of agents (or "agent pools"). Each switch may support a certain number of agents (or "seats") logged in at the same time. At any given time, a logged-in agent may be available and waiting to be connected to a task, or a logged-in agent may be unavailable for a variety of reasons, such as being connected to another contact, performing some post-call function, such as logging information about the call, or taking a break. Figure 1In the example shown in FIG, central switch 110 routes tasks to one of two sub-centers via switches 120A and 120B, respectively. Switches 120A and 120B are shown each having two agents. Agents 130A and 130B can log in to switch 120A, while agents 130C and 130D can log in to switch 120B.

[0033] The task distribution center 100 may also be communicatively coupled to integration services from, for example, third-party vendors. Figure 1 In the example of FIG, the behavior pairing module 140 can be communicatively coupled to one or more switches in the switch system of the task distribution center 100, such as the central switch 110, the switch 120A, and the switch 120B. In some embodiments, the switches of the task distribution center 100 can be communicatively coupled to multiple behavior pairing modules. In some embodiments, the behavior pairing module 140 can be embedded within a component of the task distribution center 100 (e.g., embedded in a switch or otherwise integrated with the switch).

[0034] The behavior matching module 140 can receive information about agents logged into the switch (e.g., switch 120A) (e.g., agents 130A and 130B) and information about inbound tasks via another switch (e.g., central switch 110), or in some embodiments, receive information from a network (e.g., the Internet or a telecommunications network) (not shown). The behavior matching module 140 can process this information to determine which agents should be paired (e.g., matched, assigned, distributed, routed) with which tasks and other dimensions (e.g., offers, actions, channels, non-monetary rewards, monetary rewards or compensation, etc.).

[0035] For example, in the L1 state, multiple agents are available and waiting to connect with contacts, and tasks arrive at the task distribution center 100 via the network or central switch 110. As described above, without the behavioral matching module 140, the switch typically automatically distributes new tasks to: under a FIFO strategy, the available agent who has been waiting the longest; or under a PBR strategy, the available agent identified as the highest-performing agent. Under the behavioral matching module 140, contacts and agents can be assigned scores (e.g., percentiles or percentile ranges / bandwidths) based on a matching model or other artificial intelligence data model, so that tasks can be matched, paired, or otherwise connected to preferred agents. High-dimensional analysis of BP decisions will be described in more detail below.

[0036] In the L2 state, multiple tasks are available and waiting to be connected to an agent, and an agent becomes available. These tasks may be queued in a switch, such as a PBX or ACD device. Without the behavioral matching module 140, when agent selection is unavailable, as in a FIFO strategy or a PBR strategy, the switch will typically connect the newly available agent to the task that has been waiting the longest in the queue. In some task distribution centers, as described above, priority queues may also be incorporated. When the behavioral matching module 140 is present in this L2 scenario, as in the L1 state described above, tasks and agents may be assigned percentiles (or percentile ranges / bandwidths, etc.) based on a model, such as an artificial intelligence model, so that the agent that becomes available can be matched, paired, or otherwise connected to the preferred task. The high-dimensional analysis of BP decisions will be described in more detail below.

[0037] Figure 2 A block diagram of a task assignment system 200 according to an embodiment of the present disclosure is shown. The task assignment system 200 can be included in a task assignment center (e.g., the task assignment center 100), or incorporated into a component or module of the task assignment center (e.g., the behavior pairing module 140) to help assign agents between various tasks and other dimensions for grouping.

[0038] The task assignment system 200 may include a task assignment module 210 configured to pair (eg, match, assign) an incoming call task with an available agent. (High-dimensional analysis of BP decision making will be described in more detail below.) Figure 2 In the example of , m tasks 220A-220m are received in a given time period, and during the given time period, n agents 230A-230n are available. Each of the m tasks can be assigned to one of the n agents for service or other types of task processing. Figure 1 In the example of , m and n can be arbitrarily large finite integers greater than or equal to 1. In a real-world task distribution center (such as a contact center), there may be tens, hundreds, etc. of agents logged into the contact center to interact with contacts during a shift, and during a shift, the contact center may receive tens, hundreds, thousands, etc. of contacts (e.g., phone calls, internet chat sessions, emails, etc.).

[0039] In some embodiments, task assignment strategy module 240 can be communicatively coupled to task assignment system 200 and / or configured to operate within task assignment system 200. Task assignment strategy module 240 can implement one or more task assignment strategies (or "pairing strategies") for assigning individual tasks to individual agents (e.g., pairing contacts with contact center agents). Task assignment strategy module 240 can design and implement a variety of different task assignment strategies. In some embodiments, a FIFO strategy can be implemented, where, for example, the agent with the longest wait time receives the next available task (in an L1 environment) or the longest wait time task is assigned to the next available agent (in an L2 environment). In other embodiments, a PBR strategy can be implemented, whereby higher-performing agents are prioritized for task assignment. For example, under PBR, the highest-performing agent among available agents receives the next available task. In other embodiments, a BP strategy can be used to optimally assign tasks to agents using information about tasks, agents, or both. Various BP strategies can be used, such as a diagonal BP strategy or a network flow ("off-diagonal") BP strategy. See U.S. Patent Nos. 9,300,802, 9,781,269, 9,787,841, and 9,930,180.

[0040] In some embodiments, when a task is assigned to a particular agent, the task assignment strategy module 240 can implement a decision-making BP strategy that considers suboptimal actions for the task. For a task-agent pair, the decision-making BP strategy can also assign available actions or action sets to the agent to complete the task. In the context of a contact center system, the actions or action sets can include quotes or quote sets that the agent can present to the customer. For example, in a contact center system, the decision-making BP strategy can pair a contact with an agent and a quote or quote set available for the agent to make for the customer based on the expected outcome of the contact-agent interaction using that particular quote or quote set. By influencing the choices or options among the quotes available to the agent, the decision-making BP strategy goes beyond pairing a contact with an agent by optimizing the outcome of a single interaction between the agent and the contact.

[0041] For example, if agent 230A loves sports, they may be more adept at selling sports packages. Therefore, sports packages may be included in agent 230A's offer set for some or all contact types. On the other hand, agent 230B may enjoy movies and be more adept at selling premium movie packages; therefore, premium movie packages may be included in agent 230B's offer set for some or all contact types. Furthermore, based on artificial intelligence processing such as machine learning, a decision-making BP model can automatically segment customers based on various variables and data types. For example, a decision-making BP model may recommend offering a package that includes sports and is suitable for a specific age range, income range, or other demographic or psychographic factors to a first type of customer ("Customer Type 1"). A decision-making BP model may recommend offering a premium movie package that is suitable for a different age range, income range, or other demographic or psychographic factors to a second type of customer ("Customer Type 2"). The decision-making BP strategy may preferentially pair Customer Type 1 with Agent 230A and an offer set featuring sports packages, and Customer Type 2 with Agent 230B and an offer set featuring premium movie packages.

[0042] As with previously disclosed BP strategies, the decision-based BP strategy optimizes the overall performance of the task assignment system, rather than optimizing each individual immediate task-agent pairing. For example, in some embodiments, the decision-based BP system will not always offer a campaign to customer type 1, nor will it always give agent 230A the option of quoting a deal based on a campaign package. This situation may arise when the marketing department of the company operating the contact center system may have a limited budget, a limited number of deals (e.g., a limited number of discounted campaign packages), other constraints on the frequency of certain offers, a limit on the total amount of discounts offered in a given time period (e.g., for any discount or discount package), and so on. Similarly, a campaign-based deal may sometimes be offered to customer type 2, while agent 230B may sometimes be given the option of quoting a deal based on a campaign package.

[0043] To optimize the overall performance of the task assignment system, the decision-making BP strategy can take into account all types of customers waiting in the queue, the agents available to the customers, and any other dimensions used for pairing (such as the number and type of remaining quotes, agent compensation or other non-monetary incentives, suboptimal actions, etc.). In some embodiments, a probability distribution can be assigned based on the likelihood that an inbound task or customer type will accept a given quote level based on the agent being paired with the task or customer.

[0044] For example, for customer type 1, if the quote has a 0% discount, the customer has a 0% probability of accepting the quote from the average agent, and specifically, a 0% probability of accepting the quote from agent 230A and agent 230B. For a 20% discount quote, the customer might have a 30% probability of accepting the quote from the average agent, a 60% probability of accepting the quote from agent 230A, and a 25% probability of accepting the quote from agent 230B. In a scenario where the average agent, agent 230A, and agent 230B are all assigned to a queue and available, it is possible that agent 230A performs much better than either the average agent or agent 230B.

[0045] In some embodiments, output volume measurements can be attached to each task before and after an interaction with an agent. For example, revenue figures can be attached to each caller before and after a call. The decision-making BP system can measure changes in revenue and the influence of the caller based on the quote or set of quotes provided by the agent. For example, regardless of the proposed discount or the ability of the individual agent, customer type 1 is more likely to renew their existing plan. Customer type 2 may be preferably assigned to a lower-performing agent in the quote set with a higher discount cap. Conversely, if customer type 2 is paired with a higher-performing agent or an agent authorized to provide a larger discount, she is more likely to upgrade her plan.

[0046] In some embodiments, the decision-making BP strategy can perform sequential pairing of one or more dimensions in any order. For example, the decision-making BP strategy can first pair agents with tasks, then pair quote sets with agent-task pairs, then pair rewards with agent-task-quote set pairs, and so on.

[0047] In other embodiments, the decision-making BP strategy may perform "fully coupled" simultaneous multi-dimensional pairing. For example, the decision-making BP strategy may consider all dimensions at once to select the best 4D agent-task-offer-reward pair.

[0048] The same task may arrive at the task assignment system multiple times (e.g., the same caller calling the call center multiple times). In these "multi-touch" scenarios, in some embodiments, the task assignment system can consistently assign the same items to one or more dimensions to promote consistency. For example, if a task is paired with a specific quote set the first time it arrives, then each time the task arrives at a subsequent time (e.g., for a given issue, within a given time period, etc.), the task will be paired with the same quote set.

[0049] In some embodiments, the historical assignment module 250 can be communicatively coupled to the task assignment system 200 and / or configured to operate within the task assignment system 200 via other modules such as the task assignment module 210 and / or the task assignment strategy module 240. The historical assignment module 250 may be responsible for various functions, such as monitoring, storing, retrieving, and / or outputting information about task-agent assignments and higher-dimensional assignments that have been made. For example, the historical assignment module 250 can monitor the task assignment module 210 to collect information about task assignments within a given time period. Each record of a historical task assignment can include information such as an agent identifier, a task or task type identifier, a quote or quote set identifier, outcome information, or a pairing strategy identifier (i.e., an identifier indicating whether a BP strategy, a decision BP strategy, or some other pairing strategy such as a FIFO or PBR pairing strategy was used for the task assignment).

[0050] In some embodiments and for certain contexts, additional information may be stored. For example, in the context of a call center, the historical assignment module 250 may also store information regarding the call start time, call end time, the phone number dialed, and the caller's phone number. For another example, in the context of a dispatch center (e.g., a "truck roll"), the historical assignment module 250 may also store information regarding the time the driver (i.e., the live agent) left the dispatch center, the recommended route, the route taken, the estimated travel time, the actual travel time, the amount of time spent at the customer site handling the customer task, and so on.

[0051] In some embodiments, the historical allocation module 250 can generate a pairing model, a decision BP model, or a similar computer processor-generated model based on a set of historical allocations within a period of time (e.g., the past week, the past month, the past year, etc.), which can be used by the task allocation strategy module 240 to make task allocation recommendations or instructions to the task allocation module 210.

[0052] In some embodiments, instead of relying on a predetermined offer set in generating a decision-making BP model, the historical allocation module 250 can analyze historical outcome data to create or determine a new or different offer set, which can then be incorporated into the decision-making BP model. This approach is preferred when there are budget constraints or other restrictions on the number of specific offer sets that can be made. For example, a marketing department may limit a contact center system to 500 discounted sports packages and 500 discounted movie packages per month. The company may want to optimize total revenue, regardless of how many sports and movie packages are sold with or without discounts. In this case, the decision-making BP model may be similar to the previously disclosed BP diagonal model, except that in addition to the "task or contact percentile" (CP) dimension and the "agent percentile" (AP) dimension, a third "revenue or offer set percentile" dimension is included. Furthermore, all three dimensions can be normalized or processed using mean regression (e.g., Bayesian mean regression (BMR) or hierarchical BMR).

[0053] In some embodiments, the historical assignment module 250 can generate a decision-making BP model that optimizes task-agent-quote set pairings based on single-channel or multi-channel interactions. The historical assignment module 250 can treat different channels differently. For example, the decision-making BP model can pair a contact with different agents or quote sets depending on whether the contact called a call center, initiated a chat session, sent an email or text message, entered a retail store, etc.

[0054] In some embodiments, the task assignment strategy module 240 can proactively create tasks or other actions (e.g., recommending outbound contact interactions, suboptimal actions, etc.) based on information about contacts or customers, available agents, and available offer sets. For example, the task assignment system 200 may determine that a customer's contract is set to expire, that the customer's usage is declining, or that the customer's credit rating is declining. The task assignment system 200 may further determine that the customer is unlikely to renew their contract at the customer's current rate (e.g., based on information from the historical assignment module 250). The task assignment system 200 may determine that the suboptimal action is to call the customer (contact selection, channel selection, and time selection), contact a specific agent (agent selection), and offer the agent the option to downgrade to a specific discount or range of discounts (offer set selection). If the customer does not reach an agreement during the call, the task assignment system 200 may further determine that the customer is more likely to accept the downgrade discount offer if the agent follows up with a text message with information about the discount and how to confirm it (multi-channel selection and optimization).

[0055] In some embodiments, similar to how the Kappa (κ) parameter is used to adjust / skew agent percentiles or percentile ranges (see U.S. Patent No. 9,781,269) and the Rho (ρ) parameter is used to adjust / skew task or contact percentiles or percentile ranges (see U.S. Patent No. 9,787,841), the task assignment strategy module 240 can apply the Iota (Iota) parameter to a third (or higher) dimension, such as the quote set percentile or percentile range in a decision-making BP strategy. Using the Iota parameter, the task assignment strategy module 240 can, for example, adjust the quote set percentile or percentile range (or other dimension) to bias task-agent-quote set pairings toward higher-performing quotes and unbalance quote set availability. The Iota parameter can be applied in either L1 or L2 environments and can be used in conjunction with the Kappa or Rho parameter, or can be applied in conjunction with both the Kappa and Rho parameters in an L3 environment. For example, if the task assignment strategy module 240 determines that a contact's expected wait time has exceeded 100 seconds (high congestion), it can apply the Iota parameter, making it more likely that the agent has a higher discount available for an offer, which can be sold more quickly to reduce congestion and expected wait time. More generous offers can expedite the task-agent interaction, thereby reducing average handle time (AHT). When congestion is low, expected wait time is also low, and the Iota parameter can be adjusted to make only less generous offers available. These calls may take longer, and AHT may increase, but sales and revenue may also increase.

[0056] In some embodiments, the task assignment strategy module 240 can optimize performance by simultaneously optimizing for multiple business goals or other objectives. In situations where the objectives compete with each other (e.g., discount volume and retention rate), the module 240 can balance the tradeoffs between the two objectives. For example, the task assignment strategy module 240 can balance increasing (or maintaining) revenue with maintaining or minimizing retention rate, or it can balance reducing (or maintaining) AHT with increasing (or maintaining) customer satisfaction, etc.

[0057] In some embodiments, the task assignment strategy module 240 can implement a decision-making BP strategy that considers agent compensation instead of quotes or quote sets. This framework is similar to the one described above, except that instead of influencing customers through quotes or quote sets, this decision-making BP strategy influences agent performance through the compensation that the agent can receive. In other words, instead of a three-dimensional pairing of task, agent, and quote set, this decision-making BP strategy makes a three-dimensional pairing of task, agent, and reward. In some embodiments, the decision-making BP strategy can make a four-way pairing of task, agent, quote, and reward.

[0058] Decision-based BP strategies that provide variable agent compensation based on task-agent pairings may result in greater transparency and fairness. For example, some task-assignment systems (e.g., contact center) may experience a mix of more challenging and less challenging contacts and employ a mix of higher-performing and lower-performing agents. Under a FIFO or PBR strategy, agents of all abilities are equally likely to be paired with more or less challenging contacts. Under a FIFO strategy, the overall performance of the contact center system may be lower, but average agent compensation may be transparent and fair. Under a PBR strategy, agent utilization may be skewed, and compensation may also be skewed toward higher-performing agents. Under the previously disclosed BP strategy, more challenging contact types may be preferentially paired with higher-performing agents, while less challenging contact types may be preferentially paired with lower-performing agents. For example, if high-performing agents, on average, experience more difficult calls, this "call type skew" may result in lower conversion rates and compensation for high-performing agents.

[0059] Therefore, adjusting agent compensation higher or lower for a given task (or contact) type may improve the fairness of compensation under previously disclosed BP policies. When an agent is paired with a task or contact, a decision-making BP policy can inform the agent of the contact's expected value to the contact center and / or the commission or other non-monetary reward the agent will receive for handling the contact or achieving a specific outcome. A decision-making BP policy may influence agent behavior by proposing variable compensation. For example, if a decision-making BP policy determines that an agent should handle a task or contact quickly (lower AHT, lower revenue), the decision-making BP policy may select lower compensation. Consequently, the agent may have less incentive to spend a significant amount of time to earn a relatively low commission. Conversely, if a decision-making BP policy determines that an agent should focus their efforts on higher-value calls (higher AHT, higher revenue), it may select higher compensation. Such a decision-making BP policy can maximize agent rewards while improving the overall performance of task assignment system 200.

[0060] The historical assignment module 250 can model this decision-making BP model based on historical information about task type, agent, and compensation amount, thereby making simultaneous selections of task, agent, and reward. To improve the overall performance of the task assignment system 200, the amount of upward or downward compensation change can vary and depends on each combination of individual agent and contact type.

[0061] Similar to applying the Iota parameter to the bid set or suboptimal action percentile or percentile range dimensions, and as described above, the task assignment strategy module 240 can apply the Iota parameter to other dimensions, such as skewing agent compensation to a greater or lesser degree, or to generally higher or lower values. In some embodiments, the amount and type of Iota parameter applied to agent compensation or other non-monetary incentives can be based at least in part on factors in the task assignment system 200 (e.g., the expected wait time for callers in the call center).

[0062] In some embodiments employing a strategy similar to the diagonal model BP strategy, variable compensation can be viewed as temporarily influencing the available agent's active agent percentile (AP) higher or lower in order to move the available contact-agent pairings closer to the optimal diagonal. Similarly, adjusting the offer value higher or lower can be viewed as influencing the waiting contact's active contact percentile (CP) higher or lower in order to move the available contact-agent pairings closer to the optimal diagonal.

[0063] In some embodiments, benchmarking module 260 can be communicatively coupled to task assignment system 200 and / or configured to operate within task assignment system 200 via other modules, such as task assignment module 210 and / or historical assignment module 250. Benchmarking module 260 can benchmark the relative performance of two or more pairing strategies (e.g., FIFO, PBR, BP, Decisive BP, etc.) using historical assignment information received from, for example, historical assignment module 250. In some embodiments, benchmarking module 260 can perform other functions, such as establishing a benchmarking schedule for cycling through various pairing strategies, tracking groups (e.g., a base group and a measurement group of historical assignments), and the like. Benchmarking is described in detail for contact center contexts, for example, in U.S. Patent No. 9,712,676, which is incorporated herein by reference.

[0064] In some embodiments, the benchmarking module 260 can output or otherwise report or use relative performance measures. The relative performance measures can be used to evaluate the quality of a task allocation strategy to determine, for example, whether a different task allocation strategy (or a different pairing model) should be used, or to measure the overall performance (or performance gain) achieved within the task allocation system 200 when the task allocation system 200 is optimized or otherwise configured to use one task allocation strategy instead of another.

[0065] In some embodiments, the benchmarking module 260 can benchmark the decision-making BP strategy against one or more alternative pairing strategies (such as FIFO) in conjunction with offer set availability. For example, in a contact center system, agents may have a matrix of nine offers—three tiers of service level, each with three discount levels. During "closed" calls, the agent with the longest wait time may be connected to the caller with the longest wait time and may offer any of the nine offers. Higher-performing agents are more likely to sell higher tiers of service at higher prices, while lower-performing agents may be less aggressive and offer the largest discount immediately. During "on" calls, the decision-making BP strategy may pair contacts with agents but restrict them to a subset of the nine available offers. For example, for certain contact types, if the task assignment strategy module 240 determines, based on the decision-making BP model, that the overall performance of the contact center system can be optimized by selectively restricting the offer set in a given manner for a given contact-agent pairing, the higher-performing agent may be authorized to make any of the nine offers, while the lower-performing agent may be restricted to offering smaller discounts for certain tiers. Additionally, if the provider (eg, vendor) of the decision-making BP strategy to the task assignment system uses a benchmarking and revenue sharing business model, the provider can contribute a share of the benchmarked revenue gains to the agent compensation pool.

[0066] In some embodiments, the task assignment system 200 may provide dashboards, visualizations, or other analytics and interfaces to improve the overall performance of the system. For each agent, the analytics provided may vary based on the agent's relative abilities or behavioral characteristics. For example, competitive or higher-performing agents may benefit from ranking tools or other "gamification" elements (e.g., badges or achievements that unlock points and scoreboards, notifications when an agent surpasses another agent in the rankings, etc.). On the other hand, less competitive or lower-performing agents may benefit from periodic encouragement messages, recommendations for training / educational courses, etc.

[0067] Figure 3 A task allocation method 300 according to an embodiment of the present disclosure is shown.

[0068] Task assignment method 300 may begin at block 310. At block 310, task assignment method 300 may rank a plurality of tasks that may be queued in a task assignment system (e.g., task assignment system 200). Task assignment method 300 may rank the plurality of tasks by assigning the plurality of tasks percentiles or percentile ranges based on, for example, a model such as an artificial intelligence model. In a contact center system, contacts may be ranked based on how long each contact has waited to be assigned to an agent relative to other contacts, or based on how much each contact contributes to the performance of the contact center system relative to other contacts, according to some metric. In other embodiments, the plurality of tasks may be analyzed based on an "off-diagonal" model (e.g., a network traffic model).

[0069] The task assignment method 300 may then proceed to block 320. At block 320, the task assignment method 300 may rank the plurality of agents available in the task assignment system. The task assignment method 300 may rank the plurality of agents by assigning them percentiles or percentile ranges based on a model, such as an artificial intelligence model. In a contact center system, agents may be ranked based on how long each agent has waited to be assigned to a contact relative to other agents, or based on how much each agent contributes to the performance of the contact center system relative to other agents, according to some metric. In other embodiments, the plurality of tasks may be analyzed based on a network flow model.

[0070] The task assignment method 300 may then proceed to block 330. At block 330, the task assignment method 300 may pair a task from the plurality of tasks with an agent from the plurality of agents. In some embodiments, the task-agent pairing may be based at least in part on a ranking of the plurality of tasks, a ranking of the plurality of agents, and compensation to be received by at least one agent from at least one or more of the plurality of bids to be proposed to the plurality of tasks. In some embodiments, the task-agent pairing may be based on alternative behavioral pairing techniques, such as a network flow model.

[0071] In some embodiments, the task assignment method 300 can then select a quote or quote set and / or agent reward to be used in conjunction with the task-agent pairing. In other embodiments, the task assignment method 300 can perform three-way pairing (e.g., task-agent-quote, task-agent-reward, etc.) or four-way pairing (e.g., task-agent-quote-reward). A given quote may skew or adjust a percentile or percentile range for a task, while a given compensation may skew or adjust a percentile or percentile range for an agent. Thus, by considering at least one quote or compensation, the task assignment method 300 is able to select a task-agent (or task-agent-quote, task-agent-compensation, task-agent-quote-compensation, etc.) pairing, thereby improving the overall performance of the task assignment system.

[0072] At this point, it should be noted that the task allocation according to the present disclosure as described above may involve, to some extent, processing input data and generating output data. This input data processing and output data generation can be implemented in hardware or software. For example, specific electronic components may be used in a behavior pairing module or similar or related circuits to implement the functions associated with the task allocation according to the present disclosure as described above. Alternatively, one or more processors operating according to instructions may implement the functions associated with the task allocation according to the present disclosure as described above. If this is the case, it is also within the scope of the present disclosure that these instructions may be stored on one or more non-transitory processor-readable storage media (e.g., disks or other storage media), or transmitted to one or more processors via one or more signals embedded in one or more carrier waves.

[0073] The present disclosure is not limited to the scope of the specific embodiments described herein. In fact, in addition to those described herein, various other embodiments of the present disclosure and improvements thereof will be apparent to those of ordinary skill in the art from the above description and the accompanying drawings. Therefore, these other embodiments and improvements are intended to fall within the scope of the present disclosure. In addition, although the present disclosure is described herein in the context of at least one specific embodiment in at least one specific environment for at least one specific purpose, those of ordinary skill in the art will recognize that its usefulness is not limited thereto and that the present disclosure can be advantageously implemented in a variety of environments for various purposes. Therefore, the following claims should be interpreted in view of the full breadth and spirit of the present disclosure as described herein.

Claims

1. A method for determining behavior pairing in a contact center system, the method comprising: determining, by at least one computer processor communicatively coupled to the contact center system and configured to operate in the contact center system, at least one agent available for pairing; determining, by the at least one computer processor, at least one task available for pairing; determining, by the at least one computer processor, a set of one or more subsets of items according to the dimension; adjusting, by the at least one computer processor, a parameter, wherein a first value of the parameter affects a likelihood of selection of each subset of items of the collection that is paired with the one or more subsets of items; as well as At least one task-agent-item subset pairing of the set of the at least one agent, the at least one task, and the one or more item subsets is selected, by the at least one computer processor, for allocation in the contact center system based at least in part on the adjusted parameters.

2. The method of claim 1, wherein: The dimension includes one of: offer, agent compensation, action, and channel.

3. The method of claim 1, wherein: The dimension is based on the at least one agent and / or the at least one task.

4. The method of claim 1, wherein: Adjusting the parameter includes adjusting a percentile or a percentile range of items in the set of the one or more item subsets to skew the item pairings of the at least one task-agent-subset toward higher performing quotes and unbalanced quote set availability.

5. The method of claim 1, wherein: If the expected waiting time for the at least one task exceeds a threshold, the parameter is adjusted so that the at least one agent has a more generous offer available, thereby reducing congestion and the expected waiting time.

6. The method of claim 1, wherein: If the expected waiting time for the at least one task is less than a threshold, the parameter is adjusted so that the at least one agent has a less generous offer available.

7. The method of claim 1, wherein: The parameter is adjusted to balance the average handling time with the generosity level of the available quotes for the at least one agent.

8. A system for determining behavior pairing in a contact center system, comprising: at least one computer processor communicatively coupled to the contact center system and configured to operate in the contact center system, wherein the at least one computer processor is further configured to: determining at least one agent available for pairing; determining at least one task available for pairing; Determine a collection of one or more subsets of items based on the dimensions; adjusting a parameter, wherein a first value of the parameter affects a likelihood of selection of each subset of items of the collection that is paired with the one or more subsets of items; At least one task-agent-item subset pairing of the set of the at least one agent, the at least one task, and the one or more item subsets is selected for allocation in the contact center system based at least in part on the adjusted parameters.

9. The system of claim 8, wherein: The dimension includes one of: offer, agent compensation, action, and channel.

10. The system of claim 8, wherein: The dimension is based on the at least one agent and / or the at least one task.

11. The system of claim 8, wherein: Adjusting the parameter includes adjusting a percentile or a percentile range of items in the set of the one or more item subsets to skew the item pairings of the at least one task-agent-subset toward higher performing quotes and unbalanced quote set availability.

12. The system of claim 8, wherein: If the expected waiting time for the at least one task exceeds a threshold, the parameter is adjusted so that the at least one agent has a more generous offer available, thereby reducing congestion and the expected waiting time.

13. The system of claim 8, wherein: If the expected waiting time for the at least one task is less than a threshold, the parameter is adjusted so that the at least one agent has a less generous offer available.

14. The system of claim 8, wherein: The parameter is adjusted to balance the average handling time with the generosity level of the available quotes for the at least one agent.

15. An article for determining action pairing in a contact center system, the article comprising: a non-transitory processor-readable medium; as well as instructions stored on said medium; wherein the instructions are configured to be readable from the medium by at least one computer processor communicatively coupled to the contact center system and configured to operate in the contact center system, thereby causing the at least one computer processor to operate to: determining at least one agent available for pairing; determining at least one task available for pairing; Determine a collection of one or more subsets of items according to a dimension; adjusting a parameter, wherein a first value of the parameter affects a likelihood of selection of each subset of items of the collection that is paired with the one or more subsets of items; At least one task-agent-item subset pairing of the set of the at least one agent, the at least one task, and the one or more item subsets is selected for allocation in the contact center system based at least in part on the adjusted parameters.

Images