In marketing, there is no substitute for knowing your work is driving better outcomes for your business. Just like the got milk ads implied, there are alternatives, but they are not the same.
I recently had a spirited debate with a friend about whether there was such as thing as non-incremental return on ad spend. While I was forced to admit that I have heard a retailer or two mention this metric, I refused to acknowledge that it’s a real thing. We agreed to disagree – at least for the moment.
As with many arguments, I realized after the fact what I should have said: non-incremental ROAS is a non sequitur. It contradicts itself. ROAS is what you get back from your marketing investment.
Similarly, incrementality is what changed as a result of your marketing. So non-incremental ROAS is “what didn’t change” “that you got back” from your marketing investment.
Intellectual honesty
I had a great boss at P&G who used to say “don’t be intellectually dishonest” if someone was trying too hard to rationalize a decision or explain a market change by using sketchy logic. While it took me a minute to process, the phrase has stuck with me for a long time. I know what that boss would say if someone tried to sell him non-incremental ROAS. I hope you agree.
So now that we are being intellectually honest, we can admit among us marketers that not every campaign improves revenue or profit for our business. There are many reasons for this but this isn’t a marketing textbook. So let’s just say that when you consider the 4Ps, there are campaigns with a chance to succeed and ones that do not. If you consistently find yourself in the latter category, consider Warren Buffett’s quote about brilliant management and difficult industries,
“When a management with a reputation for brilliance tackles a business with a reputation for bad economics,
it is the reputation of the business that remains intact.“
For campaigns with a chance to succeed, there are many levers you can pull to drive incremental revenue for your business. Some of these are bread and butter tactics that are sometimes forgotten in programmatic advertising. Others are more like white truffles from Alba that seem like small details but make an outsized impact.
- Increase reach
The most important metric in traditional media, reach, is often ignored in digital buys because it’s difficult to calculate and plan. Many digital campaigns only reach 10% of retail shoppers. This may work for smaller e-comm businesses, but large ones require some reach>purchase intent>recall>buy math to know if 10% reach will move the needle with your business. Ads can only drive action if they are seen, after all. - Improve targeting
In many digital measurement studies we run, the buyer penetration rate among the exposed audience is the same as the U.S. population. This is surprising as you would expect campaigns to be focused on either attracting new shoppers or driving more revenue through existing shoppers. And if so, effective targeting should result in a higher or lower penetration rate than the national average. But in programmatic, there is often a disconnect between brand objectives and activation. It occurs when the third party audiences are selected for the campaign.
Marketers can ensure their digital audiences match their desired target audiences by asking their media agency one simple question and then asking the logical follow ups. Start with “How do we know the audiences we select are accurate?” More detail on this topic is available in our Adweek article. Whether your goal is conquesting, loyalty, frequency, basket size or something else, audiences built from actual consumer purchase behavior ensure you aren’t wasting digital spend by mass blasting your ads. - Get noticed
In many marketing contexts, assuming you are similar to your target consumer is fraught with issues. But here’s one case where it’s not. You know how well trained you are to ignore the banner ad spaces at the top and right of the articles you read online. And you are in marketing! Consumers brains are also trained to ignore most of the 2,000 ads they see each day. Blend in at your own peril. - Use the biggest data set you can
Just like other forms of research, the larger the sample, the greater your ability to see statistically significant differences. In campaign measurement, large data sets are especially important to prove the impact of your campaigns because of the match rate waterfall from ad impressions to outcomes data. Here’s a sample waterfall with made up but representative data:- 30% of ad impressions are resolved to identity (ID)
- 25% of IDs are unique (assumes 4x frequency)
- 75% of IDs are matched to an ID in the outcomes data
- 0.5% of IDs took an action during the campaign measurement period
With this waterfall, you’d expect to see one outcome for every 3,500 impressions. That’s actually a relatively good scenario. The size of the measurement dataset drives the numbers in both c and d. And the size of your business and length of campaign drives step d as well. Smaller data sets can make it 10 times harder (or more) to prove the incremental impact of your ad campaigns.
5. Optimize
Increasing sample sizes also allows you to drill into the data with more confidence. Knowing that a recent campaign drove a 2% sales lift is nice, but the information isn’t actionable. If you could further see that one audience had a 4% sales lift but two other audiences showed zero increase, that becomes highly actionable information.
6. Compare apples to apples
Measuring incrementality requires a test vs. control approach. A high-quality control group should be as similar to the exposed group as possible. Often this is defined in terms of demographics and location, but you also need to ensure that the test and control groups have the same likelihood of purchase before the advertising period. If they don’t, all the measurement results will be misleading. Poor controls can lead to “optimization” that actually hurts the sales of your business. Ask your measurement partner to see a comparison of your test and control groups pre-campaign to ensure they are similar based on outcomes.
7. Speak the CFO’s language
The leaders of your company are not interested in impressions, engagement and clicks. Their objectives are sales and profit. Tie your activity to sales impact to avoid credibility issues. This helps marketers in a more indirect way than our other tips, but there is nothing wrong with driving the business and looking good in the process.
8. Test, test, test
When I started my career in marketing, test vs. control methods were time-consuming. I had to air ads in one market and find a similar market to compare it to. Running multiple experiments at once was difficult, if not impossible. In digital marketing and increasingly with CTV, it’s possible to isolate people who have been exposed to your ad and compare their outcomes to individuals who haven’t seen your ad. Take advantage of this ease of execution to constantly test and learn what works for your brand.
9. Be compelling
Tip #9 is job #1. If you can’t convince your target why they should buy from you, tips #1 through #8 won’t matter. And creating a compelling 300×300 pixel display ad is tough. Review your marketing in the context of a sample webpage or app as a consumer would see it. Seeing things through the consumer’s lens is well understood when reviewing packaging of CPG products. It’s critical for display ads, too.
About Commerce Signals:
Commerce Signals is the largest source of credit and debit card data for marketing use cases. With a permissioned and anonymized view of consumer spending behavior, Commerce Signals’ powerful insights, accurate audiences and campaign measurement help eliminate waste and boost marketing ROI. Its solutions are used by some of the largest retailers, direct to consumer and AdTech companies in the country