|
Well, hello, and welcome to this week's edition of the eCommercer.
Here's a thing I keep bumping into as an ecommerce owner. The numbers that make you feel best about your ads are often the ones telling you the least. You look at the dashboard, the ROAS is up, everyone's happy, and the bank account quietly disagrees.
This week's guest has spent eleven years living inside that gap, and he's got a fix that sounds completely mad the first time you hear it. So naturally, I loved it.
The Question That Started It All
My guest is Brenden Delarua, Co-Founder and CMO of Stella, and the story he tells goes right back to his days as a media buyer. He had a big account running across Meta, Google, Criteo and a few others. He hit every ROAS target he was given. Smashed them, actually. Record-breaking stuff.
Then came the quarterly review, and the CFO asked him a simple question. If our ad numbers are this good, why is revenue down? Why isn't any of this showing up in our bank account?
That question sat with him for a long time, and the conclusion he came to is one a lot of us need to hear. ROAS is not correlated with profit, even though most people assume it is.
His point is that ROAS is a correlation metric. Somebody saw or clicked an ad and later bought something. That's all it tells you. It doesn't tell you whether the ad caused the sale. And once you're being graded on a number like that, the temptation is to chase easy wins like retargeting and branded search, because they push ROAS up fast. The dashboard looks brilliant. The business stops growing.
| | If you want a high ROAS, I can get you a high ROAS. But eventually it catches up to you, because the business is not growing.
— Brenden Delarua |
The YouTube Ads That Looked Like a Disaster
Here's the story that really made the penny drop for me.
Brenden runs YouTube ads for a client, and YouTube is famous for a terrible click-to-convert ratio. So on the platform, the ROAS was sitting at 0.2. On paper, a money pit. The kind of number that gets a channel switched off in a meeting.
But he pulled the post-purchase survey data, and around 7% of buyers said they'd come from YouTube. Once he weighed that slice of revenue against what they were actually spending, YouTube wasn't a disaster at all. It was roughly a 3x return and genuinely profitable.
Same ads. Same spend. Two completely different stories, depending on which number you trusted.
Where a Non-Technical Founder Actually Starts
Now, before you panic, the big-brand version of this involves geo-holdouts, where you deliberately turn your ads off in certain regions and watch what happens to your real Shopify revenue. It's the cleanest way to find out which ads truly cause sales. It's also not where a smaller brand needs to begin.
For most of us, Brenden's advice is refreshingly doable. Get two things in place.
- A proper multi-touch attribution tool (he rates Triple Whale and Northbeam), and
- Post-purchase surveys, so your actual customers tell you why they bought.
The surveys are the bit I'd not thought enough about. His favourite question, which he credits to Jarrell Blades over at Tushy, is asking people what almost stopped them from converting. That one question tells you where your messaging is leaking, and it costs you nothing but a moment of a customer's time.
Why This One Lands for Me
I sell omega-3, and I've lost count of the times I've stared at a channel report trying to work out who really earned a sale. Brenden reminded me of a chat I had a couple of years back with a chap who'd tracked one woman buying a single pair of shoes, and there were 236 touchpoints before she bought. Paid, organic, social, all of it. So which one gets the credit?
The honest answer is that no single platform deserves 100% of it, yet every one of them claims it. That's the whole ROAS problem in a nutshell. What I like about this conversation is that it doesn't just name the problem. It gives you a first step you can take this week without a data science degree.
| 
