- Industry Watch: Adobe expects October online spending to grow faster than 2026 holiday season spending.
- The Take: Operation Smile Canada found which channels create donations and which ones capture them.
- From the Blog: Can MMMs measure campaigns, or only channels?
- Prescient Voices: Sharon MacPherson, VP of Direct Response Marketing at Operation Smile Canada, on stewarding donor funds.
- Three Things: Three reads on how MMMs work, cross-channel vs. omnichannel marketing, and balancing your full funnel.
Operation Smile Canada funds free cleft surgery and long-term cleft care for children, so every dollar donated really counts. To grow the donations it receives, the team runs paid media across five channels at once: Google Ads, Meta, Connected TV, Pinterest, and Microsoft.
Last-click reports told them to put more into search and pull back on Connected TV because branded search reported about 9x ROAS, while CTV looked like a near-total loss.
Prescient's marketing mix model read all five channels together, and the reports gave the team at Operation Smile a new way to consider the efficiency of each of them.
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.
Let's dive in.
INDUSTRY WATCH
October online spending is predicted to grow faster than 2026 holiday spending
Adobe expects US shoppers to spend a record $275.1 billion online from November 1 to December 31, up 6.7% from last year, including $47.5 billion during Cyber Week, up 7.4%. But October may be growing faster: Adobe projects $95.8 billion in online spending this month, up 8% YoY.
Giving runs on a similar calendar and has also seen year over year growth. GivingTuesday raised an estimated $4.0 billion in the US last year, up 13% from 2023.
Know what you're looking at going into this holiday spending season: Last-touch reporting credits the final click, so the channels that build demand early in the season tend to look weaker in those reports than they are.
Sources: Adobe holiday forecast (Sep 28, 2026), GivingTuesday 2025 results (Dec 3, 2025).
01 · THE TAKE
Which channels create giving, and which ones capture it
Last-click reporting showed search as Operation Smile Canada's most efficient channel and Connected TV as one of the worst performers. Each platform's reporting could only see its own contribution, and Connected TV reaches people who see an ad, feel moved to donate, but give days or weeks later through a different channel. In-platform reporting can't follow that journey.
Our marketing mix model effectively reassessed the efficiency of each of their channels and found that search actually came in last. Connected TV climbed to their second most efficient channel, with roughly 6x the contribution last-touch had credited to it. Pinterest moved from last to first, even on a small budget. Inside Google Ads, the branded search campaigns that reported about 9x ROAS ranked lowest in the MMM, because they mostly reached donors already searching for Operation Smile by name (meaning, they were capturing demand created elsewhere).
With that view, the team's question became which channels create demand and which ones meet it. They moved more budget into Connected TV through the first half of 2026. Spend scaled 3x from Q1 to Q2, and modeled efficiency held close to its Q1 level, with no sign of saturation.
A channel can look like your weakest performer and still be one of your strongest if your measurement has blind spots.
02 · FROM THE BLOG
Can MMMs measure campaigns, or only channels?
Operation Smile Canada's weakest spend sat inside a single channel: the branded search campaigns within Google Ads. Spotting that took a model that reads campaigns, and most MMMs stop at the channel level.
The post explains why. When dozens of campaigns rise and fall together, a traditional model can't tell them apart, so grouping them into channel buckets keeps it stable while hiding which campaigns are working. It also covers what decides how clear a campaign-level read can be: spend, how long a campaign has run, how much history exists, and how much it overlaps with others.
The payoff is much more transparency into potential budget moves. Instead of funding channels with a mixture of effective and inefficient campaigns, you can identify how to shift budget between them in order to maximize your return.
03 · PRESCIENT VOICES
At Operation Smile Canada, we take the stewardship of our donor funds very seriously and want to make wise, impactful decisions with every dollar spent. Prescient has given us the insights we need to make investments that ultimately help more children receive the cleft care they need and deserve, closer to their homes. Thanks to Prescient, every dollar is working harder for the children, which fulfills our donors' wishes and our mission!
04 · THREE THINGS WORTH READING
1. How do MMMs work? What happens between your data and your budget recommendation
A walk through traditional, open source, and Prescient's approaches to marketing mix modeling, including why two models built on the same data can disagree. It also covers the question worth asking of any MMM: how its accuracy gets checked against the revenue that came in. Read →
2. Cross-channel vs omnichannel marketing: The differences that matter for your budget
Operation Smile Canada needed one cross-channel view before moving another dollar. This post separates cross-channel from omnichannel and explains why both make channel-by-channel reporting unreliable, since connected campaigns drive revenue that shows up in other channels. Read →
3. Top of funnel vs bottom of funnel: How they differ and how to balance your budget
The general version of Operation Smile Canada's lesson: top of funnel creates demand, bottom of funnel captures it, and lower funnel channels often take credit for buyers already on their way. This post covers how to set your split and measure both ends in terms of revenue. Read →
The Halo
Exclusive insights, every week.
Subscribe to The Halo for sharper marketing thinking.