sage

crossed 1,000 subscribers last week on an email newsletter that i've mostly automated with ai since day one you can just do things

lol @ everyone giving chatgpt shit for running ads when every single other ai chatbot will follow suit within the next two years

@copyimperialist ai is also making it 100x easier to create high quality landing pages

The ad's only job is to get the click. If you disagree, you're judging the wrong part of your funnel. Too many DTC brands mistakenly blame their ads when sales are low, complaining about creative fatigue or declining ROAS. But if your ad successfully got a qualified prospect to click, it did its job. The ad is just the opening line, not the whole conversation; the real problem is what happens after the click, on your landing page. Think about it: you spend $10,000 on ads and get a 2% CTR at a $1 CPC, which brings in 10,000 visitors. Your ad worked perfectly. But if your product page only converts at a dismal 0.5%, you'll only make 50 sales. At a $100 AOV, that's $5,000 in revenue on a $10,000 spend, resulting in a $5,000 loss. Was it the ad's fault? No, it was the landing page's. If that same page converted at a healthy 2%, you'd have 200 sales, generating $20,000 in revenue and a 2x ROAS. It's the same ad and the same traffic, but a radically different outcome. So stop blaming your ads. Start fixing your funnel.

@dtc_alchemist 'unlock your potential' imo, better to focus on specific outcomes instead people want to know what they'll gain, not just vague promises

Biggest ad scriptwriting sins: ❌“I’m obsessed” ❌“I’ve tried everything” ❌“Game changer” ❌“The best part?” ❌“You NEED this in your life” If your agency/strategist uses these they are extremely lazy and/or have serious skill issues Anything else I Missed?

@tomfgoodwin what % of people really care about privacy?

The big flaw with LLM ads is obvious: No user data → cheap, irrelevant ads Lots of user data → privacy freakout They can only balance trust vs relevance so well. End result: either unremarkable ads or destroyed trust in the product.

@sarthakgh as much as it pains me, i'd rather wait an extra year or two for severance s3 if it means the quality bar stays where it has been

Apple TV’s hit rate is insane Virtually every show is fantastic It is what HBO used to be and arguably even better given the low volume My only complaint is they take too long to release new seasons of really great shows

@garrytan taste for evals > data for evals?

AI just moves the excuses out of the way. You don’t get to say “I can’t edit” “I don’t know how to code” or “I’m not technical” Intelligence is officially too cheap to meter. Your taste is your evals. Your desire to win is your agency. Time to build. garryslist.org/posts/ai-didn-…

wonder how their nps is doing over the same time period?

@directRmarketer got it. thanks, that's helpful!

@laplacemonk Yes to see how far I am plus how mamy ads are in my date range vs hidden. Today i will finalize it with some upgrade ideas i had. Why it is important at large data sets you barely reach the end.. a well known problem

Upgraded Facebook Ad library hacking/scraper Chrome Extension, to sort by exact date range when it started running, sort by high impression and floating count of how many ads loaded of the total. And how much are hidden (out of date range) + percentage of total for the keyword

@raisperhar_ are you using any ai tools to help generate more ad creative, given the required volume is so much higher?

How to Get 10+ ROAS After the Meta Ads Andromeda Update? While most advertisers are panicking about poor Meta Ads results, I’m consistently getting 10–20 ROAS for multiple clients — and here’s exactly what you need to do if you want to see the same results. Forget whatever campaign structure you were using before September 2025. After the Andromeda update, there’s only one structure that can deliver maximum sales: ✅ 1 CBO Campaign ✅ 1 Ad Set ✅ Up to 50 diverse ads inside that ad set The key is diversity. Meta’s algorithm now thrives on variety — using just one ad type for your brand is old news. Your ad set should include at least 5 of the following creative types: ✅ Founder’s Story ✅ Us vs. Them ✅ Product Demo ✅ Customer Testimonials ✅ AI-Generated Ads ✅ Features & Benefits ✅ Face-to-Camera Videos ✅ Whitelisting Content ✅ Positioning Test ✅ UGC (User-Generated Content) And here’s a golden rule: If you use ad formats 1, 2, 4, 6, and 10, you’ll get guaranteed results — no matter your budget. For new ads, always launch a separate testing campaign first. Once an ad performs well in the test campaign, move it into your main scaling campaign (same ad set) and pause it in the testing one. It’s really that simple. If you understand this structure, you’ll start getting great results on Meta again — even after the Andromeda update.

@alex_momeni clean 😮‍💨

+5% conversion rate my team sautéed for Wellbel

@HR_starryeyes dang, sorry to hear that what was the original reason you were thinking of scaling down budget on the weekend?

Made a stupid mistake of reducing spends over weekend. Swear to god I’ll never do it again. It’s the stupidest idea out there. Very easy to say for someone who doesn’t understand abc of perf.

@Tochukwu2001 @nicktheriot_ do you use any tools for managing these processes? or do you just naturally remember when to do certain things?

@laplacemonk @nicktheriot_ That's where discipline and frameworks come in. Test creatives continuously, scale with rules, and align messaging with audience sophistication. This transforms ads from guesswork into a predictable growth engine.

The 10 commandments of “10k days”: 1.Testing 30+ creatives weekly. 2. 90% of hooks are emotional. 3. Retargeting ads handle objections. 4. Angles mapped to market sophistication levels. 5. Ad copy hits 3–5 buying triggers in the first 2 sentences. 6. Testing in CBO with broad targeting. 7. 3-day budget rules dictate scaling decisions. 8. Test new upsells to increase AOV 9. Offer structured around bundles, upsells, and LTV math. 10. Every creative gets an equal opportunity to spend and decrease CPA to scale Follow these and avoid the “numerical death” of your ROAS.

@bananaaafish @foreplay_co @pipiads @denote_net this is awesome what was your favorite of the three?

Want faster ad creative research? I compared @foreplay_co , @pipiads and @denote_net for ad inspiration workflows — here’s what I found Workflow tip: 1-use Foreplay to find top creatives 2-Pipiads to validate performance 3-Denote AI to generate new angles #Marketing

@shortgratis do you use any particular third-party ad spy tools? or just the native ad libaries?

Competitive intelligence is legal. Monitor your competitors' ads in the Ad Library.

@HustlesMedia do you use any particular tools to drill into user sessions?

Your ROAS isn't shrinking because ads got expensive. It's your funnel leaking cash. Stop blaming platforms. Drill into user sessions, eliminate checkout friction, and personalize post-purchase offers. That's how you turn clicks into profit. 📈

@Tochukwu2001 @nicktheriot_ anything work particularly well to make your ads practice more systemic?

@nicktheriot_ When you treat ads like a system instead of a gamble, you protect your ROAS and unlock growth.

@KanishDigital aren't they both important for profitability? because higher cac might be okay if the purchase is larger?

For first-time profitability, CAC matters more than ROAS ROAS gets inflated on single large orders, but CAC goes right into unit economics to instantly know whether this order would ever make money or not It's quite surprising CAC isn't talked about enough here

@Austen why do you like codex > claude code?

In the next 6 months a lot of them will switch to Codex

generational buy opportunity 80% of white collar workers will be replaced by ai agents aws will provide much of this compute