Solving the ProServ Forecast Challenge

iStock_000010840637Medium_Binoculars_Crop1d

Every industry has a certain economic model and its own set of operational dynamics that distinguishes it from other industries.  Some characteristics may be similar across industries, or may lend themselves to useful analogies between industries that can often lead to creative viewpoints.  The Professional Services, or “ProServ” industry is no different.  Defined broadly, it encompasses any business where the value proposition is largely centered on people with specialized skills providing services to clients and where the time value of these individuals constitutes the major cost of such services.  Labor in ProServ businesses usually comprises anywhere from 60% to 75% of the entity’s cost structure.  It includes not just consulting, accounting, legal, engineering, architecture and similar firms, but also industries such as advertising which, while not traditionally defined as Professional Services, largely and increasingly follow a ProServ economic model.  Forecasting is often a difficult exercise in many ProServ scenarios.  This article seeks to highlight some of the obstacles and other considerations that must be overcome to successfully meet the forecasting challenge.

The ProServ Backdrop

It’s important to understand that even within the ProServ industry there is a wide range of variability between business models when it comes to components of the economic model:

  • Revenue models can range from straight time and materials (i.e., hours billed plus direct costs) to arrangements based on fixed fees, retainers, or milestone deliverables.
  • Client contract structures can range from single discrete Statements of Work (“SOW’s”) to Master Agreements (often with individual SOWs underneath) to Retainer Agreements that may act like a bank account of hours to be drawn upon.
  • Labor operating models can vary from discrete pools of labor with a specialized skill set that correlate directly to a client SOW, to profit centers containing several pools of talent that service client contracts specific to that profit center, and all the way to a more expansive and fluid organization that shares labor across profit centers and geographies.
  • The approach to Work Breakdown Structures (or “WBS Elements”) varies across ProServ sub-industries, but can also vary within entities based upon the nature of the client contract or scope of work or the monitoring style of the client team.  Law firms have “legal matters,” accounting firms have “engagements,” and advertising agencies often view their work in the context of “campaigns.”  But there can be other levels in the service definition also useful as a measure of work effort that are therefore incorporated into the project structure.

It’s important to understand how each of the above variations might fit into and impact the forecast approach.  It’s also important to understand where Forecasting fits into a broader ProServ platform landscape.  The four core components of any ProServ platform are Budgeting & Forecasting, Project & Resource Planning, Entity-Level Historical Reporting, and Client/Project-Level Historical Reporting.  The former two are focused on forward-looking views while the latter two correspond to actual results.  A second correlation is between Project & Resource Planning and Client/Project-Level Historical Reporting, both of which deal with more granular levels of detail, while Budgeting & Forecasting and Entity-Level Historical Reporting deal with summary information.  These relationships are presented visually in the following diagram:

Platform Diagram with Gears_Bigger Font

These four key components have to be aligned to mesh properly.  For example, the chart of accounts and entity codes have to be standardized between the two top components and information passed back and forth. Project structures and labor structures have to be standardized between the bottom components which together should provide a complete project management view.  And there should be data feeding from the detailed layers into their respective summary layers to facilitate those upper-tier views.

So when we discuss the challenges of Forecasting in ProServ scenarios it’s necessary to address considerations in both the summary (Budgeting & Forecasting) and detailed (Project & Resource Planning) layers as these tend to bleed into one another depending on the specific business involved and process strategy.  However, within the discussion below I’ve highlighted areas where the distinction may be particularly important.

Key Considerations in ProServ Forecasting

ProServ shares many forecasting challenges with other industries, but some carry special considerations.  These cut across foundational, system and process elements of the overall forecasting approach.

Hierarchical Levels and Planning Entity.  One of the most fundamental decisions is the designation of the “planning entity” and the relevant hierarchical levels to be incorporated into the forecasting process.  Put another way, how granular do you want to get with the forecasting process and where are boundary lines necessary to reflect management accountabilities?  Upper-tier levels such as consolidation entity, business unit, profit center and cost center can generally be approached as a Budgeting & Forecasting exercise.  But if the process reaches down into more detailed forecasting focused on clients, projects, account executives, lines of business, or specific labor pools (i.e., departments, labor grades, or individual employees), then a more robust Project & Resource Planning tool and process is warranted.  This is likely the case for a ProServ business with any degree of scale and complexity, in which case the upper-tier levels are simply an aggregation of these detailed forecasts (perhaps with some degree of informed refinement across business units).

User Constituency.  Who are you aiming to support with the forecasting platform and protocols?  Are you purely interested in a high-level view to help senior management manage the overall business?  The problem with this (besides potential data integrity questions) is that it will generate limited actionable information at the lower operational levels.  Or are you seeking to implement an operational tool – woven into the organization’s operational processes – to drive front-line decisions (for client-handling executives, project managers and labor managers)? In this case the high-level view becomes a by-product of the operational tool.

Accountabilities and Ownership.  Closely related to the question of hierarchical levels is that of accountability for any specific component of the forecast.  Who in the organization “owns” the client?  The project?  The labor resource?  It derives from the basic operating model of the business and from there extends to the forecasting framework, and becomes more challenging in scenarios involving global and regional clients that cross organizational boundaries.

Sales Cycle.  ProServ entities have varying sales cycles depending on the profile of their clients, projects and sub-industry segments.  This can range from large engagements extending over many months or even years to small quick-turnaround projects.  The forecasting design should consider three primary categories of revenue projections:

  • Contracted business – This essentially corresponds to the estimate-to-complete on any in-process work and should correlate to an executed legal contract.
  • Proposed business – This corresponds to potential revenues from work for which the company has a bid outstanding or in process. The work has largely been defined with respect to both scope and pricing (although there may still be multiple versions in play).  A forecasting approach should assign a probability of conversion to each individual bid.
  • Budgeted business – Any forecast with a medium or longer term time horizon must also incorporate some portion of work where the source is currently unknown.  This may correlate to additional work from existing clients as well as work from entirely new clients.  It’s important to assign clear accountabilities for each underlying target and in many situations a “haircut” adjustment to reflect some appropriate degree of conservatism.

An important forecast consideration is the degree to which there is clear visibility into these cycle components underneath the forecast, and the extent to which there is a desire to track the progression of a particular forecast element through the stages.

Client Contract Association.  Consideration should be given to correlating the forecast elements to the associated client contract covering each major portion of work.  The contract may represent a Master Agreement or smaller Scope of Work documents.  This may prove useful in assessing actual performance against each contract and providing insights for future negotiations.  It could also provide a higher degree of visibility into scope creep, potentially providing the justification for a contract renegotiation for factors out of the servicing entity’s control.

Labor Planning & Management.  Of all the benefits of a robust and effective forecasting process, none is more important than labor planning and management.  This is where immediate, actionable operational decisions come into play, especially in environments where there is a lot of volatility in the business.  It’s also the area where one finds the greatest contrast between detailed Project & Resource Planning tools and upper-tier Budgeting & Forecasting platforms.  It leads to many questions in developing a strategy and process.  Do you want a robust, integrated view of labor supply and demand for each class of employee or labor pool? What is the tactical approach around “ringfencing” labor (department definitions, grade definitions, pool definitions)? Is there integration with the talent management system to leverage master data?  An approach of “tapering granularity” can (and likely should) be applied whereby the level of specificity is matched to the firmness and imminence of the sales cycle phase (for example, Budgeted Business would forecast labor at the overall labor pool level while Contracted Business would forecast at the labor grade or employee level).  Labor availability has to be incorporated to provide real-time visibility to department heads with respect to labor utilization and a basis for action and capacity leveling.  In many situations freelance contractors are used to deal with ebbs and flows and specialized expertise, and access to this type of information makes for better decisions. All of this ties back into ProServ forecasting.

Rate Cards & Standard Costing.  The approach toward costing hours and pricing work is a fundamental decision in any ProServ business and it leads to cultural questions.  What degree of willingness is there to share information across the broader employee population? Are standard costs – a powerful finance tool – adequate for the business situation?  This also impacts the definition and structure of labor grades.  Widely-banded labor grades provide more “masking” of employee compensation information but greater potential for variance from actual, while narrow banding more closely approximates actual but may effectively broadcast salaries.  On the revenue side is the question of price discounting, where a gross presentation of standard rates and discounts provides greater visibility into “realization” rates and trends.  The forecast should mirror any transfer pricing methodology in use if accurate profit center attribution is critical to the forecast.

“Cross-border” Resource Sharing and Transfer Pricing.  Extending from the question of accountability and ownership is the approach for resource sharing across organizational boundaries and the treatment of such situations in the performance measurement systems.  Accountabilities may be structured around discrete profit centers, but capacity and client servicing demands may necessitate cross-border resource or project sharing.  The operating model has to incorporate fluid mechanisms around assignment of resources and transfer pricing that foster collaboration and cooperation, and by extension any forecasting platform has to reflect these at both a detailed level (for accurate resource planning) and at the upper tier (to ensure the integrity of business unit forecasting across the organization).

Estimate-to-Complete.  ETCs are a critical aspect of managing and controlling the delivery of services to clients.  They’re no different in principle than general forecasting, but certain variables are much more defined (mainly contractual pricing and scope) and it requires more immediate and definitive action to correct problems (such as adjusting the staffing profile or controlling scope creep).  ETCs truly sit at the juncture of forecasting as a front-line operating protocol versus forecasting as an overlaid control process.

Project and Work Breakdown Structure.  This represents an intersection of the question of hierarchical levels and accountabilities.  One must determine the appropriate and most relevant level of granularity to incorporate into Project & Resource Planning tools.  Even if the revenue aspect of a major piece of client work is captured at an aggregate level, it is often beneficial to break the cost collection into smaller logical groupings for purposes of control and analysis (for example, a separate project element for each discrete activity).  In an integrated operational platform, these could factor into the overall forecasting design.

Staffing Profile Templates.  Continuous improvement entails building upon existing knowledge.  By capturing information, it can be leveraged for future use and serve as a guide for improved future forecasting and planning through the use of profile templates that suggest a staffing mix for each major type of work. This is especially relevant in the scoping and pricing of work.  Profiles such as these can be used on a detailed basis within a Resource Planning tool, but also have applicability for general estimating at an upper-tier of forecasting.

Operational Workflows.  Budgeting & Forecasting and Project & Resource Planning are meant to capture the financial implications of operational activities.  So the workflow around preparing and utilizing these should align with the operational flow to the greatest degree possible.   These are TOOLS and so they should provide value, not expend energy beyond the benefit gained. This will help drive proper adoption and improve the integrity of the data collected.

Rolling Forecast.  There seems to be ongoing discussion and debate over the value of traditional budgets versus a rolling forecast, with various permutations and definitions of what each represents.  But one thing is certain.  In ProServ models, some form of a rolling forecast incorporated into the approach is often critical if performance is to be effectively managed.  This is driven by the importance of estimates-to-complete (“ETCs”) in managing projects and scope creep, by the sometimes volatile nature of labor demand, and in some cases by the use of contractors and freelancers to level the peaks and valleys of labor utilization.  This rolling forecast can solely reside within the more detailed Project & Resource Planning layer, or can be integrated into a true aggregated rolling forecast (updated monthly with actuals) to be a more effective tool for senior management with a wider view of the business trends.  A detailed planning horizon would likely be shorter term (three to six months) while a consolidated view should likely extend 12 to 18 months (with tapering levels of detail).

Master Data Management.  MDM is a foundational consideration in any system design.  Similar to other industries, a ProServ forecasting design would inherit these Master Data frameworks with particularly critical linkages in the areas of Client Master Data and Employee Master Data.  The latter would likely involve integration with the company’s Human Resource Information System.  Challenges arise in the propensity for employee status changes throughout the year via promotions and labor grade changes.  These can alter existing resource and project plans (and the related forecasts) and so they must be considered in the forecasting approach and design.

Gauging the Forecasters.  ProServ forecasts usually involve inputs from many parties with the greatest degree of subjectivity and “art” coming from the revenue side, where client-handling executives and rainmakers must assess the probability of landing new business.  A feedback loop that monitors the accuracy of individual forecasters over time can be extremely useful, as this provides guidance for senior leadership to adjust future forecast inputs up or down to reflect the propensity for overly conservative or overly optimistic estimates from key individuals.  It is also useful to incorporate other analytics into the forecast such as line of business roll-ups, since this also affords the opportunity to apply targeted adjustments to account for macro trends in the business.

*   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *   *

These are all aspects of managing a ProServ business.  Someone in the organization is either proactively addressing every one of these points or reactively managing crises.  They all have implications for forecasting, but many are much more fundamental and represent basic operational design and performance measurement decisions.  Most are questions that can only be answered by:

  1. Understanding the organizational business strategy.
  2. Understanding the organizational process maturity.
  3. Developing a platform and process strategy that melds with (1) and (2) and a transformation roadmap to get from current state to future state.

In this way, a forecasting process can be structured that balances the optimal with the practical . . . one that is achievable in light of the specific business profile.