Agile enterprise performance management using the example of a call center
Many companies use call centers to provide their customers with the best possible customer service, to expand telephone sales or to offer additional services. Call centers are often criticized for not being efficient enough and for missing cost or contribution margin targets. Therefore, an intensive analysis of optimization potentials is worthwhile. Efficiency improvements can be brought about in a wide variety of ways – in this example, we present a value driver model that can be used to analyze financial improvement opportunities and thus show an example of agile enterprise performance management.
Reduction of effort through the use of value driver models
Value drivers have the character of operational levers. These enable financial (corporate) management at a higher level. The focus is no longer on many details; rather, only significant influencing factors are included. In the context of corporate planning, value driver models can be used to derive goals directly from corporate strategy. Through additional integration of (strategic) measures, options for action and their effect on important financial target figures are shown transparently.
The structure of a value driver tree is quickly explained. It consists of data, operators and KPIs.
- Die Daten besitzen, jedes Datenelement für sich, Merkmalsausprägungen. Diese können Zeit, Lokation, Produkttyp oder andere Klassifizierungen sein. Anhand dieser Ausprägungen werden die Daten tabellarisch geordnet und können getrennt betrachtet oder aggregiert werden.
- In den Operatoren werden die Daten nach entsprechender Vorgabe verarbeitet. Typische mathematische Operationen, die durchgeführt werden, sind Additionen und Multiplikationen. Es sind jedoch auch komplexere Beziehungen möglich, die mit Formeln beschrieben werden.
- Die Zielgrößen eines Treibermodells, KPIs (Key Performance Indicators, Kennzahlen), werden aus den Datenelementen und den Operatoren berechnet. Man kann sich das Ergebnis dann ganz einfach ausgeben lassen.
Basically, there are different ways to construct such a driver model. These include common planning and BI solutions, as well as Microsoft Excel. Depending on the complexity of the model, characteristics or scope of data, these tools only fulfill the purpose to a limited extent. Flexibility (in case of change requests in the model), transparency (of the calculation logic and the interaction between inputs and results) and agility (when calculating different alternatives) often fall by the wayside. Implementation is time-consuming, and there is no time for the essential aspects of the “exercise”: defining goals and forecasts, evaluating and analyzing options for action, discussing scenarios.
Transparency and overview through modeling software
For a company in the telecommunications industry, we designed a business model-specific driver tree and built it to support simulations in planning in our Agile Enterprise Performance Management software, Valsight. In the following, we present the individual work steps.
Step 1: Define the key drivers and analyze the KPIs
In this business case, a typical call center was modeled in Valsight. The logic of the model is explained as follows:
- Verkaufskapazität: Es gibt eine bestimmte Anzahl an Calls, die bewältigt werden muss. Diese ergibt sich aus den Verträgen, die mit den Kunden ausgehandelt werden.
- Bestandskapazität: Der Kapazität steht eine bestimmte Anzahl an Calls gegenüber, die durch die Belegschaft bewältigt werden kann.
Inventory capacity is influenced or “driven” by the following two value drivers:
- Personalbestand: Die Anzahl der Callcenter Agents die im Callcenter angestellt sind.
- Calls pro Agent Die Anzahl an Calls, die ein einzelner Agent im Durchschnitt absolvieren kann.
Both drivers are subject to certain influencing factors. For example, the number of calls an agent can complete can be increased through training to communicate more effectively with customers.
Due to the improved capabilities, the agent gets the information he needs faster and can therefore devote himself to the next call more quickly.
The main driver in the present model is headcount. On the cost side, this item determines how big the office needs to be, how much electricity is consumed, the cost of office supplies, etc. Increasing the number of call center agents increases these cost items. On the revenue side, it drives the capacity that is sold to generate revenue.
The basic headcount is adjusted in the system under “Personnel planning”. This level is influenced by diseases. In addition, agents are also not available for calls during their training sessions. In this way, an actual headcount is then obtained, in this case “Available Agents”.
The sickness rate, in turn, is influenced by the “target/actual deviation”, i.e. the relationship between the number of minutes that agents are currently allowed per call and the number of minutes they will be allowed to spend in the future. If the pressure on employees is now intensified by reducing the number of minutes, the average number of calls that an agent manages increases, which increases capacity, and thus ultimately revenue. At the same time, this increased pressure on employees also means an increase in sick leave, as agents cannot withstand the pressure in the long run. This reduces the capacity again.
Such multilateral relationships are difficult to trace in a spreadsheet solution such as Excel. In a driver model, they are quickly explained and easy to understand.
Here, the “sold capacity” and the “inventory capacity” immediately result in a first KPI: the “capacity utilization”. Utilization is used to measure and indicate efficiency on the revenue side, which is always exactly 100% in the best case. This can be explored via the other drivers and their influences.
Step 2: Analyze the effects of changes in important influencing factors
For example, if the workload is 110%, this means that there are more calls than the agents can handle. To change this, there are several options:
1. adjust the default duration of the call. However, this can reduce capacity again at the same time due to the resulting increase in sick leave.
2. make personnel settings. This increases the existing capacity, but involves an investment.
3. training activities for the agents. However, the impact is comparatively small, costs are also incurred, and agents are not available in the short term. This would initially cause a further drop in available capacity before it increases.
The measures to be taken and the exact effects can be defined in the so-called scenario manager and analyzed in various diagrams.
Assumptions can be added at any time with a few clicks. Each assumption group (here, for example, “Personnel planning”) contains sub-items (such as “Personnel planning (additional)” here) that describe the change in individual elements in the model.
- Jede Unterposition hat einen Bezug zu einem Element im Modell. Bei diesem Line-Item ist es die GuV-Position „Personalplanung“.
- Die einzelnen Unterposition können für die verschiedenen Szenarien aktiviert beziehungsweise deaktiviert werden.
The underlying base data, as well as the data entered in the assumptions, are then active in the corresponding scenarios. If the values are changed via these assumptions, they automatically pass over the selected element, as shown in Image 3, into the driver model. In conjunction with the operators (calculation formulas), KPIs and other dependent elements are calculated via the model logic. Thus, the increase in personnel automatically changes the costs, but also the revenues. The ratio in which this happens, and whether the result will be positive, depends on the driver tree that reflects a company’s business model.
Subsequently, the simulated scenarios can be analyzed in various evaluation forms, such as line and bar charts or tables. In order to visualize the effects of individual assumptions on a specific key figure, a variety of diagrams can be selected. Here, we use an Assumption Bridge as an example, which separately shows the impact of adjustments in workforce planning on the contribution margin for 2018. In this way, planning measures under certain assumptions becomes very clear, simple and transparent.
But it’s not just medium- or long-term financial planning that can be optimized with value-driver-based software solutions like Valsight. Simulation solutions for planning new business models and calculating business or investment cases are also quickly set up. Business cases in particular, in which different possible development routes are to be compared, can be presented quickly and clearly with the help of scenarios and assumptions to enable direct comparison and thus facilitate decisions. The example shows how agile enterprise performance management can be implemented with Valsight.
