Practical Optimisation – A Guide to Budget Allocation for People Who Actually Have to Make the Decisions: Part 1

Practical Optimisation – A Guide to Budget Allocation for People Who Actually Have to Make the Decisions: Part 1

“How do we optimise our marketing budget?” is a question we get asked on occasions.

And, honestly, in our approach to it we have found again and again that simple things done well beat mathematical sophistication every time. Not that we are anti-mathematical sophistication (we do that too), but we have learned the hard way that it must be kept resolutely in its place, or risk wreaking havoc.

It’s a simple enough question – but with just enough devils lurking in the detail to make fools of our best intentions. With that in mind we present, herewith, our collective hard-won experience from years of optimisation.

 

Yet another Millenial Chief Exec, making yet another high-powered decision

 

Define what you mean by ‘optimum’

An obvious one, this, and so often overlooked. What are you trying to do? Maximise sales? Maximising sales is all very well within a fixed total budget, but sales always increase with budget size; more spend will always generate more sales. So then, we need to think about whether those sales are worth having for the additional cost it takes to generate them, i.e. marginal revenue or profit.

Maximising ‘market share’ is something of a red herring. Unless your optimisation software predicts what competitors will do (something we have NEVER seen) you will always have to assume their sales to predict market share. Therefore, to all intents and purposes, maximising your own sales amounts to the same thing as maximising share.

What about non-financial goals, such as brand building, or the need to amortise the fixed costs of a production line? You tell the business that you can increase profitability by doing 10% fewer promotions. The production manager points out that the business has just spent £20m on new machinery it’s still paying for. What’s going to happen? (And, in fact, this really did happen.)

For that matter define what you mean by ‘budget’. For example, if you are optimising your media budget, are you including the cost of agency fees? What about the fixed costs of production? Or even the salary of your media manager?

 

Start with ROI

Budget optimisation is affected by considerations such as cut through and diminishing returns (basically the shape of the sales response curves), but the single most important factor is the measured ROI. Often this is a good ‘quick’ guide to where a full optimisation will allocate the money. It won’t tell you the exact amounts to spend and what to spend them on, but it will give you a sense of which budgets would increase and which would decrease, were you to run a formal optimisation.

On the other hand, trying to make do without proper ROI measurement is asking for trouble. We understand why ad hoc rules of thumb such as share of voice are used, but they can only ever be broad generalisations at best and at worst lead you to very inefficient deployment and sub optimal ROI. Get some measurements.

If you do decide that directional results based on ROIs are not sufficient and you want a full optimisation solution, then here are some tips for getting the most out of it.

 

Recognise the effort and commitment involved

Optimisation is a process that always raises as many questions as it answers. Handing it over to consultants with the injunction to ‘optimise this’ will result in you being left with a plan which you are told is ‘optimum’ but without knowing why it is, or whether it is feasible to implement in practice. There is no short cut to committing serious time and effort to defining the problem and its parameters.

 

Making breakfast – the chicken’s involved, but the pig is committed. I don’t think the cow helped much at all.

 

That’s Part 1 of our article on practical optimisation. Read more in Part 2…

 

Philip Gaudoin

By |2018-05-01T16:13:05+01:00April 10th, 2018|Modelling, Optimisation|

About the Author: