We all live Life at the Margins so why do we insist on analysing the Average?
In the 21st Century, we’re all living increasingly rapid lives. The phrase “life at the margins” reflects this – we’re often making decisions about the next best activity or trying to work out the benefits of tiny differences between things.
Scientists call these decisions "marginal decisions" – decisions that we make where we try to judge the outcome of very small changes in the impact of an activity. Good decision-making requires us to compare the additional (marginal) costs of alternative actions with the additional (marginal) benefit that each action might generate. Most choices involve doing a little more or a little less of something; few choices are all or nothing decisions.
Marketers are not immune to having to make these marginal decisions on a daily, weekly or annual basis. Decisions which require this thinking might include which particular promotion to run in a specific period, in which key account to provide extra support and through which advertising channels to promote products.
So how do we make these decisions? What information do we require? If we’ve been good and are following best practice, we will no doubt be looking for data to help us make these decisions. The most frequently used metrics when trying to make these decisions is that of “Return on Investment” or ROI.
When most people talk about ROI, they normally mean something called the Average ROI. For example, suppose a promotion costs £50,000 to run and the profit generated is £100,000, the Average ROI ratio is 100,000 / 50,000 = 2*. Typically we’ll have a series of these average ROI ratios to look at and then we’ll need to use this information to make our investment decisions.
So the next question is this - does the Average ROI ratio provide enough insight to help marketers make the best decisions? Do we just invest in the one with the highest average ROI? The answer to this questions is “no” and the reason should be obvious if you’re familiar with the Law of Diminishing Returns.
One of the most important things we learn, when marketing our products and services, is that the law of diminishing returns applies to marketing. This means that we acknowledge that doing twice as much of something rarely if ever gives twice the impact. We therefore need to evaluate our spending options at different levels to establish just what impact diminishing returns will have on the levels of return that we can expect. In this example, we will estimate the returns from an activity at three different levels of investment.
Take a look at figure 1. Here the Average ROI figures are calculated for us at three points on the chart – you can see that the value of the Average ROI is changing as we move along the curve. This is important and often missed in more simplistic analyses.
Now look at Figure 2 – here we’ve joined the dots and we can now quickly see what the relationship is between investment and the return we expect to see from an activity. The “Marginal ROI” is the return we get for the next unit invested – in mathematical terms this is known as the gradient of the line.
If we were to compute the marginal return at each point on the curve, we’d quickly be able to work out where the marginal return went from being greater than 1 to less than 1. It’s this point that’s key – at this point, your marketing is going from being profitable to being unprofitable**.
Computing the marginal return involves using some fancy maths or more simply we can use a package like modelQED to work this out for us.
A little bit more work with a tool like optimiseQED will let us know how much more or less we should be investing in the activity given our ROI objectives. This is a far more powerful way to address the resource allocation challenge than by either using guesswork or simply investing in the activity which generates the highest average ROI.
As the Marketing Effectiveness industry evolves, we will increasingly see marketers talk about Marginal ROI as the killer KPI – the one that matters most and helps them make decisions. Make it a part of your vocabulary now and begin to share this knowledge with others.
*There is variation in the definition of ROI with some preferring the formula (Return – Cost) / Cost. The definition used in this article (Return / Cost) is directly comparable to the definition above although care should be taken when using either ratio to ensure the definition is clearly given.
**For the purposes of this article, we have assumed that all response functions are linear or diminishing and so-called “S-shaped” curves are excluded. Interpretation of this family of curves should be done with a little more care using a tool like optimiseQED to determine the right investment strategy.