Performance indicators in most organisations are not aligned to the goal of the organisation. In addition, they are rarely designed to measure a single process and be unaffected by factors outside the single process.
The usual result is that members of the organisation only concern themselves with high level process indicators. For example, the value of sales ($’s) or the quantity of sales (widgets or services), operational, sales, marketing and financing costs ($’s).
However, decisions made in organisations are usually at much lower level processes. For example, to work an extra shift, to invest in more productive machinery, or to train sales people in product knowledge.
The result is that organisations make decisions without a quantitative link between the decisions and their outcomes, which when aggregated, generate the high level indicators.
This can be avoided fairly simply, albeit with a bit of trial and error. Cascading the organisation goal down through a cause and effect tree is a simple way to arrive at performance indicators for a process. The key step people tend to forget is to cascade the processes before cascading performance indicators.
If we imagine that we want to develop KPIs for a sales team as part of their effort in achieving a profit after tax of $2M, our analysis of the processes which give rise to the profit might look something like this:
[net profit after tax] is composed of:
[sales], [costs], and [tax]
[sales] is composed of:
[sales to existing customers] and [sales to new customers]
[sales to existing customers] is composed of:
[average current sales per customer in a defined period], [number of current customers in a defined period] and [customer churn in a defined period]
[sales to new customers] is composed of:
[prospects], [average sales potential per prospect] and [sales conversion]
[prospects] is composed of:
[marketing reach] and [marketing efficiency]
[average sales potential per prospect] is composed of:
{product relevance to customer}, {product range} and {product price}
[sales conversion] is composed of:
[customer visits planning effectiveness], [sales process effectiveness] and [number of sales people]
Without completing the tree, the idea should now be clear. The tree may also be built to further levels of detail. At some point, however, the detail will not be measurable efficiently and effectively.
For each of these processes and sub-processes, performance indicators may be proposed. For example:
[net profit after tax] $’s
[sales]: Sales ($’s)
[sales to existing customers]: Sales ($’s) to customers who have been buying for more than three months
[average current sales per customer in a defined period]: Sales $’s per customer for the previous month to customers who have been buying for more than three months
[number of current customers in a defined period]: Number of customers on our books at the beginning of the month who have been buying for more than three months
[customer churns in a defined period]: Percentage of customers who have been buying for more than three months who stopped buying this month
[sales to new customers]: Sales ($’s) to customers who have been buying for less than three months
[prospects]: Customers with whom we have communicated
[average sales potential per prospect]: Average spend per product range for a prospect
[sales conversion]: % of prospects we communicate with who become customers
Cascading KPIs continues down the tree until we find a point where we cannot measure the indicator or where the independence of one KPI from another becomes compromised.
The work in effectively cascading KPIs, however, is only half done at this stage.
Defining the KPI in terms of specific data and its point and time of origin is important. Data is often labelled similarly when it actually represents a different entity or the same entity at a different time in a cycle. For example, when is a sale a sale? Is it a sale when the offer is accepted, the product delivered, the money received or the possibility of returns eliminated?
Further, it is important that the range and target values be set. If we set a target of $1M sales, are we happy with $0.99M sales? What is considered to be the reasonable upper limit of our sales target range? That is, what is exceptional; $1.01M? $1.1M? $1.3M?
Our ability to cascade appropriate KPIs in this manner is limited by our knowledge of the real cause and effects within our business. It is further limited by our ability to define and measure KPIs. Not being able to adequately measure and not being able to adequately determine cause and effect should not stop us from cascading KPIs.
The risks associated with not being able to adequately measure and, therefore, be in control of our business as a whole, across diverse departments, should act as an appropriate justification to the business case to invest in processes necessary to be able to cascade KPIs.
