The second part of our guide to real-life optimisation. Read Part 1 here.
Define the trade-offs you are interested in… and those you are not
Incredible as this may seem, not everyone realises that optimisation invariably creates winners and losers. But when you think about it, someone taking a bigger slice of a fixed pie is always going to leave less for everyone else. Even in the happy case that the board has the money machine switched to ‘blow’, and talk of bigger budgets is filling the air like the excitement of an extra Christmas, everyone is still competing for a bigger share of the increase.
Our advice is to make sure everyone understands this in principle before the results start coming in. You can greatly increase your chances of success by defining the trade-offs in which you are – and are not – interested. Reallocating money from France to the UK (to take a random example) is always going to cause consternation on the south bank of la Manche. So, unless you really mean to do that, make it clear you don’t.
Let the maths be your guide – reality trumps theory every time
Take it through to execution.
You’ve optimised. You are tripling the budget on your main brand. You brief your media agency. They point out that the new budget means deploying 400 ratings a week for 4 weeks. Of course, that is far too high to be sensible. It may not even be possible to buy that many ratings without paying a hefty premium which, of course throws out your calculations. Therefore, be wary of simple annual optimisations that are not practical (it’s why we do all our optimisations weekly so that we can set real-world constraints – quick plug. Ed.)
Unicorns are pretty but impractical and not real
Maths is continuous, reality is lumpy
You’ve optimised. The optimiser said spend 25% less a year on TV. So, instead of the two campaigns a year you currently run, you do… what? One and a half? Which one do you halve? Or do you do a heavier deployment of a single campaign? But maybe that is too heavy for a single campaign. Again, only a weekly optimisation will tell you the answer to this.
Chunks, rather than lumps, but you get the idea (it’s a metaphor)
Other people have a say
Ever tried paring back a sales budget? Even if you can persuade the board it’s the right course, even if you win over the notoriously recalcitrant Sales Director to your side, even if the Production Direction can flex the line to make fewer units… you’ve still got to persuade the buyers at your customers that it’s a good idea. And – guess what? – they refuse point blank.
When roused, Sales Directors are notoriously quick to reach for the nearest available weapon
Which brings us to the next point…
Create consensus among stakeholders
Make the model work for your stakeholders. Suppose it tells you to spend 20% less on Brand X, but the brand manager believes that will kill the brand. You might be better off agreeing to a 10% cut this year, and to measuring the impact, than fighting tooth and nail for the original 20% cut.
Above all, be sensitive. When budgets are discussed, the thought of job security is never far from people’s consciousness. And, not only do people worry about their jobs, but, believe it or not, they also feel passionately about the brands they manage, treating them on occasions as their own children.
They can never say they weren’t consulted
Optimising a budget fully always takes several years in practice
No one ever wants to go straight to the optimum budget allocation in a single bound. Nor should they, for the simple reason that it’s too risky. Optimisation can produce extremely polarised outcomes. At the same time, it cannot possibly factor in all the real-world considerations that might make us think twice about slavishly doing what the model says. So, unless it’s an emergency and you are fighting for survival, it’s better to go slowly, taking a few steps each time, with each win building confidence in the process. We’d suggest optimising within brand first. Start with media only. Moving money between ad campaigns within the same brand. Then maybe within brand group – brands that sit within the same Marketing Manager’s remit. It’s key to identify who is making the decisions here.
The pyramids – also took a long time. And many camels.
Let the final say be human
Our final suggestion is an obvious one, but still worth saying: models don’t make decisions – people do. We are all grappling with reality in one way or another, and anything that appears to make our lives simpler can have the appearance of an oasis in the desert. Two easy – but false – paths should be resisted. The tendency to deify the model (usually meaning we have not understood the problem) or the tendency to reject the model (usually meaning we have not understood the model).
We recommend, as a middle way likely to yield the best outcomes, the path of reasoning, questioning engagement with the model. Yes, listen to it. Yes, understand why it says what it does. But do also question whether it is right, consider what turn of events might make it wrong, and proceed with eyes open.
To get optimised, contact us.
Philip Gaudoin