Look, there is nothing to blame Meta algo for. Yeah, performance is shitty in past weeks. But it's not your ads underperforming.
This exact same narrative I've seen all across X —"CPMs are up, ROAS is down, nothing changed on my end."
... and that's right, because nothing actually changed on your end, that's directly related to customer's wallet.
I mean, let's make a quick rewind of something you already know: on Feb 28, the US and Israel launched airstrikes on Iran. Iran closed the Strait of Hormuz. That waterway moves 20% of the world's oil supply. Within days, oil went from about $70 a barrel to over $110.
You saw gas prices climb, and so did your customer. Gas going from $3.20 to $4.50 costs a normal American household an extra $80-120/month. That money comes out of somewhere, and the first cut is the impulse buy. The "oh that's cool, let me grab it" purchase that your entire TikTok and Meta funnel is built around.
So, your ad didn't get worse, creative didn't stop working, audience didn't change, but their wallet? Their wallet did.
It goes deeper than gas: shipping costs are up, packaging materials cost more because plastics come from petrochemicals; grocery prices are climbing because 1/3 of world's fertilizer comes through that same strait.
So, your customer is paying more for everything and obviously gonna think twice before buying anything.
What do you actually do right now:
> Shift your product angle from impulse to justified.
Pure "cool gadget" purchases under $30 are getting hit hardest. Those are what people skip first. Products that solve a visible problem or save money still convert because the buyer can justify the purchase to themselves. "I want this" is losing to "I need this" right now.
> Raise your AOV instead of lowering prices.
Bundles, upsells, "complete kit" offers. A $45 bundle converts better than a $19 single product right now because the customer is already making fewer purchases. When they do buy, they want to feel like they got everything handled in one shot.
> Tighten your testing.
Every failed product test costs more when shipping is up and conversion rates are down. Don't test 10 products hoping 2 hit. Validate the market before you spend on ads. If the data says a niche is dead, believe it the first time. The margin for expensive lessons just got a lot thinner.
> Front-load your value prop in the first 2 seconds of your creative.
You've most likely done it already, if not — it's a 'must'. When people are spending cautiously, they scroll past anything that doesn't immediately answer "why should I spend money on this." Your hook needs to hit the problem or the outcome, not the product.
> Extend your attribution window and give campaigns more time.
People are still buying. They're just taking longer to decide. A purchase that used to happen in 24 hours might now take 3-4 days. If you're killing campaigns after 48 hours of bad data, you might be pulling the plug too early.
This isn't permanent. Oil dropped to ~$98 yesterday on news of possible negotiations. If the strait reopens soon, you'll see things normalize within weeks. If it won't open in following two months — we will see COMPLETE shift in buying behavior.
So don't count on it. I won't speculate on this regard whether it opens or not. Prepare for worst, hope for better. Right now, don't gut your ad account trying to fix a problem that lives outside your ad account.
@ecombatman Dashboard refresh gives you dopamine, notes give you pattern recognition. I started finding winners once I logged hook, offer and page change next to spend every day.
@conortrains Shopify will undercount that lift if people read email, leave, then come back through direct or branded search and still use the code. We learned more from holdout segments plus code redemption than Shopify attribution, first-order lift was bigger than the dashboard showed.
@Peter_Quadrel learned this on cold traffic, we got views and no carts when we copied polished ads from category leaders because we skipped the job of teaching. New brands need first-frame proof, one objection answered, and customer phrasing you pull from reviews or support inbox.
@ron_ecomm meta can run holdouts, they choose spend optimization because incrementality would tell half of us to buy less ads. After a point I get more lift from post purchase surveys, call reviews, and offer fixes than from another week staring at frequency charts.
@ecomrudolfs file it same day and send ad set IDs, budget change timestamps, and yesterday vs today CPA in the first message, cuts out 5 rounds with support. I’ve gotten credits more often on daily budgets, after cost cap or budget edits they love calling it normal delivery.
@flips4miles Big trap is checking one hero ASIN and calling the brand dead. I’ve gotten back into Nike and LEGO through odd child ASINs and stale bundles first, those pockets carry fat spreads since everyone else quit after the first denial.
@MrOverpaid Reps count more once a buyer can say no. I got sharper when I wrote product pages, emails, and ads against real objections, tweet volume built muscle, selling showed me where words leak cash.
@ByJEcom Daily profit will mess with your head at this stage. I’d watch 7 day contribution margin by country, one fat cart can make a weak funnel look healthy for a day.
@oliverkenyon Ran into this on oral care, Hismile’s bigger leak is choice overload on first purchase. New buyers need a fast “which one fits me” path across PAP+, v34, strips, bundles, because that confusion kills more intent than weak trust cues on a brand this known.
@DTCMidas Did this on one store, WISMO tickets dropped once we swapped fake 1 day copy for our real dispatch window. Missed promise creates refunds way faster than a 5 day ship time ever will.
@ArijanJanes Cart quantity upsells print when the price ladder matches PDP, same savings, same pack names, same ship promise. We saw CVR dip every time cart introduced new math, shoppers treat that like a second sales page.
@GadzhiIman I’ve watched info businesses hit a ceiling when all the trust lives in founder’s feed. The ones that reach real net worth turn that trust into owned customer base, recurring offers, and deal flow for buying boring cash machines outside the brand.
@ecomTrevor In this tier, buyers click fast and then sit on the page looking for proof, power draw, setup time, freight-to-door, damage handling. We got lift when we sold headache removal, same playbook plus stronger trust layer.
@EcomKostnchko_ In my tests, Claude gets closer when you feed it post-purchase survey answers and creator raw takes over polished “winning UGC.” Best review scripts steal exact buyer wording, then keep one friction detail in the read, that tiny hesitation makes the testimonial feel real.
@tristanrbl676 Mindset mattered for me in one spot, it kept me from killing a winner after two bad days. On US, I keep one promise, tone down claim language, and rebuild proof for each pocket, that scales better than flooding the BM with new concepts.
@FedotOff90 in Colombia, money shows up on rebill, PSE or Mercado Pago plus a WhatsApp reorder flow saves more dog-supplement subs than another ad angle once owners see the dog move better.
@ecomstreur also add your hotel and factory addresses in Chinese, as screenshots and inside Amap before you fly. I lost 90 minutes in Guangzhou with data on my phone because the driver couldn’t use the English address (and airport WiFi was useless)
@markdmei Big leak sits between receipt and delivery. In my stores, the winner was a plain text “what happens next” email with ship window, usage tips, and a reply-to address, it cut WISMO tickets hard and lifted 30-day repeat more than the founder note.
@oliverbrocato Brand search lags. I’ve cut Meta on stores where branded volume looked small and email plus repeat buyers still carried week one, search catches up after retention does.
@ecomtalent Retainers show up when that editor can read an ad account and turn ugly metrics into next week’s hooks. Copy and psych get you in the room, knowing why thumbstop held and CVR died keeps you on payroll.