Welcome back to The Halo. In this issue we cover:
- Industry Watch: OpenAI released ads. Real accounts are hitting 1.49x ROAS in 15 days. Prescient launched a new native connector to measure it.
- The Take: How Quiet Owl bet on YouTube for BillieBars, and why platform reporting almost killed it.
- From the Blog: What YouTube actually contributes when a marketing mix model can see the halo effects.
- Prescient Voices: David Sinton, Co-Founder at Quiet Owl, on unlocking top-of-funnel spend with confidence.
- Three Things: Reads on top-of-funnel measurement, MMM data quality, and model refresh cadence.
Do you remember the memes and the marketing from competitors when OpenAI announced ads? Six months later, they have not aged well.
Real accounts are hitting numbers that would make Meta and Google campaigns look average. Most brands are still watching from the sidelines. The ones moving early get one advantage the rest cannot buy: history in the model.
Marketing agencies get that advantage faster than single brands. They run the same play across multiple clients and learn from each one. Quiet Owl is this week's example. They bet on YouTube for BillieBars, a high-AOV brand where most agencies would not push top-of-funnel. It paid off in three record sales months in a row.
Welcome to The Halo. Each week: one piece of measurement analysis, the best of the blog, one real customer result, and three reads worth your time. Plus a new release you'll want to know about.
Let's dive in.
INDUSTRY WATCH · NEW RELEASES
OpenAI released ads. Real accounts are hitting 1.49x ROAS in 15 days.
Opascope ran a benchmark with 15 days of real spend data. The sample seems small, but the results are worth looking at. One of the brands inside the agency was among the first to run OpenAI Ads. They got a 1.49x blended ROAS. $60K in spend, $89K in revenue.
It's hard to achieve these results in your first move on a new channel, especially in 15 days. It generally takes about 3 weeks to learn the ins and outs of the new channel and generate a good ROAS. In this case, ROAS scaled from 0.6x to a peak of 2.94x in 10 days.
But there's a caveat: you can't fully trust platform reporting, even with last click data. A platform is good at tracking cost per click, but it's never able to tell you anything actionable about your spend.
That's why we built a new API connector between OpenAI and our MMM. With this integration, you can compare the platform dashboard against the full picture of channel performance including its halo effects. You'll start building performance history to power forecasts from the moment the API key goes in.
Source: Opascope, ChatGPT Ads Benchmarks From 15 Days of Real Spend. Connect OpenAI Ads →
01 · THE TAKE
How Quiet Owl bet on YouTube for BillieBars, and why platform reporting almost killed it
Quiet Owl agency ran a top-of-funnel test on YouTube for BillieBars, a high-AOV brand where most agencies would not push awareness. The short-term results looked disappointing, and the in-platform numbers suggested only one path forward: kill the campaign.
David Sinton, Co-founder of Quiet Owl, asked a different question: what if the YouTube dashboard wasn't showing the full picture?
Prescient's Halo Effects report gave him the answer. YouTube was building real demand, but the in-platform reporting couldn't see or quantify it. With that data in-hand, he scaled instead of cutting.
Example from an anonymized account: 75% of YouTube's value here is halo, the demand last-touch never credits.
David didn't get lucky. He had a source of truth the YouTube dashboard couldn't offer.
02 · FROM THE BLOG
What YouTube actually contributes when you can see its halo effects
YouTube can look lackluster when you're looking at in-platform reporting. That's because YouTube's dashboard can't see the spillover effects of your campaigns.
The post walks through what YouTube actually contributes when you can see its halo effects. You'll also get tips on how to confidently scale campaigns before YouTube dashboard reporting justifies it.
Quiet Owl's 600% YouTube spend increase for BillieBars is the real world example of every claim this post makes.
03 · PRESCIENT VOICES
Prescient allowed us to unlock top-of-funnel spend in early April. Our client has reached record-breaking sales levels every month since then.
04 · THREE THINGS WORTH READING
1. Top of funnel visibility loss: new causes and what to do
The bridge to both threads this week. Top-of-funnel visibility loss is not new, but AI overviews and LLMs are making it worse. The post walks through why awareness spend keeps failing to get credit, what changed recently, and how measurement architecture can surface the actual contribution.
2. Your MMM is only as good as the data going in
The direct companion to this week's Industry Watch. The model cannot forecast what it cannot ingest. Connecting new channels early is how the model learns what to do with them later.
3. How often should your MMM refresh?
Do you have enough information to act quickly and confidently on your campaigns if your MMM reports weekly? The refresh cadence of your marketing mix model directly affects every budget decision you make and its efficacy.
The Halo
Exclusive insights, every week.
Subscribe to The Halo for sharper marketing thinking.