🪩 Volume 145 | July 22, 2026 |
|
I need to confess something that cost me about six months and an embarrassing amount of ad spend. Last year, we ran ads. A LOT of ads. All pointing at a quiz and a lead magnet, all built to do the thing you're “supposed” to do: Capture the lead, nurture the lead, convert the lead. I was IN the data. I watched cost-per-click, cost-per-lead, the little ROAS (return on spend) number that either makes you feel like a GENIUS or like you should've kept your corporate job. Some weeks the numbers looked great, other weeks they looked like a library on a Saturday night (quiet, mildly depressing, one guy asleep in the corner). But heeeey like all the wrenches that life likes to throw, I learned that people didn't actually want a self-serve funnel (WHAT HAPPENED TO WANTING CHOICE????). They wanted more of us. They wanted hand-holding, a human, and high-touch. Not a single damn column in that ad platform said “hey Lauren, your audience is quietly begging for a different relationship with you,” but that's what was happening. We only found it because we went looking past the numbers. This week is about being NIMBLE (our current favorite word around here), about testing your messaging like a MAD SCIENTIST instead of clinging to it like a security blanket, and about the metrics that actually tell you something versus the ones that just make you feel a certain type of way. And a little treat: the paid audio companion (where Sarah and I get into the EXACT metrics we watch (and the vanity ones we ignore)) is usually behind the Substack paywall. This week it's on the house. Consider it a sample of the good stuff. Click here for it, and if you like what you hear… you know where to find next week's. |
|
-- This week's read time: 3-4ish mins For you skimmers: 2 mins (hit the bold headers and bullet points) |
|
Last week in DESSERT I name-dropped Bobbie, the infant formula brand, for publicly posting a code of ethics that basically said we're here to earn trust, not just grow a business. Well. This week they earned a full slice, because someone on their growth team wrote a LinkedIn post that put words to something we’ve also been shouting from the rooftops. Setting the stage: Airbnb exec has a 3 a.m. panic in the formula aisle → can't find a clean, organic option → quits her job to build one → launches Bobbie in 2018 as the first new formula brand in ~6 years → survives a humbling first-year recall → gets catapulted into the spotlight by the 2022 national formula shortage → becomes the first USDA Organic whole-milk formula made in the U.S. → now runs one of the most trust-forward marketing operations in a category where two companies (Abbott + Reckitt) control ~86% of the market. ( CNBC has the full lowdown.) Alina Fitsai, who works on growth at Bobbie, posted a breakdown of how they think about paid media, and the headline is that their paid media is “more than CPA & ROAS.” They zero in on education, awareness, and trust, run as one full-funnel system with the customer at the center. She sees growth marketers make the same mistake constantly, an ad angle or a message gets switched off because ROAS [return on ad spend] is low.” The brand message gets killed before it ever had a fair shot. Ah!! A trap!! She points out that nearly 70% of searches now happen without a click, and she's right; the freshest data (SparkToro, early 2026) puts U.S. zero-click Google searches (zero click Google search is where a user Googles something, gets the answer on the first page, and leaves without taking action) at about 68%, up from ~60% in 2024. INSANE. Which means a HUGE share of the value of your message; the impression, the “oh I've heard of them,” the slow build of trust… never shows up as a click. Your CPA (cost per acquisition) column literally cannot see it. If you only trust what the dashboard can measure, you'll systematically defund the exact brand-building that's working. Her other big reframe: it's less about the one perfect message and more about the diversity of your messaging angles. Validate that a message resonates (through other channels, not just ads), and if it does, don't kill it when the CPA dips!!! Move it to a different part of the funnel. Trust doesn't convert on the first impression; that's not a broken ad, that's just… how trust works. For your business: Before you kill a message because “the numbers are bad,” ask which numbers, and whether they can see what that message is actually doing. Now, the reason this hit me so hard is that it's exactly the lesson our own six months of ads taught us. We ran messaging that was VISCERAL and painfully specific (“your website is giving you less action than a library on Saturday night,” “your website should be employee of the month; instead it's showing up late with coffee stains on its shirt”). And that messaging worked, but not in the way the dashboard defined “worked.” The real signal was qualitative: what people said back to us, what they asked for, the fact that they wanted high-touch. The numbers told us WHAT was happening. They never once told us WHY. (a glimpse at that ad creative… if you were interested): |
|
For your business: Any metrics dashboard you have at your disposal is a smoke detector, not a diagnosis. It's great at telling you something's off. It is genuinely terrible at telling you what to do about it. Pair every number with a human signal… a reply, a DM, a sales call, a comment, or you'll optimize your way straight into the wrong strategy. So how do you actually TEST messaging? If you can scrape together even a small ad budget: this is the single fastest way to get real feedback, provided your targeting is tight. A few hundred dollars pointed at a well-defined audience will teach you more about which words land than months of guessing. You're not buying conversions, you're buying answers. (Watch engagement quality: saves, shares, replies, not just the vanity clicks.) If ads aren't in the cards right now (and that's completely valid) you can test organically. Here's how: - Split-test your email subject lines. Most email platforms let you A/B test a send to a slice of your list, then send the winner to everyone. It's the closest thing to a free lab you own.
- Run the same offer with different hooks across your organic posts and watch which framing gets saved and shared (not just liked). Likes happen on autopilot. Saves and shares don't.
- Use your own audience as the focus group. Story polls, “this or that,” a question in your caption or DMs. Ask which pain point actually keeps them up at night and let them hand you the language.
- Mine the words your audience already uses. Comments, reviews, competitor comment sections, Reddit threads, your own sales-call notes. Social listening is the most underused free research tool small brands have; the exact phrasing that makes someone go”"yes, THAT” is sitting in there right now.
- Two headlines, one landing page test. Even swapping the hero headline for two weeks and watching which converts is a real test.
The thread through all of it: organic IS the “other channel” Alina's talking about; the place you validate a message before you ever spend a dollar putting paid muscle behind it. 🎙️ FREE dessert, anyone?? the audio's free this week (before you go!). we get into why checking your data daily = stalking your ex, the lead magnet that got 1,500 emails and zero paid sales, and why 3 good leads beat 100 bad ones. → [listen free on Substack] |
|
How'd ya like this cake drop?? |
|
|
|
🪩 443 SW 2nd Ave Fort Lauderdale, Florida 33301, United States |
|
|
|