The DUC-DOVE Method
An Approach to Resource Requirements Analysis and Management

Brian J. Mistler, Ph.D.

Humboldt State University

[PUBLIC DRAFT - Last Updated 2017-10-01]

Overview of Concepts and Terms

The DUC-DOVE Method is so termed for its seven principal components, Demand, Utilization, and Capacity, as well as the key Disparities between these items and Outcomes that result from failing to resolve such disparities, how such current or potential outcomes align with a hierarchy of Values, and subsequently Executing the values-based strategic plan to achieve the desired outcomes.It also includes necessarily an analysis of the factors which produce or limit capacity (e.g. in services analysis, return on investment (ROI), staffing, facilities, knowledge, or other capacity limitations). The DUC-DOVE Method is one approach to capturing key factors necessary for resource requirements analysis, summary, and allocation of resources decisions especially in service-related fields.

Demand or “Demand for Services” summaries some quantifiable measure of the requests for services. As utilization is naturally capped by both capacity and awareness of services, tracking changes in utilization alone is as a poor indicator of the true demand for services, and fairly meaningless predictor of potential future utilization should capacity or awareness be altered. In the case of health-service delivery where demand exceeds

Population Demand Maximum or “Need” captures the total potential demand for services in a given population. For purposes of nonprofit service delivery and the analysis described herein aimed at tertiary intervention, need can be assumed to be externally determined an unmodified by services. However, these same principles can be used when engaging in primary and secondary prevention to understand population level changes. When using sampling plus projection to determine future need, if the need is assumed to be externally determined, unmodified by services, and occurring at a fixed percentage, the Population Demand Maximum in a given time period is then, of course, equal to:

(total current population size) x (population growth factor)
x (percentage of individuals with the need)

Modeling changes to the base-rate of the demand maximum in a population is simply a matter of reproducing this model recursively -- taking there to be some higher-order limiting factor and modeling the population demand maximum in one model as the demand in another; this beyond the scope of the current overview.

Utilization in its simplest form is the number of times a particular service is provided. Where multiple services are aggregated together, utilization should be expressed in roughly equivalent common units (e.g. number of visits). In this model utilization and demand are defined such that utilization can never exceed demand.

Capacity or “Reasonable Capacity” when charted/graphed in this model should capture “reasonable capacity” as a final conclusion, having properly accounted for capacity benchmarking and the desired amount of discretionary capacity, rather than “physical capacity maximum”. Utilization may slightly exceed capacity, especially for short periods of time. When utilization exceeds reasonable capacity, the risk of lower quality service including errors and poor customer service increase.

Physical Capacity Maximum is the absolute upper bounds of capacity, beyond which it’s impossible to further increase output. This level of performance is actually undesirable as a framework in most circumstances, though it can be estimated in controlled environments. As reasonable capacity approaches Physical Capacity maximum for sustained periods of time, staff are likely to experience burn-out and error rates are likely to increase.

Overage Percentage can be calculated as the percentage of time Utilization of Services exceeds reasonable capacity.

Capacity Benchmarking across other similar environments has several advantage and disadvantages. All forms of capacity are inherently variable in any industry involving both humans and problems of varying complexity. For example, if we ask how long does it take to fix a computer problem and then try to calculate how many computer problems a given technician can address in a day, we’re stuck not knowing things like how difficult the computer problems will be or how quick a particular technician will be. To achieve a capacity measure that can be applied uniformly for making projections, we take an average as a starting point for both (e.g. the average problem is of x complexity and the average technician can address 2 of these problems in an hour). But, the average where -- how do we know if our average is reasonable or not? The largest advantage of capacity benchmarking is it helps to adjust reasonable per-provider capacity expectations if an entire team of providers is under-performing. As a caution, however, capacity benchmarking must also take into account context and any systematic influences on problem complexity or operation conditions -- a remote computer repair shop in the freezing conditions of Antarctica next to a military training base, might see a disproportionately high number of computers with all sorts of severe damage requiring a much higher average time that has to be addressed in more difficult conditions, than technicians in a computer repair shop in a mall that spend half their day quickly replacing batteries, and referring the most complicated problems to specialists six doors down.

Discretionary Capacity Factors are elements of capacity that can be used to adjust capacity by skipping or adding steps which are “nonessential” to the primary task. For example, the time it takes to repair a phone could be shortened or lengthened depending on if I clean the phone’s screen before returning it to the customer. Technicians may be able to repair 12 phones a day without cleaning each, but only 11 a day when they take the time to clean them before returning them to the customer. Perhaps they could do 13 a day if they skipped all pleasantries entirely in their interactions with customers and focused entirely on the technical aspects. However, we can quickly imagine the impact of both of these decisions on customer satisfaction and return business. As customer service expectations increase -- because of competition for services or the expectations of a client base -- discretionary capacity factors should be considered when determining reasonable capacity.

Connected Capacity Impact Weights are useful when multiple pieces of a team work together to establish capacity. For example, in a health setting, capacity is clearly related to the availability of providers (i.e. 4 doctors have more capacity than 3), as well as the team of personnel which work together to provide services -- nurses, medical assistance, medical records support personnel, administrators, and so on -- who each make an impact on capacity. A team with 10 doctors will clearly have more capacity than a team with 5 doctors. And, a team of 5 doctors and 3 nurses we would understand to have more capacity than a team of 5 doctors and no nurses. However, how can we aggregate capacity into a single number across a team -- how does the capacity of a six-person team of 4 doctors and 2 nurses compare to a six-person team of 3 doctors and 3 nurses? To address this complexity while still allowing a single aggregate total capacity, resources of varying types which contribute to overall capacity can assigned weighted value based on their proportional impact on the specific service being examined. Note well that this number does not capture in anyway the overall value of the role to the organization, the ideal balance of team composition (which is more of an art requiring domain and context specific expertise), or account for “catastrophic” minimum thresholds (e.g. a team of 8 nurses might be efficient but simply couldn’t function legally without a medical doctor, and 8 doctors would be unable to help patients if there weren’t medical assistants or medical records support staff in sufficient proportion to room patients and track them). But, will all of these things in place, connected capacity weights (e.g. assigning an MD a relative Connected Capacity Impact Weight of 8 and a Nurse 5, can help quantify total capacity for connected teams. These weights can be established through domain-specific expertise and refined empirically by  looking at varying utilization over time as staffing models shift slightly.

Disparity is the gap (calculated as the arithmetic difference between two items) between any of our metrics for Demand, Utilization, and Capacity, and its these gaps that help point to future actions. Key disparities produce important information:

Demand-Utilization/Demand-Capacity Disparity is present when Demand significantly exceeds Utilization and/or Capacity. When Service Demand exceed Utilization, it generally because Capacity is limiting utilization, and suggests the following objective to resolve the disparity:

-> Increase Capacity

Population Demand Maximum-Service Demand Disparity is present when the identified need in the population exceed the demand for services. This is generally due to either insufficient marketing or a decision on the part of consumers or providers to self-limit further attempts to increase service demand due to capacity limitations, stigma, costs of services, or other barriers to service utilization. As capacity and marketing are increased, without any other inhibiting factors, service demand will approach population demand maximum asymptotically. To resolve this disparity:

-> Increase Marketing

-> Decrease Barriers

Utilization-Capacity Disparity is present when utilization exceed reasonable capacity. As this disparity increases the risk of lower quality service including errors and poor customer service increase. Capacity can either be increased to meet demand or marketing can be used/reduced to decrease demand.

-> Increase Capacity

-> Decrease Marketing

Capacity-Utilization/Demand Disparity is present when capacity greatly exceed demand for services or utilization of services, and indicates services are being underused and resources are being over-allocated. Some target Capacity-Utilization or Capacity-Demand Gaps may be created intentionally in cases where a wait time for services is less acceptable (i.e. urgent care or other health services, utility repair services), however in all cases a reasonable upper bound must be identified based on probabilities (e.g. even Emergency Rooms or 911 services can’t be designed to handle certain uncommonly large events). Capacity-Demand Disparity is similar to Capacity-Utilization disparity, except even more conclusive. When Capacity exceed Demand for Services, it should be determined whether or not to increase marketing or decrease capacity. When Capacity exceeds Population Demand Maximum, Capacity should definitively be lowered.

-> Decrease Capacity

-> Increase Marketing

Outcomes are those impacts on the client which have a somewhat knowable mathematical relationship to the other factors (Demand, Utilization, Capacity, and Disparities). In service delivery these may include things like wait-times in days or weeks for appointments, average wait time in minutes or hours for a walk-in appointment, total length of a visit (if there are multiple resources required in succession to provider a service), or peak times where there is high variability in demand.

Values determine which outcomes are desirable, acceptable, or unacceptable, and in a limited resource model a hierarchy of values helps to determine what decisions to make when two outcomes come into conflict. This includes return on investment (ROI) from both a financial and institutional values perspective.  While this model approaches values in the decision making phase, the most successful organizations will have clearly defined their values at the beginning, and used them to guide areas more critical for DUC-DOVE analysis.

Summary

By examining Demand, Utilization, and Capacity, as well as the key Disparities between these items and the Outcomes that result from failing to resolve such disparities, and an individual/groups/institution's hierarchy of Values,  it is possible to determine whether to increase or decrease capacity, marketing, or to address other external barriers, as well as weighing the trade-offs in setting reasonable capacity. The final step is executing objectives to change the situation using values-based strategic plan to achieve the desired outcomes. Through repated collection of data on Demand, Utilization, and Capacity the effectiveness of the plan can be monitored iteratively.  The DUC-DOVE Method is one approach to capturing key factors necessary for resource requirements analysis, summary, and budgeting or other allocation of resources decisions in service-related fields, and may be useful on its own or in conjunction with others (e.g. SWOT, Value-Proposition Analysis, Free-Market, Resource-Demand Flattening, etc.) as part of the strategic planning process.

 © 2017 Brian Mistler. Free use for all educational or nonprofit purposes without written permission but with attribution. All other rights reserved.