Study notes · Nabeal Khan, The Ecommerce Roadmap · Meta module, lessons 34–43 (+24)
Give Meta what it wants, then get out of its way.
His whole Meta method on one page: how he launches, how he reads day one, how he scales one campaign to about $3k a day, how cost caps work, and the account structure he uses past $50k a day. Each point is restated in plain English, with the lesson it comes from.
It sees billions of people and every brand's data. Yours is a rounding error. He says the real wins happen outside the ad account: product, offer, avatars, angles, creatives and funnels. L35
If Meta spends on an ad, it's a good ad. If it won't spend, something about the ad is wrong. He's seen forced spend work about one time in a hundred. L35
Total revenue ÷ total media spend, across every channel. Platform ROAS figures fight over credit, so he watches MER against his breakeven. L34
Spend cannot be restricted and spend cannot be forced.Nabeal, lesson 41
Notes, not a transcript: everything here is his method restated in my words, with no outside advice added. Numbers are the ones he gives. Where he calls something a rough rule, it's marked that way.
Drag through the stages
The spend ladder
Slide your daily ad spend. The panel shows how his ad account is set up at that level, which channels he has switched on, and what he does next.
The ad account
Channels switched on
What he does at this stage
Run your own numbers
His rules as calculators
Each tool applies his thresholds to your numbers. They're his rules of thumb, so treat the output the way he does: a signal to look closer, not a law.
Defaults are his lesson 34 example: a $50 order with $16 of costs, and $30k revenue on $18k spend.
Breakeven MER = 1 ÷ margin, where margin = (order value − product − fulfilment − fees) ÷ order value. He calls a breakeven of 1.4–1.55 good. The +25% target is something to grow into over 3–5 months, and 2+ is where he says you're really making money. L34
Read these only after the campaign has run a full day. L37, L40
This is for highest-volume campaigns. On cost caps, use the cost cap tab. L40
Switch the same campaign: bid strategy "cost per result goal" at campaign level, then set the amount on the ad set. No need to duplicate. L40
Lesson 37 · launch protocol
Launch day setup
The exact settings he walks through in Ads Manager. He says the buttons move around every week; what matters is the structure, the ads, the landing page and the offer. Tick them off as you go (saved in this browser only).
Lessons 35, 36, 38, 41
The ecosystem rules
His core idea: Meta is an ecosystem, not a set of isolated tests. People see about a dozen of your ads before buying, so every ad that gets spend is playing some role.
Lessons 35 and 24
The five metrics he watches
He keeps the column set small. Soft metrics are only for understanding why an ad wins or loses so the next one is better, never for deciding where spend goes.
Lesson 42 · scaling ads
The golden formula
He takes this from Meta itself: how much an ad spends comes from three things multiplied. Scale horizontally (more landing-page campaigns) and vertically (each one spending as much as it can).
On highest volume you only control the budget, and the bid moves around with everything else.
1 · Raise the bid you can afford
Higher order value
- Price test in $1 steps ($50, $51, $52…) until conversion rate or revenue per session breaks. He doesn't like split-test apps for this.
- Add free gifts to justify a higher price, as long as they fit the avatar.
- Test a 3-product bundle against "2 plus a free gift" in its own campaign.
- Post-purchase upsells. His example: a turmeric hand cream for people who type all day.
Higher LTV, lower churn
- Teach long-term use inside the funnel before they buy.
- Better customer experience: packaging, design, even a physical book instead of an ebook.
- Milestone gifts for subscribers: extra bottle at month 3, mystery gift at 6, a voucher at 12.
- Follow-up ads from the creator already carrying spend, plus long education ads.
- Bottom-of-funnel ads on subscription perks, and a flood of testimonials.
Better margin
- Negotiate suppliers, ideally with Chinese-speaking help, and visit China yourself.
- Cheaper, faster shipping through volume.
- Trim waste, but he'd rather grow revenue than cut costs.
2 · Raise the estimated action rate
Signals Meta reads
- Past conversion rates: click-through, add to cart, checkout, purchase.
- Hook strength and watch time.
- Landing page engagement: bounce, time on site, page speed.
- Fatigue, which he thinks barely happens if you keep feeding new ads in.
- Likes, shares, saves and positive comments; fewer hides and reports.
- Purchase density: higher spend gives steadier results.
What he does about it
- Novel ads that feel native. Never rip other brands' ads. His $5-CPM US ad takes minutes before you realise it's an ad.
- Study trending organic content for ideas.
- Send cold traffic to a pre-sell page. People read and scroll, which Meta rewards, and they arrive more convinced.
- His example of the idea: a product page that tells a long story with an enemy before the price appears far down.
3 · Raise ad quality
The questions he asks
- Would someone watch this if it weren't an ad?
- Would they send it to a friend? Would it do well organically?
- Does it tell a story with a hero and an enemy?
- Does it teach before it sells? His top ad gives five minutes of value before the product.
- Will the comments say things like "this is exactly what I needed"?
- Does it teach something, back up what they believe, stir an emotion or solve a problem?
Lessons 36, 37 and 24
Ad volume and intent
After Andromeda, Meta only lets genuinely different ads into the auction and matches each one to a small slice of people. So the job is lots of different, high-intent ads, and the creative does the targeting.
How many ads
L36There's no magic number, but more high-intent ads means more chances. A friend who launched with 5 ads failed; one with 300 succeeded. When asked how to reach $100k a day, his answer was about 5,000 ads. Spend lags behind ad count, then catches up.
Launching
L37About a week to build store, research and ads. Launch with as many high-intent ads as is reasonable, at least 5 per awareness stage. Statics alone are fine, though 1–3 videos help. Get feedback loops going once 40–50 ads are live.
Go niche, not broad
L36Avatars split into personas (belief systems), and each persona gets dozens of ads. Every ad opens a new pocket of people. Adding a high-intent ad should never hurt the account; if it does, the testing setup is wrong.
Quality vs quantity
L36A realistic hit rate is 20–30%, and he cares less about it than about volume of high-intent ads. Roughly an hour per ad, not ten. At least 90% of your ads should be high intent.
Big swings, not hooks
L36, L24Test angles and personas instead of 100 hooks on one ad, since Meta already shows ads to relevant people. Iterating now means same winning message, new concept: about 70% winning angles, 30% new ones. Example: one angle as UGC, then a street interview, an animation and a 5-minute science video.
Where winners come from
L24Iterate on what customers say, not on your ads. Reddit, reviews, post-purchase surveys, YouTube and TikTok comments, and competitors' comments and reviews. Every line should make them want to read the next.
Is this ad high intent?
His checklist for every ad (L36). The best statics answer all of it in a few words: a stranger should be able to tell you who it's for, what it does and why it's different.
Lesson 39 · "whitelisting" from day one
Run ads through third-party pages
Strictly it isn't whitelisting (that's using a creator's handle). It's running ads from niche-named Facebook pages instead of the brand page, because nobody trusts a brand name they've never heard of.
Why
A page named for the niche gets past people's guard. You get higher click-through and lower CPMs than the brand page, which means a better experience for the viewer and a cheaper CAC. Big brands run many such pages; search the brand as a term in the Ad Library to see them all.
Which page for which ads
Third-party pages carry most of the top-of-funnel spend. The brand page is mainly for middle and bottom of funnel. Once you're big and want to be remembered, use the brand page more.
How to set one up
- Brainstorm niche names with Claude (for his magnesium brand: Dialed In Daily, The Operator's Edge, Wired But Tired).
- Business Manager → create Page, category Magazine, bio from Claude.
- Connect it to the ad account, switch into the page, add a logo and cover.
- Post on it. Warm it up with a few posts; once a page carries traffic, a VA posts daily.
Run several
His team runs dozens of pages per brand, so the same person meets several "outlets" all pointing to the brand. A page outlives a failed launch: the people who saw it on one test are already familiar with it on the next. They even share pages across brands.
Lesson 34 · the numbers game
Launch timeline with subscriptions
His brands make their profit over the customer's lifetime, so losing money on the first order is normal. If ROAS is too high, you're probably not spending enough.
Validate demand, build cash and momentum. A subscription option lowers conversion on a fragile launch. Bundle offers help build cash.
Turn subscriptions on: now you've launched. Running 10–15% under breakeven MER is OK if you have cash, because rebills arrive next month.
Meta ROAS can sit around breakeven because rebills fund it. Keep spend steady and test as many ads as you can.
Work out 90-day LTV against CAC; he aims for 1:3. Give Claude your subscription dashboard, spend and costs as a stand-in CFO, then double-check it. Keep Meta ROAS above 1 while the brand is under 6 months old.
With six months of retention data you know your 6-month LTV, so you can scale aggressively. If you're profitable on the first order with subscriptions, he says scale as fast as you can.
Growth pace
L34The tech-company "40 rule": revenue growth % plus profit growth % should add up to about 40 points a month (20 + 20, or 30 + 10). Seasonal spikes are the exception.
Tools he names
L34, L35Loop for real brands (cohorts, cancellation flows), Kaching for dropshipping stores. Purchaser exclusions at ad-set level (longest window) so rebills don't flatter CPA. Server-side tracking from about $2–3k a day.