No point claiming a £5 to £1 return if we were always going to get the £5. The question we should ask of every campaign is simple: did we change behaviour? If we didn't, we're being busy fools.
The problem with ROAS as most people measure it
A true return on advertising spend should exclude sales that would have happened anyway. Yet many of the numbers presented in marketing meetings don't.
Here's an example. If a customer searches for Colgate and we serve them a Colgate ad, that's hardly going to change behaviour. Preaching to the converted comes to mind. The ROAS looks fantastic, but most of those sales were coming regardless.
If we serve that same customer an ad for mouthwash or replacement toothbrush heads, we're much more likely to grow the basket. Then we can confidently claim credit for persuading them to spend more.
I'd like to see a new metric called IROAS, where the I stands for incremental. Or we simply agree that the R means incremental revenue.
Measure return on margin, not sales
I've worked for retailers where gross margin ranged from 12% to 75% depending on the category. A sales-based ROAS hides that completely. That's why I prefer ROAM, return on margin.
Our colleagues in commercial finance can set us clear targets with simple maths. If we spend £75k on marketing while trading at a 15% gross margin, we need £500k of additional sales just to break even. It focuses the mind, doesn't it? Higher-margin sectors need a smaller sales lift to wash their face, which is exactly why ROI should be based on incremental pounds of margin.
Get past the vanity metrics
It's easy to get bogged down in media stats: reach, impacts, impressions and share of voice. They're all very well, but they don't tell us whether the activity drove more profit once cost of goods and marketing costs are factored in.
Attention is the missing overlay. Digital can generate millions of impressions and reach the very best prospects, but only if we win their attention. Most impressions are now served on a small mobile screen to someone doing something else. A 0.03% click-through rate may be "above the industry average", but the CFO is unlikely to be impressed.
What we say matters as much as who we say it to. Creative and media need to work hand in glove. Spending precious advertising pounds telling customers what they already know is not going to change behaviour.
How to prove what's incremental
- Test and control. Hold some stores, regions or customer segments back so you can attribute any lift with confidence.
- Marketing mix modelling. Back in the day this was called econometric modelling. It shows how weather, pricing, competitor activity and marketing each move the needle, and where diminishing returns kick in.
- Set objectives up front. Higher footfall or a bigger basket? New customers or more from existing ones? Picking a lane makes success easy to judge.
- Include lifetime value. Modelling repeat purchases can quickly move the ROAS scales for new-customer campaigns, a bit like a loss leader.
But don't let the maths run the business
There's a flip side. Marketing driven purely by maths is a dangerous game. Those who obsess over the lowest CPA or the highest ROAS won't grow a business in the long term.
Brands grow by selling more things to more customers. If we only target hot prospects to keep costs down, we ignore everyone who might buy next week, next month or next year. Defaulting to PPC, social, Performance Max and CRM puts too much emphasis on people who need something right now.
Shoppers are also far more spontaneous than they used to be. Most haven't decided what to buy before they walk into a shop. Tangible, visible things like a well-targeted leaflet on the doormat build mental availability. It's not sexy and doesn't offer real-time analytics, but that doesn't mean we shouldn't do it.
The balance to strike
Be ruthless about incremental return on the activity that's meant to drive sales now. Be clear, and honest with the CFO, about the brand activity that's meant to pay back later. Play the long game and the short game, and measure each on its own terms.
Done well, you can spend less and grow more at the same time. A good marketing review almost always finds money being spent on customers who were going to buy anyway.