The goal of every business is to make a profit and so, by extension, the goal of every strategy should be to move that business towards optimal profitability. To this end, all strategists should first identify those actions which a business undertakes in its day-to-day operations that enable to it make a profit. These actions are called profit levers. As with mechanical levers, an adjustment made to a profit lever will result in a change in profit. Profit levers are not limited to the sales process, however. Profit is the difference between the money that comes into a business and the money that goes out of it and so, while some profit levers affect the ‘money in’ side of the business, others affect the ‘money out’ side and every business is an amalgamation of several such profit levers.
Although it is possible for profit levers to reinforce one another, it is more common for them to compete. One can therefore not optimise total profit by optimising each profit lever in isolation. To do so, one must take into account the nature of interactions between profit levers. A discount that increases sales volumes does so at a cost to gross margin and a new target market that is more responsive to an offer may also be more risky. In each case, maximum profit is achieved only when the correct balance for each profit lever is found.
The relationships between each of a business’ profit levers can be mathematically – or visually – represented in a profit model. The first step in building a profit model is to deconstruct a business into its most basic profit components. The profit model for almost any business can be begun as a function of sales revenue less all variable and fixed costs.
Profit = Revenue – Variable Costs – Fixed Costs
The simplicity of this equation is what makes it broadly applicable but it is also what limits its usefulness. To be useful, each of these components must first be further broken down into their more detailed constituent parts. For example, it is clear from the equation that an increase in revenue will lead to an increase in profit. However, ‘an increase in revenue’ is too nebulous a term to inform a specific course of action. The ‘revenue’ component must therefore be broken further into its component parts. In this case, revenue is equal to the number of units sold multiplied by the selling price of each unit, so revenue can be affected by adjusting one of these profit levers – by changing the selling price, etc. The other components must also be broken down until each fragment of the equation is a fully formed profit lever.
Profit = No. Units Sold (Sales Price) – No. Units Sold (Cost per Unit) – Fixed Costs
At this stage though, the profit model is only a template that shows the directional impact that an adjustment to one profit lever might have but not the size of that impact. Without knowing the size of any one impact, it is also not possible to know the likely combined impact of those adjustments. For example, by adding a new feature to its product – and therefore increasing the cost per unit – a company might be able to increase the number of units it sells. The profit model shows us that an increase in the number of units sold will lead to an increase in revenue but it also shows us that an increase in the cost per unit will lead to an increase in variable costs. What the model can’t do, therefore, is tell us if that increase in revenue will be sufficient to cover the increase in costs. Before it can do this, the model needs to be populated with organisation-specific information.
Customising a profit model is a two stage process. The first stage is used to describe the organisation’s current situation and the second stage to describe the likely impact of a new strategy. Describing the current situation should be a relatively simple process as most of the data needed is usually readily available from historical databases and published financial reports. A good understanding of the status quo is important but profit models are most valuable when they are used as forward-looking tools. For example, it is good to know the current profitability of a business but it is so much more valuable to be able to estimate how that profitability will be changed by a new strategy that, for example, discounts the selling price of goods or one that increases fixed production costs but lowers unit costs.
Building such a forward-looking profit model requires an understanding of two important concepts – marginal performance and test-and-learn analytics. It is important to differentiate between performance on average and performance at the margin or, in other words, to differentiate between performance in a stable environment and performance in that same environment once a change has occurred. Because new strategies always change the existing environment, their success is determined by marginal performance. A company that sells one thousand items at one price might sell two thousand if it drops it price by fifty percent; or it might still just sell one thousand. Before it launches a campaign to offer discounted prices it vital to know which of these new environments is most likely.
Historical data sources only show average performance. To correctly account for marginal performance, therefore, it is imperative to have access to forward-looking analytical tools that can calculate the change a strategy will have on its environment. It is clearly not possible to know the future impact of a strategy before it is implemented so test-and-learn analytics is not, strictly speaking, a forward-looking technique. Rather, it is a technique that gathers real-time marginal performance data in a series of small and controlled experiments. The results of these experiments are used to populate the profit model which can then be used to extrapolate the likely impact of the strategy where it to be rolled-out on a large scale. Rather than offering a fifty percent discount on all items, for example, the firm could randomly offer a few different discount rates to small groups of customers, measure the marginal performance of each of the groups, populate the profit model accordingly, identify the most profitable strategy and then offer that discount rate when the campaign is officially launched. All major decisions can therefore be made using profit models populated with real marginal data.