Editors note – This was supplied to NZ Marketing Magazine pre Covid-19.
Thanks to Daniel Kahneman the idiosyncrasies of human behaviour are increasingly well known.
A nice gambling example involves a thought experiment of a lottery with ten tickets, a $1 ticket price and a $20 prize. When presented this way almost everyone buys in. With enough money and time you are guaranteed to double your money. Minimal risk, lots of reward. However, when the same lottery is presented with your mate Dave having the nine other tickets most people don’t play. Same lottery, same risk-reward but it feels different. Casinos are also full of idiosyncratic behaviour and poor risk-reward management, possibly due to too many G&Ts. Putting it all on 25 is far more speculative than the measured tactics displayed by (some) blackjack players. One is a total punt with fixed, low odds. The other uses data and strategies to make the odds much better.
Where a business chooses to invest its money is pretty similar to gambling. Balancing risk- reward to maximise return from a finite amount of money. Limit risk too much, there’s little reward. Too much risk, no job and/or company.
There are numerous ‘games’ we can lay our marketing chips on – Marketing, UX, call centres, NPD, distribution and events all have different risk-reward profiles over different time frames. Where businesses differ is
they tend to be full of different people and departments all fighting over the same chips (unlike gambling where one person makes all the decisions over where to lay the chips and gets all the feedback re losses and gains).
Too often there isn’t a helicopter view of the whole game and the total chip situation is unclear. This is one of the reasons there have been so many articles written and spoken about recently on the ratio between top and bottom-funnel marketing, an issue largely created due to different departments focussed on disparate metrics and different timeframes.
What hasn’t been written or spoken about anywhere near as much is the optimal amount to spend on broadcast advertising.
The IPA (The Institute of Practioners in Advertising in the UK) have been researching this field for some years now. Their data clearly shows that not only are most brands not spending at an optimal level, they’re underspending so much that it is reducing the effectiveness of advertising overall. This can easily turn into a vicious cycle whereby broadcast advertising has a lowering ROAS (return on advertising spend) which results in a reduced broadcast advertising budget which results in even less effect etc etc.
This is one of the dangers with using efficiency metrics like ROAS or ROI (return on investment) as the key metric. Over short to medium time frames it’s easy to increase by reducing the investment or ad spend part. As long as the return drops at a lower rate than the budget cut, which in the short term it usually does, ROAS goes up even though less product has been sold.
Last year there probably wasn’t a higher profile example of the negative effects of underspending in broadcast advertising than the merged Kraft-Heinz. In 2017, the year after new majority owners 3G Capital and Berkshire Hathaway merged the two companies, advertising spend was down 39 percent.
Ultimately this led to the writing down in value of over $15 billion in brand value and the current share price is around a third of where it was in 2017. ROI would probably have looked great for a year or so but over a longer period not so much. Which is somewhat ironic given Warren Buffett’s investment strategy to play the long game. Recently, a senior analyst from analytics consultancy Analytic Partners, told me that in the last few years he’d seen only one client across the globe spend as much as they could (or should) be spending to maximise profit. The concept of an excess share of voice (ESOV), the ratio of a brands share of category spend versus its share of the market, was discovered about 30 years ago by John Philip Jones.
Historically it was based on TV ratings (hence share of voice) but now incorporates all spend (essentially ESOS). It shows that, all things being equal, a share of spend well over a brands share of category will eventually result in share of category increasing. Conversely, a share of spend well beneath a brands share of category will eventually result in a reduction in share.
A 2016 paper in The Journal of Marketing Research found it still held true in today’s post digital era. In July last year Mark Ritson spoke about ESOV for Marketing Week. He referenced discount supermarket Lidl who struggled for decades to compete with Tesco, Sainsbury’s, Asda and Morrisons. It’s market share never got any higher than 3 percent until between 2013 and 2017, they increased their share of voice from 5 percent to 19 percent and their market share doubled by 2018. A 3 percent share increase may not sound like much but the category is worth over £190bn and it resulted in £2.7bn in incremental sales. Last Christmas sales were up 11 percent yoy.
Obviously marketing budgets aren’t finite but unless something else is wrong it’s the gamblers version of counting cards in blackjack. You won’t win every hand but over time it’s a good bet. Spend more, get more. Minimal risk, high reward.
Disturbingly, alongside not spending enough on broadcast advertising, there is evidence to suggest there is now more being spent on making new products than advertising them. Obviously this makes sense for tech firms but an article last year in the Harvard Business Review reported that manufacturing firms are now spending more than 6 times as much on R&D as they do on advertising. Manufacturing ad spend has dropped from 1 percent of total expenses in the 1980’s to circa .08 perent today, whereas R&D has increased from 1 percent to more than 5 percent.
It’s clearly difficult for large companies to optimise these spends if the decisions are being made in different departments but with NPD failure rates reported to be somewhere between 40-80 percent, spending at this level on R&D is the equivalent of new product roulette. Furthermore there’s not much point in creating new products if you haven’t got enough budget to advertise them. For all their innovative new products, Apple spend quite a lot advertising them and few companies make products spoken about as much as theirs.
A few years ago Jeff Bezos said advertising is a tax companies pay for making unremarkable products. According to Kantar, in 2018, Amazon was the fifth biggest advertiser in the US (with 37 percent of spend on TV) and this month Campaign UK reported that they are now the largest advertiser on the planet with a spend of $11 billion (almost 2 percent of global advertising). It’s not that the products aren’t good but they now have competition from Apple, Google, and Disney but Amazon products aren’t any more or less remarkable than the competitors.
It seems making a product so remarkable it doesn’t need advertising is rather hard.
Some of the other new (and large) spenders in broadcast advertising in recent years have been Uber, Netflix and Airbnb. Once the markets these brands created increased in size, competitors launched. In competitive markets, just like chocolate bars or insurance, the role for advertising is to help get potential customers to choose one company over another – Uber not Lyft, Netflix not Disney and Airbnb not Trivago.
No doubt R&D has a role to play in the growth of companies but perhaps there is something to be said for taking a more Darwinistic approach. Wait until a start-up company has grown to a size that it is worth buying and buy it. It’s probably more expensive than internal NPD but it’s a known return for a lot less risk and overall likely a more efficient use of expenses.
So, in marketing as in gambling, you can’t lose what you don’t bet but you could be losing anyway. And, you’re probably betting off spending less time at the roulette table of NPD and more time making sure you’re spending enough chips at the blackjack table of broadcast advertising.
Simon Bird is the group head of strategy & measurement, PHD New Zealand