This week: a case study
This week I’m talking about a brand I personally love: Ka’Chava.
Ka’Chava tested out Northbeam Apex, one of our newest features, to incredible effect. Like most of you, Ka’Chava is focused on new customer acquisition, and their success with Apex was so amazing we had to bring you this banger case study to explain how they did it.
Let’s tee it up.
The brand
Ka'Chava is an all-in-one nutrition shake built to deliver comprehensive plant-based nutrition from whole-food ingredients. In a crowded health and wellness category, Ka’Chava differentiates on ingredient quality and taste, which has earned them a massive base of loyal customers.
Like most of you, dear readers, Ka'Chava budgets a meaningful share of its media spend on Meta. Rather than relying on ad-platform-reported metrics, Ka’Chava uses Northbeam to measure Meta’s true impact and decide where their next marketing dollar goes. ‘
The challenge: Meta wasn’t optimizing for new customer acquisition.
Ka'Chava's growth depends on first-time buyers. Returning purchases matter, but the marketing team’s performance is measured against the hardest goal: new customer acquisition.
The problem: Meta's default ad delivery system doesn't distinguish between new or returning customers. In Meta’s standard attribution, the delivery algorithm applies its own logic to decide which ads earned purchases. It then prioritizes delivering ads that drive lots of conversions, regardless if a conversion is a new or repeat buyer.
That works well for plenty of advertisers. For Ka'Chava, it meant the algorithm was spending in the wrong direction.
The team at Ka’chava needed Meta’s ad algorithm to focus on delivering ads to potential net new customers. Anything else is wasted budget.
This need created a dissonance. While every budget decision was based on Northbeam's measurement, every ad delivery decision was based on Meta's algorithm.
Both systems were trying to optimize performance, but they disagreed on what good performance looked like. Meta is a “walled garden
Ka’Chava needed a solution to both properly identify new customers, and a way to advertise exclusively to them on Meta.
The solution: Ka'Chava closed the optimization gap with Apex.
Northbeam Apex is a direct integration between Northbeam and Meta that feeds Northbeam's multi-touch attribution data directly into the platform's delivery algorithm.
This means Ka’Chava could optimize their Meta ad delivery using Northbeam’s performance data.
On the Northbeam side, Ka'Chava turned on Apex to begin sharing data with Meta.
Then, using the Custom Attribution setting in Meta Ads Manager, Ka’Chava used the Northbeam data to optimize their ads for new customer acquisition.
Instead of letting Meta decide which ads deserve credit for a purchase, Northbeam does the attribution first and tells Meta what actually drove the sale. This means Meta’s algorithm can distinguish between new or repeat buyers, and only target ads toward net-new prospects.
For Ka'Chava, the workflow looked like this:
Northbeam measured performance. Using its first-party data and multi-touch attribution model, Northbeam determined which marketing touchpoints contributed to each purchase: not just the Meta clicks, but the full customer journey across channels.
Ka'Chava chose the goal. Instead of optimizing for all purchases, Ka'Chava told Northbeam to optimize for what mattered most: revenue from first-time customers.
Apex shared that signal with Meta. Rather than sending Meta raw purchase data, Apex sent Northbeam's attribution results, so Meta could see which conversions actually matched Ka'Chava's goal.
Meta optimized toward that goal. Meta's delivery system began finding more people who looked like Ka'Chava's highest-value customers, using Northbeam's attribution instead of Meta's default model.
One of the biggest advantages was that Apex fit into Ka'Chava's existing setup. It didn't replace the Meta Pixel or Conversions API, didn't duplicate conversion events, didn't reset campaign learning, and required no engineering work from Ka'Chava's team.
The test: how well does Apex work?
Naturally, Ka’Chava wanted to understand how using Apex could improve their performance beyond their baseline.
To explore this, Ka'Chava ran a controlled A/B test at the Meta ad-set level with identical campaigns. The only difference was how the algorithm was optimized.
Cell A, the test: These ads had delivery optimized against Ka'Chava's Northbeam Apex signal via Custom Attribution.
Cell B, the control: These ads had delivery on Meta's standard attribution settings (business as usual).
Results were read in Northbeam, the same system Ka'Chava uses to compare performance and allocate budget across every channel.
The impact
The results were immediate. Optimizing Meta against Ka'Chava's own attribution produced a massive lift in new customer efficiency. According to Northbeam data:
Metric (measured in Northbeam) | Apex vs. BAU control |
|---|---|
New customer return on ad spend | +43.2% |
New customer acquisition cost | −25.2% |
By using Apex, Ka'Chava acquired more first-time buyers at a materially lower cost, making new customer growth more sustainable.
Critically, these gains reflected incremental new customer growth, not a reshuffling of existing attribution credit between channels.
This is a huge unlock with low setup effort and low downside. All Ka’Chava did was funnel their Northbeam data back to Meta with a few clicks, and found +43% new customer ROAS nearly overnight.
It sounds too good to be true. I like to think of it as “putting rocket fuel in your car engine.” As any data scientist would tell you, ad delivery algorithms can only be as strong as the data and signals fueling them.
These huge lifts in performance are what happens when you feed the world’s most accurate marketing data into the world’s highest-converting ad algorithm.
In Ka’Chava’s own words:
“We hold every channel accountable to our own attribution, so it matters that the Meta delivery system is now optimizing against the same standard. A 43% lift in new-customer ROAS isn’t just incremental improvement: it’s a fundamentally different level of efficiency that lets us acquire customers we couldn’t reach profitably before.”
Ka’Chava is not the only advertiser seeing success with Apex. We ran similar A/B tests across a variety of brands this summer. Here’s what we saw:
By the way, Northbeam Apex is available now for all Northbeam customers at no extra cost. Simply go into your settings and activate it today.
✂️ Meta is removing placement controls from ad sets. Media buyers are still losing ground in Zuck’s war to cut out the middleman. It’s time to start adapting.
🤖 Meta’s AI can now audit your campaigns and provide recommendations. Once again, if you are a human media buyer, you need to be worried.
🤔 Adding the “created with AI” label to ads cuts clickthrough by 31.5%. Beware of your own marketing echo chamber: the average American is distrustful of AI. I mean, haven’t you all seen The Terminator?
Love this newsletter? Please forward it to your friends, coworkers, or Jed York. Subscribe here to get this data in your inbox every week.