IT Financials Glossary

ABC (Activity-Based Costing)

Posted in Costs by mgentle on August 26, 2010

ABC (Activity-Based Costing): a method for assigning costs to products and services based on the activities required to produce them. Knowing how much products and services really cost allow companies to justify expenditure, identify what’s profitable vs unprofitable and explore opportunities for cheaper alternatives.

ABC is best understood via an example. Let’s imagine a customer service department in a Business-to-Business (B-to-B) telco with a team of multi-skilled call centre agents who handle both external customer enquiries and internal enquiries from the sales force. In addition, they correct orders that have wrong or missing information by calling the customers and doing it on the phone.

You build an ABC model by first interviewing the agents and asking them how much time they spend on these three activities, which they estimate at 60%, 30% and 10% respectively (for simplicity’s sake we will avoid breaking down these activities into “cost pools”). You then count the number of “cost drivers” – ie, the number of external enquiries, internal enquiries and rejected orders – which for our example is say, 10 000, 2 500 and 1 000 per year respectively. Finally you factor in the total costs of the department, which are $1m per year.

This data is then entered into a model in an ABC system, which calculates the following activity cost driver rates (Fig. 6.1 below – NB click on the image to get a full-screen view):

An example of Activity-Based Costing

Needless to say, this simplistic example hides a number of inaccuracies and complexities, the main ones of which are:

  • The time spent is based on people’s subjective estimates of their behaviour – and assumes 100% annual productivity, with no idle time.
  • Any changes in activities and processes due to new or changed products, or to exception processing, would require re-interviewing and maintenance of the model.
  • If the approach were applied to 100 people instead of only three, and across multiple activities more complex than those for a call centre, the effort and cost involved in interviewing people, setting up the model and maintaining it could soon become prohibitive.
  • If required, as in IT, it would be difficult to break down people’s activities into capex and opex.

The conclusion is that ABC is more suited to relatively stable, industrial, commodity-type processes with little exception processing. It would also need to be a high-volume activity to justify the overhead of maintaining an ABC system.

Even in a stable industrial environment, the above constraints can still become a barrier to ABC, which has led the inventors of the original model, Robert S. Kaplan and Steven R. Anderson, to propose in 2003 a simplified version called time-driven ABC (see “Further reading” in Appendix 2).

Needless to say, IT projects do not correspond to this type of environment. The most common method for evaluating people costs in a project is through time-entry. ABC in IT is therefore only really applicable to stable production applications and services.



Posted in Costs by mgentle on August 26, 2010

Capex: see Costs.

Capital costs

Posted in Costs by mgentle on August 26, 2010

Capital costs: see Costs.

Cash out

Posted in Costs by mgentle on August 26, 2010

Cash out: the sum of capex and opex.


Posted in Costs by mgentle on August 26, 2010

Costs: Costs can be categorized in different ways.

The most fundamental cost categories are fixed costs and variable costs:

  • Fixed costs: these are costs not directly influenced by business activity or usage – eg, rent or annual software maintenance. Fixed costs are generally easy to plan and to manage, but they are risky in that they represent long-term commitments and therefore need to be entered into with caution.
  • Variable costs: these are costs that vary with business activity, or usage, such as telecoms charges or disk storage. Because variable costs are usually driven by decisions or events outside of IT control – eg, changes in BU usage policies, or a peak in customer service activity following a product launch – they are more difficult to plan and manage.

Fixed and variable costs are useful from an overall perspective in terms of pricing and negotiations, but they don’t help in understanding who or what is driving the totals. It is therefore useful to be able to assign these costs directly or indirectly to a customer or an activity, so as to help justify IT expenditure, facilitate pricing or explore opportunities for cheaper alternatives:

  • Direct costs: these are the portions of the fixed or variable costs that can be directly attributed to a cost centre, activity or customer. Examples include dedicated hardware, software or application support costs.
  • Indirect costs: these are the portions of the fixed or variable costs that cannot be directly attributed to a cost centre or an activity – eg, shared infrastructure and network services. Such costs have to be apportioned or allocated based on criteria like number of users or BU revenue.

All costs, fixed or variable, direct or indirect, are ultimately classified as either capital costs or operational costs:

  • Capital costs: more commonly known as capex (short for capital expenditure), this represents the substantial assets of the company, like plant, property, equipment – and IT systems. Capital costs figure in the balance sheet the year in which the asset is acquired, and are depreciated as expenses in the P&L during its useful life. Capex in essence pushes out expenditure incurred today to subsequent years, thus making the current year look “better” (which is why most CIOs like it…). It is therefore closely monitored by the CFO, lenders and financial markets – eg, via the capex-to-sales ratio – though clearly not closely enough to prevent a major telco of creative accounting fame from incorrectly capitalizing $3.8b of expenses to make its 2001 and 2002 numbers look better (what has the WorldCom to…?).
  • Operational costs: more commonly known as opex (short for operational expenditure), this represents day-to-day running expenses whose effects can be measured within a short timeframe. Unlike an asset, an operational expense has no intrinsic value. It’s just that – a one-time expense. When companies go through a round of cost-cutting as explained at the start of Chapter 1, it’s usually the opex that’s being cut. This is what the CIO monitors closely each month in the IT financial reporting.

There are clear rules for what can be capitalized and what must remain an operating expense. This is explained in detail in Chapter 2 (Budgeting rules for capex vs opex).

The sum of capex and opex is called cash out, which is the “real money” that the company has to pay for goods and services, regardless of how it will be accounted for later.

Direct costs

Posted in Costs by mgentle on August 26, 2010

Direct costs: see Costs.


Posted in Costs by mgentle on August 26, 2010

Expenses: Operating expenses or opex. Sometimes used as a verb, as in “expensed”.

Fixed costs

Posted in Costs by mgentle on August 26, 2010

Fixed costs: see Costs.


Posted in Costs by mgentle on August 26, 2010

Forecast: actuals plus estimated remaining costs. The forecast is in essence a revised budget, based on better visibility on what’s really happening and is likely to happen next. The original budget will often have been defined at least six months beforehand, based on assumptions and estimates that almost always have to be adjusted as reality sets in.

A forecast should cover, at the very minimum, the rest of the financial year and, ideally, a rolling 12-month period (called a rolling forecast).

Indirect costs

Posted in Costs by mgentle on August 26, 2010

Indirect costs: see Costs.