OF Icon

Outsourcing Fundamentals

Independent guide to Outsourcing: BPO strategy, Service Provider selection, contracting and implementation projects. Assessing and improving your current service.

Contracting > Commercials

The main commercial elements of the Agreement

In this commercials page, there are a set of interrelated topics which are critical to getting value from an outsourcing service:

  • Pricing: the way in which the services are charged with the variations in work volumes, resources required.
  • Service Levels and Service Level Frameworks: how performance of the services is measured and rewarded.


The Pricing Challenge

If the demands and therefore the output of a service would remain constant right across the term of a BPO agreement, then pricing would be simple. In practice, the service volumes change, the respective size of difference components of the service vary by week, by month and year on year. In this reality:

Client goals will tend to be:

  • Value for money
  • Predictability of charges
  • Flexibility to increase and decrease demand

Provider goals are:

  • To get a fair return for resources deployed
  • To get a fair return for investments (cost of sale, infrastructure set up etc.)
  • Predictability of income stream (and ability to plan resources)

Main Types of Pricing

  • Fixed: The charge for the entire scope of service is fixed
  • Cost plus: The provider accounts for costs of providing the service and is paid a margin on top of costs.
  • Anchored: This is a variable scheme anchored around a baseline volume. The charge for a baseline volume of services is set. Volumes above or below the defined level are costed at a marginal rate and the charge adjusted.
  • Variable: The charge is proportional to the volume of service.

In practice, most pricing frameworks are a variant of "anchored" and one structure for this type is given below.

The Anchored Pricing Framework (ARCs & RRCs)

In this pricing framework, a baseline volume is estimated by the client. This is the projected volume for the service over a specific period (can be defined by month or year). The pricing agreement sets the charge to be made for the baseline volume. It also sets how much this charge would be increased for volumes in excess of the baseline (an Additional Resource Charge or ARC) and how much the charge would be reduced when the volume is less than the baseline (a Reduced Resource Credit or RRC). The ARC and RRC rate is typically less than the baseline volume rate so that increased volumes get the benefit of scale economies and the total charges for lower volumes still cover fixed costs.

Pricing Units

The pricing unit identifies what you are paying for. It can be any of the measurements made on the process being outsourced, for example, FTEs utilised, transactions performed or outcome such as collections made. In selecting a pricing unit, you need to test how it fits with your and the provider's objectives for pricing. The main types of pricing unit are:

  • Capacity: for example FTEs (also systems, desks)
    • can be used when other volumetric data is not available
    • the resource capacity is committed
    • Does not motivate providers to increase efficiency
    • Need to agree level of resource required to perform service
  • Work: for example invoices processed, training days delivered.
  • (note some aspect of HR outsourcing use number of employees as a pricing unit. This is taken as a proxy for work, because from year to year, the amount of work over the whole workforce is proportional to number of employees)
    • chosen well, cost of providing services is a function of the volumetric
    • The initial rate for the pricing unit needs to be defined in most cases using client's history of volumes and resources for the internal activity. If this history is not well documented, the provider may reserve the right to "true-up" the rate after service has been implemented.
  • Outcome: for example candidates recruited
    • Client pays for results
    • Providers have to take a risk and will be unwilling to do so unless the potential reward is high.
    • Client may, especially after some time, feel a successful provider is being over paid.
Limits and Bands

Both Anchored and Variable pricing schemes will only satisfy provider and client objectives over a range of volumes. An increase in volume above a certain point may require the provider to invest in a new service centre; a decrease below a certain point may make provision of the service unviable. Pricing schemes will define the top and bottom limit beyond which there will need to be a re-pricing negotiation.

Providers may ask for a minimum volume or revenue commitment in order to cover investments and overheads. If given, it should fit with the pricing scheme.

There may also be a dead band defined. This is a band around the baseline volume within which ARCs and RRCs are not applied. A dead band avoids trivial charge variation in the region of the projected volume.

Other factors

Exchange rates and Inflation: for offshore outsourcing, a large portion of the provider's costs will be in a different currency to that of the charges. The provider will either mitigate the risk of exchange rate and inflation variation by indexing their charges, or build the risk into the rates. The position on inflation and exchange rates should be established as part of the RFP process. You would normally request pricing in your preferred currency and it is worth checking with finance whether there is a preferred approach for dealing with exchange rates.

For multi-country outsourcing services, taxation can be a very complex topic and, although usually does not affect the overall shape of the service very often, it is best to consult your tax specialist early on.

Service Levels and Service Level frameworks

The service levels and Service Level Framework:

  • Set the performance levels for the services
  • Define the consequences for under performance
  • Provide mechanisms for changing the priority of different performance indicators and changing service levels

There is a large degree of standardisation of methodology and of language across the outsourcing industry and the structure described below should be familiar to providers. However, there remains enough variation that you should test for common meaning for everything being discussed with your potential providers.

The sole objective of the service level framework is to motivate the provider to deliver services to defined performance levels and, in the case of failure, to restore good performance as soon as possible. Do not distort this objective by using penalties to reduce the cost of the services.

The Basic framework

The basic premise of this framework is that the performance of the service is measured by a set of performance indicators. Under performance, by definition, is not meeting the levels set for the performance indicators. The performance indicator measurements are reported and under performance triggers corrective action. Missing the level for the highest priority performance indicator triggers financial penalties. Repeated under performance would ultimately lead to the client being able to terminate the contract for cause.

Performance Indicators

Performance Indicators are the metrics of the service performance elements. They would include:

  • Timeliness: typically, the percentage of transactions completed within a specified time; percentage of calls answered within 30 seconds.
  • Quality/accuracy: typically, the percentage of transactions completed error free.
  • Satisfaction: typically, the percentage of users responding to the agreed survey with satisfied or better.

Some of the challenges of defining performance indicators are:

  • There is a compromise to be made between having too many measures and not having coverage of the whole of the services. It is better to focus on the most important processes.
  • Transactions must be well defined, or more specifically, the hand-offs between client and provider be defined precisely in order for the performance indicators to work.
  • Providers will want to put limits on their commitment to perform. For example, to agree the response time for a transaction, the provider will want to have a limit on the rate of transactions to be processed.
  • Satisfaction is a subjective measure. Many providers will not agree to a satisfaction performance indicator being a KPI or CPI. Even where a provider will agree to a satisfaction measure, they will want a period for services to settle before the indicator is active.

Performance Indicators each have two levels of performance defined:

  • Target: represents the level expected of the service. Dropping below this level is under performance and requires corrective action.
  • Minimum: represents a level below which the service is unacceptable. Dropping below this level demands immediate response.

The performance indicators, all measured and reported, and each with its two levels, are classified by importance:

  • General Performance Indicators (GPIs): There is no consequence for their being missed. They are for information only.
  • Key Performance Indicators (KPIs): defaults are investigated and corrective actions taken. KPIs can be promoted to become CPIs.
  • Critical Performance Indicators (CPIs): In addition to corrective actions, defaults of CPIs trigger penalties (Service Credits)

The use of this classification takes account of some performance measures being more important to the business than others, and allows focus on the most important for penalties.


A Service Credit is the penalty on a default of a CPI. The amount of the service credit is calculated from its allocation, the Pool Percentage and the At Risk Amount. Service Credits are normally deducted from the charges in a following period.

At Risk Amount: this is the value, usually set as a percentage of monthly charges, that the provider agrees to have at risk in the service level framework. The level should be set as part of the pricing proposal. The sum of all Service Credits in a measurement period cannot exceed the At Risk Amount.

Pool Percentage: In order to allocate a meaningful value to each CPI, more than 100% is apportioned. With a Pool Percentage of 150%, 15 CPIs can be allocated 10% of the At Risk Amount. A default of any of the CPIs would result in a Service Credit of 10% of the At Risk Amount. However, the At Risk Amount remains the limit and in the unfortunate situation that 11 or more CPIs have defaults, then the Service Credits would still be limited to 100%.

Defaults: A default triggers consequences, not least the award of a Service Credit. A typical arrangement would be that missing the minimum performance level of a CPI is a default and missing three consecutive target levels of a CPI is a default.

Earn Back: In some agreements, the provider is allowed to "earn back" Service Credits through good performance. For example, where there has been a default, the provider may earn back the Service Credit by achieving the target performance for, say, the following 4 months. Clients need to asses whether having higher motivation to restore good performance is worth losing some of the incentive not to drop performance in the first place.

In defining the components of the service level framework, the business consequences of service performance must be thought through. Is the consequence of a single failure of a CPI critical, or is the business requirement for a broad reasonable performance?

Flexing the performance indicators

One of the objectives of the Service Level Framework is to be able to adapt to the changing needs of the business.

The client can choose to promote KPIs to be CPIs and CPIs demoted in order to change the focus of the provider. Providers will want to have notice before a change becomes effective and limit the number of changes than can be made.

GPIs can be promoted to be KPIs, but only by agreement with the provider (change control), since this increases the provider's commitments.

Changing the target or minimum level: Because this may change the resource required by the provider to perform, changes in levels require agreement.

Introducing a new PI: Where a new measure is needed, it can be introduced, by agreement, as a GPI, allowing the measure to be set up and tested without having implications other than the cost of its implementation.

Challenges and issues

Operating Level Agreements (OLAs) are the equivalent of performance indicators except that they are obligations on the client to perform. There are areas where providers do depend on client performance, for example where a client system is used to support transactions.

Segmenting the at risk pool: It doesn't take too many transactions covering too many geographic regions before the Service Credits become too small to have any impact. It is best to focus your allocation on the few most critical CPIs.

Using up the penalties early in a service measurement period: Do check that your Service Level Framework doesn't leave significant periods where a provider may have lost Service Credits and gets no benefit from restoring good service until the start of the next measurement period (typically month).

Double jeopardy: Most agreements are explicit that a single cause of failure is only one failure. This may be sensible for most incidents, but do remember that if the outsourced processes have a point of failure impacting a large portion of the service, then still only one default would arise in the event of failure.