Predictive planning: the next step in the planning and budgeting revolution

Citation

Barrett, R. (2005), "Predictive planning: the next step in the planning and budgeting revolution", Measuring Business Excellence, Vol. 9 No. 1. https://doi.org/10.1108/mbe.2005.26709aab.002

Predictive planning: the next step in the planning and budgeting revolution

Predictive planning: the next step in the planning and budgeting revolution

Predictive planning

Many current budgeting and planning “solutions” do little more than simply automate the traditional spreadsheet-based budget. Whilst this may improve the efficiency of the process, saving time and cost, the opportunity is now available to generate value from the budgeting process. By integrating non-financial data into the budget, a process-based view of the business can be created that allows for driver-based budgeting, shared services modelling, intelligent capacity management, and fully integrated activity-based costing so that true cost and profitability data can be provided to users in a real-time closed loop solution. This allows organisations to undertake predictive planning as operational and financial managers are able to model resource requirements and capacity constraints inside the application as well as cost data, to build a uniquely dynamic budget capable of delivering a near real-time re-forecast. Changing business drivers in one department – such as resource consumption rates, conversion ratios, or forecast sales volumes – allows users to see the impact ripple across the organisation, right through to the profit and loss account.

Why is non-financial data so important?

Why is incorporating non-financial data so important? Because line managers do not run their business by looking at their financial costs every day; they typically manage by focusing on the key non-financial indicators that actually drive their costs and revenues. In fact when they come to preparing their departmental budget, they typically start with this key non-financial data, modelling the demands on their department and resource consumption rates to determine their actual resource requirements and costs. This approach is frequently called “driver based budgeting” (see Figure 1).

Figure 1

Driver-based budgeting

This approach allows managers to plan with reference to the numbers that matter to them. For a call centre manager that might mean looking at in-bound and out-bound call volumes, call duration and conversion rates. Production and logistics managers will be managing such things as the number of delivery loads, machine output rates and stock turnover targets.

The system generates their resource requirements (e.g. head count, floor space, computers, etc.), compares this against the available capacity, and generates budgeted costs by line item. Managers are alerted when their demands breach forecast capacity, or when they are under-utilising available resources.

What’s more, the resource demands in dependent cost centres can be linked so that a variance in one generates variance in another. For example, a sales manager working for a white goods company forecasts a 10 per cent uplift in sales for his territory, driving a demand in the customer service department for installation and service engineers to cover that territory. The customer service manager is aware of what is driving the demand and collaborates with the sales manager to meet this within his own planning constraints. Linking cost centres across departmental and divisional boundaries in this way ensures that all parts of the organisation remain aligned to each other, and allows financial planners to assess quickly how changes in one department will impact another.

Fully integrated activity-based costing

Another advantage of integrating non-financial data into the budgeting process is that it allows high level activities to be modelled within the same application, so that costs can be allocated through to cost objects such as products, customers and channels (see Figure 2). Adding an activity layer generates a further refinement to the budget model and allows managers to analyse budgeted performance across multiple output dimensions.

Figure 2

Using activities, cost can be allocated to discrete shared services such as PC support or payroll administration. Then, based on the resource requirements of each department (e.g. the number of PCs, head count, etc.), this cost can be calculated and the central services can be charged out appropriately to operational departments.

This brings several benefits to the organisation, aligning the budgets of shared service departments such as IT and HR with the demands that the organisation places on them, and scrutinising the cost and efficiency of the internal service delivery.

Activity-based budgeting

Once the key processes in the organisation are modelled using activities through to “cost objects”, users have the power to reverse their budget model to predict new levels of resource requirements and costs for any level of output. This approach to resource planning and forecasting (activity-based budgeting) allows users to forecast the impact on resources and costs simply by changing the volumes of product sales by customer and channel. The system will forecast new levels of activities, resources, costs and profitability for any level of output, highlighting capacity constraints, and integrating the internal demand for shared services (see Figure 3).

Figure 3

In the past activity based budgeting (ABB) has been simplistic, often involving little more than the “reverse calculation” of a standard activity based costing model. This is far from ideal. To undertake predictive planning, the process is better managed over three stages.

1. 1.

   There is a need to generate a new forecast of the volume and type of activities that are required, ideally from within the system, where users can play scenarios and adjust this forecast.

2. 2.

   The system should take these forecasts and calculate the resource requirements (including shared services). At this point users can directly adjust their resource requirements and make comparisons with capacities.

3. 3.

   Only then should the system complete the cycle by forecasting the final costs by line item and department.

All of this “driver” information has been available within organisations for many years, but it has never been included as a part of the budgeting process. By incorporating these non-financial drivers of cost, forecasting and planning become richer and more accurate, re-forecasts are available on demand, and operational managers can fully understand the impact that they have on the bottom line.

Because ABB reduces the involvement of line managers in the planning and budgeting process, it is rarely used as the sole planning and budgeting methodology. More often it is used in scenario planning to produce a top down forecast of the coming years, with driver based budgeting used to generate the annual budget.

Enhanced budgeting functionality

The focus of most budgeting and planning tools is on making the management of the core budgeting processes such as data collection, consolidation, and reporting, easier and quicker for the user organisation. Users are requiring these essential processes with an unprecedented level of speed and reliability, whilst requiring a whole new dimension that allows users to generate greater business alignment, increase strategic flexibility, and formulate quicker responses to change. These features are discussed here in turn.

The power of real time

One of the main problems with traditional budgeting and planning tools is the sheer time that it takes the organisation to complete the process. In particular, distributed systems suffer from this problem, as data has to be sent out to each user individually so that they can make their contribution before re-submitting for consolidation. Multiple iterations only make the process even longer. Many organisations find that their budgeting timetable limits the number of iterations that they can make, and therefore limits the degree of refinement and accuracy of their budget forecasts.

To do predictive planning you need to sidestep this entire problem by using a single centralised database coupled to a very powerful and efficient calculation engine, opened to a community of web-based contributors. The result is a system that enables real-time data collection, real-time consolidation, and dynamic forecasting of actual and ”what-if” budget scenarios. New data from the front line can automatically be validated and consolidated to update the group budget forecast whilst feeding into other departments to drive their budget revisions. Other applications require data manipulation and multiple steps to generate the same functionality.

Generating business alignment and collaboration

It has traditionally been difficult to ensure business alignment through the budgeting process. In many leading applications there is still a “horizontal disconnect” between individual operational plans as well as a “vertical disconnect” between operational plans and the over-arching objectives of the organisation. This is because the focus of these applications is still on silo-based planning processes, offering little collaborative planning support, and leaving managers to plan in isolation with little visibility of each other’s objectives.

Operational alignment. In most applications, some form of operational alignment is achieved by staging the budget submission process. For example, sales and marketing may start the process by forecasting sales volumes and pricing structures, operations then use numbers from this budget to develop their own budgets, and finally shared service departments complete the cycle. This lengthens budget cycles, reduces the opportunity for collaborative working as numbers are simply handed down the chain (often without justification), and restricts the number of iterations that the budget can go through.

The solution to this problem is to integrate a Work Management tool. This allows the finance department to control and administer the budget to a cascading and calendar-driven time plan, as well as allowing users to build links between cost centres that model the relationships in the business.

Given that the system is real-time, and therefore the links are dynamic, the system can make line managers aware the instant that changes in other cost centres impact their own resource requirements – not just during the planning cycle. The result is a significant improvement in the degree of collaboration between cost centres throughout the year, to the point that the process can support continuous operational alignment. For example, call centre managers can track the level of marketing activity in the pipeline and when and how it is likely to impact their department (see Figure 4). Shared services departments such as IT and HR can see forecast head count by job type and derive the internal demand for PC support and induction training.

Figure 4

Strategic alignment. Organisations commonly encounter the problem of reconciling “top-down” objectives with the grass roots intelligence and plans. Achieving an optimal balance with most systems is difficult for two reasons:

1. 1.

   At group level, Finance receives the budget at a consolidated line item level, further consolidating these figures to generate the forecast P&L, cash flow, and balance sheets for comparison with the strategic objectives. The key planning assumptions made by the business departments to generate their resource demands are lost during this consolidation process. Therefore when financial planners make “top-down” adjustments to the budget, they do so with no knowledge of how this will impact operational plans or the synchronisation between departments.

2. 2.

   At the operational level, departments budget using the operational data that is familiar to them – volumes, consumption rates, etc. – using this to feed a separate financial budget that then has to be consolidated with all the other departmental budgets for comparison against group objectives. It is not clear how their activities impact the group P&L, cash flow, or balance sheet, therefore they cannot and do not plan with these financial metrics in mind.

To truly undertake predictive planning, users can link operational plans to the top-level objectives of the organisation, consolidating financial and non-financial data in the system to give a true picture of the organisation’s resource requirements and how these relate to their strategic goals. In effect this builds a resource-driver hierarchy that explains how the operational plans add up to the business objectives. This has two important benefits:

1. 1.

   Financial planners can now run realistic simulations of the impact of top-level budget changes on every department and responsibility centre in the organisation before signing off the final budget.

2. 2.

   Line managers can plan knowing how their activities drive cost and revenues in other parts of the business, and ultimately impact the bottom line.

Achieving real time re-forecasting

In today’s turbulent market environments, organisations need to re-forecast more frequently. To do this in any detail usually means starting the budget process again, collecting data, consolidating it and then making adjustments, and finally entering into internal negotiations. As described earlier, in most systems this process can only move as quickly as the slowest member in the dependency chain, leading to an impatient wait as business decision makers anticipate the results.

Using current systems, most organisations can only hope to complete a top level or “soft” forecast in the time required. This usually involves asking departments to project a new set of high-level numbers based on a given scenario. However, using the relationships that are built into a predictive planning model, organisations can re-forecast at the touch of a button, right down to the level of individual cost centre budgets. An example of the inter-linking relationships used to generate a real time re-forecast is illustrated in Figure 5.

Figure 5

Changing assumptions in one area of the business (e.g. business objectives, net pricing, or consumption rates) can automatically trigger changes in the rest of the organisation and into the future, quickly revealing the impact on every department and accurately predicting the impact on business performance.

This has two important applications:

1. 1.

   Faced with sudden changes in their markets, financial planners can update the budget model and re-forecast a budget scenario with a single click.

2. 2.

   Updates posted by line managers can be immediately consolidated into the group re-forecast, revealing their impact both on other departments and on the bottom line.

Because the data is all held centrally, and the results are calculated as needed in real-time, organisations can skip the lengthy “submission” process; each user’s data can immediately be included in the overall picture in as many “scenario” options as required. Once the figures are agreed the new budget version can go live and the organisation can move forward.

System deployment and maintenance

Deployment

Minimising the effort required for deployment is especially important for enterprise-wide applications, where the user base may be large and widely spread. These can create significant costs, possibilities of errors and time delays. Users should take the following actions to minimize the deployment effort required:

1. 1.

   By using a web-enabled system there is no need to manually load software on budget holders’ computers. Users just access the model using their standard web browser.

2. 2.

   In this way, the same report and data collection templates can be deployed to a large community of contributors, with the exact same template behaving differently for different users according to their role in the organisation and their rights to access particular pieces of information. This avoids the requirement of many other applications to build views onto the database for each different access level, which involves much repetitive work and time-consuming maintenance.

3. 3.

   During the initial data load, the system needs to read the data relationships and automatically build the account structures and the hierarchies of the business. It then can analyse the historic relationships between data interactions and builds the main dimensions of the data model.

The system will also need multi-language and multi-currency support, allowing users across the world to interact with one global model, by presenting data in the local currency and language. At a group level, budget data can be automatically converted and consolidated into the home currency.

Maintenance

To predict, the planning software has to have the flexibility to absorb changes quickly and easily. This requirement is not just in the budget data but also in the organisational structure, consolidation paths, resource hierarchies, and business driver relationships. For example:

Budget data. The data always needs to be held in the central database. Users mustn’t be allowed take their data off-line, so that it is held within the system and their assumptions and rules are available for all to see and inspect. There should be no separate submission or consolidation process, so that there is no risk of data becoming inconsistent or outdated.

Business structure. Building in the data logic means that data collection grids or “books” can be created to continually reflect the current organisational structure. Organisational changes then don’t create a need for the wholesale re-development of the books. Similarly, when importing data from another source (e.g. General Ledger) the system must recognise and automatically update its model with new line items and dimensions instead of generating exception reports.

Consolidation. The organisation’s hierarchy and consolidation paths need to be represented graphically, and changed easily, ideally by simply dragging and dropping items into the correct structure. Multiple alternate structures can then be created to allow for scenario reporting.

Driver relationships. In many tools, the cost driver relationships between cost centres and within departments are either hard-wired into the model or not modelled at all. Changing these relationships can be a time-consuming and expensive process, usually requiring the specialized skills of application programmers. These links within and between cost centres need to be dynamic, requiring no hard-coding, so that the budget model can be continuously kept up to date with the latest driver information and relationships.

Conclusions

With change being one of the certain constants of modern business life, the planning has to be designed to enable an organisation to adapt quickly and efficiently to changing circumstances. By incorporating the non-financial drivers of cost, operational managers are free to budget and plan with numbers that are relevant to them. Furthermore, by modelling the essential cause-and-effect cost relationships across a business, powerful features such as ”single-click” re-forecasting, cross-functional capacity management, and shared service optimisation within the operational budgeting environment become a reality.

Predictive planning is a Holy Grail for many accountants, especially in the current complex and often chaotic business environment. We cannot predict in the sense of “foreseeing the future”, but with non-distributed, highly scalable, multi-lingual, multi-currency, and web-enabled software, modelling the consequences, taking corrective action and predicting the resulting outcome is now a reality.

Whilst many solutions will claim to support enhanced budgeting practices, when put to the test most are designed simply to automate the outdated spreadsheet-based system of budget management. They have been designed from a purely financial perspective to perform the financial budgeting basics, usually in a calendar driven fashion, with little operational perspective (apart from simple volume × cost rules), and with no understanding of capacity management and shared service alignment. These tools may generate an incremental cost improvement in the budgeting process, but to maintain a competitive edge organisations need to do more.

To support a step change in business performance, enhanced budgeting process improvements are not enough. New functionality is required that enables the integration of non-financial data into the system to allow modelling and prediction of the cost drivers. Quality information has to be provided in a timely fashion from both finance and operating functions so that the whole business can be reflected in the model. In that way, managers can react to change in an integrated way and predict with a degree of confidence that they don’t have at present.

Richard BarrettMBA, FCIM is Vice President of Global Marketing at ALG Software, Knutsford, UK. He started his career within the pharmaceutical industry and gained an MBA in 1981. He has a wealth of experience in consultancy and holding national and international positions in consumer marketing, insurance and as well as business-to-business marketing with DHL Worldwide Express. E-mail: rbarrett@algsoftware.com