Meta ads
Account structure, Andromeda, testing, reading the data
Start with the playbook: performance-marketing — the rules that matter most, with links into these notes.
- Analysing a Meta creative test beyond "variant B won"Dara Denney's method for explaining why an ad won — read the campaign then the top spender, tag format, creator, imagery/setting and messaging, ask which awareness stage the message reaches (scalable vs low-hanging-fruit messaging), add breakdowns and comparisons — plus the recurring Motion reports she runs.
- Broad targeting vs interests, lookalikes and retargeting on Meta (2023 → 2026)How targeting advice flipped from Smart Marketer's 2023 70/15/15 interest/lookalike/broad split to 2025–2026 broad-by-default — and the remaining cases for interests (forcing a new avatar, lead gen), geo-only (local), customer exclusions, and Advantage+ as audience research.
- Building Meta creative from personas, angles and awareness levelsReady Set's "main vs testing dimensions" framework (creative target persona, angle, awareness level → visual hook, style, format, partnership ads, talent), Motion's persona → messaging → hook pyramid, and how to find untapped personas in your account.
- Choosing what Meta optimises for — "Meta is literal" (lead quality, value, funnel events)Because Meta delivers exactly the event you ask for, optimise for the lowest-funnel event with ~20–25+ conversions a week, use instant-form conditional logic and SMS-verified "higher intent" forms (or split thank-you pages) so only qualified leads count, switch sales campaigns to maximise value of conversions, and avoid add-to-cart/landing-page-view optimisation.
- Consolidate into an "Andromeda portfolio" or keep the old-school setup?Smart Marketer's rule for choosing between the consolidated Andromeda setup (reset when things are going badly) and the granular "Forrest Gump sprint" (when you have momentum), with the CleanEase case — 8 campaigns to 2, CPM −20%, CPA −35%.
- Creative similarity and "entity IDs" — why 30 ads can get one ad's worth of deliveryHow Andromeda clusters look-alike ads into one "entity" (industry term), which creative variables make an ad count as different (format, production style, audio, setting, persona), and how to tell similarity from fatigue.
- Diagnosing why a Meta ad isn't working (metric-pattern cheatsheet)Metric patterns that locate where an ad breaks — low thumbstop with OK CTR (fix the hook), high watch with low CTR (stronger CTA), clicks but short watch (shorten), high CTR with no conversions (landing page/offer), rising CPM/CPA on a winner (fatigue), no spend at all (similarity) — and how to turn each into a testable hypothesis.
- Five levels of Meta advertiser maturity (Renault → IKEA → New Balance → HelloFresh → Huel)Ben Heath's ladder for judging a brand's Meta ads from the Ad Library — level 1 repurposed TV/billboard ads with hashtags (Renault), level 2 good ads wasted on incongruent landing pages (IKEA), level 3 one creative style and weak CTAs (New Balance), level 4 high variety but one offer and one page (HelloFresh, 78 ads), level 5 offers and angles per avatar with 300+ ads (Huel).
- Fixing Meta underspend — CTC's four phases, and why they never turn off adsCommon Thread Collective's 4-phase answer when Meta won't spend at target efficiency — (1) creative expansion with big concept swings, (2) bid-surface expansion by duplicating evergreen campaigns into other bid types, (3) an "underspend" ASC for ads that got <3× target CPA in spend, (4) relaxing the ROAS/cost constraint only when the business justifies it — plus the "never turn off an ad" principle and how others disagree.
- Getting spend onto new Meta ads — push delivery vs the creative testing toolMeta's two ad-level ways to give new ads real spend without separate testing campaigns — "Push delivery to this ad" (default 20% of budget for 7 days, rolling out to ~half of Heath Media's accounts in Oct 2026) and the Creative testing tool (2–7 ads split-tested on segmented audiences) — when each fits by conversion volume, how much budget to test with (10–60%), and the short-term performance hit.
- How many ads per ad set, and the minimum budget for a Meta campaignPractitioner numbers for ads per ad set (10 to start, 15–20 diverse concepts in a consolidated ad set, 3–6 per concept test, ~20 for ASC) and minimum daily budgets (≥3× target CPA, ≥$100/day for ASC, $100–300 per concept test).
- How Meta ads strategy changes from $100 to $100,000 a monthBen Heath's side-by-side of a $100/month and a $100K/month Meta account — structure (one ad set vs multiple campaigns + 10–30% testing), targeting (hands-off vs audience segments + value rules), testing (one test a month vs a creative "factory" of e.g. 12 ads/week), bidding (most conversions vs value / CRM-fed), scaling steps, measurement and team.
- How to raise a Meta campaign's budget — automated rules, stepped increases and the scaling ceilingBen Heath's two scaling methods (automated rules, +3%/day when 7-day CPA is under target, −3% when over; or manual steps of 50→100→150…→1,000 with shrinking percentages every 5–10 days), Jon Loomer's pre-check (is it scalable among new audiences / 1-day click?) and ~15% steps, why increases cause temporary instability, finding the scaling ceiling — and why not to duplicate campaigns.
- How to structure a Meta ads account in 2025–2026The campaign layouts practitioners actually run — Dara Denney's testing campaign + Advantage+ (ASC) core, Smart Marketer's one-campaign-per-objective "Andromeda portfolio", CTC's evergreen (acquisition/retention ad sets filled to 150 ads) + marketing-moment campaigns, Ben Heath and Jon Loomer's one-campaign-one-ad-set start, Mars Men's 40 pillar campaigns — plus multiple products, countries, retargeting and the 2020 baseline.
- Is it really creative fatigue? Diagnosing a Meta ad's performance dropJon Loomer's case that most "creative fatigue" is really Meta shifting a new ad from remarketing to prospecting (prove it with audience-segment and attribution breakdowns), when fatigue is real (small/local or isolated remarketing audiences, too few assets, very high spend in niche markets per Ben Heath), and what to do instead of reflexively launching a new ad.
- Landing pages for Meta ads — congruency, listicles and sales pagesWhy Meta traffic converts better on a page that continues the ad's angle (listicles, advertorials, quizzes, nav-free sales pages) than on a PDP or homepage, with conversion-rate benchmarks and case results (+25%, +30%, $25K→$40K/day without new ads).
- Launching Meta ads for a new DTC brand — product, pillars, statics firstThe zero-to-one sequence from Zach Stuck (Mars Men, Hollow) and Dara Denney — pick a product/AOV that Meta economics can carry, "spiderweb" every problem the product can solve into pillars, test pillars with cheap statics in one purchase-optimised campaign, then add $100 creator videos and listicle landing pages.
- Meta audience segments — proving remarketing happens on its ownDefine engaged-audience and existing-customer segments in Meta's advertising settings, then use the audience-segment breakdown on sales campaigns to see how much budget already goes to warm people (often 10–30%+) — why separate remarketing campaigns are usually unnecessary and non-incremental, and how the breakdown underpins scaling and fatigue decisions.
- Meta bid strategy — minimum ROAS vs cost per result goal vs lowest costCTC's analysis of ~$1.5B Meta spend across ~320 accounts (Mar–Jun 2026) found minimum ROAS lands within a few percent of its target at portfolio level while the cost-per-result goal hit target less than half the time (on Google tROAS 1.3× target, tCPA 0.8×); why value bidding gives the algorithm more room, how CTC sets the target from margin and incrementality, budget liquidity, and the low-volume counter-advice to avoid manual bids.
- Meta creative testing setups at $5K, $50K and $1M+ per monthConcrete creative-testing configurations by spend level — one ASC with ~10 ads at lean budgets, an ABO testing campaign with one ad set per concept (3–6 ads, $100–200/day) at growth, and 10–20 tests/week at $200/day for ~7 days at scale — plus rules on test length and not mixing new ads with proven ones.
- Meta value rules — bid more for people worth more (instead of restricting targeting)How to use Meta's value rules (Advertising settings → value rules → rule set; bid +X% by age, gender, location, placement, device…) to feed information Meta can't see — repeat-purchase rate, lead-to-customer conversion, LTV — instead of hard age/gender/placement exclusions, and the common mistake of copying data Meta already has.
- Planning Meta creative volume, budget and teamWork backward from the financial model to a testing budget and a monthly number of new concepts (Zach Stuck), how Mars Men runs 5 agencies + internal pods with a pillar each, the 10×-retainer agency rule, first hire = copywriter-turned-creative-strategist, and why creative velocity beats raw volume.
- Scaling a Meta ad from signal to winnerHow practitioners decide something is a winner (spend velocity of $2K in 4–5 days, 2× the account benchmark held 3–5 days, internal ROAS above ~2) and how they scale it — budget steps, duplication into ASC or new ad sets, bid-strategy combos, new funnels and new avatars.
- Starting a new or low-volume Meta ad account — the simplest setup that can workWhat to do (and never do) with a brand-new or budget-limited Meta account — run in Ads Manager not boosted posts (iOS boosts add ~30% Apple fee), check purchase/lead events don't fire on page load, optimise for the purchase not add-to-cart or traffic, one campaign + one ad set + 1–2 ads, broad and unrestricted, leave it 1–2 weeks — plus Jon Loomer's fallback to a cheaper front-end product or leads.
- The breakdown effect — judge Meta results at the ad-set level, not per adWhy per-ad and per-asset comparisons mislead on Meta (historical ROAS doesn't predict future delivery; low-volume slices; day-1 delivery ≠ day-8), so judge ad sets or CBO campaigns in aggregate, launch new ads in batches of uniquely different angles every ~2 weeks, and don't chase the single winning headline or image.
- Weekly Meta media-buying cadence and attribution settingsLaunch new ads on one fixed day each week and judge after 7 days; use 7-day click for evergreen campaigns and 1-day click for flash sales; starter bid/attribution defaults; how teams run weekly pillar and hypothesis reviews; when third-party attribution tools are worth it.
- What actually moves Meta ads results in 2026 — the fundamentals checklistA consolidated checklist from Ben Heath ("97% of results"), Jon Loomer (foundational elements) and CTC (the "canon") — simplify and consolidate, make creative the targeting via diversification and creator partnership ads, accurate pixel + CAPI tracking, optimise for what you really want (value, qualified leads), value rules instead of restrictions, know and grow LTV, build brand, test with Meta's tools, read attribution properly.
- What Andromeda, GEM and Lattice are, and what they change for advertisersPlain-language model of Meta's 2025–2026 ad delivery stack (Andromeda retrieval → GEM ranking → Lattice unified models), the Advantage+ sales rollout of mid-2025, and the practical consequence that creative, not targeting or structure, decides delivery.
- When to iterate on a winning Meta ad vs make a net-new conceptDiversity (net-new concepts) explores new personas; iteration exploits a proven persona × angle combination — how the 2025–2026 practitioners split effort, what counts as each, and the cases where a small tweak is enough.
- Which Meta ad metrics to track, and how to read themA minimal Ads Manager column set split into primary metrics (spend, purchases, cost per purchase, ROAS — the only ones to optimise on) and storytelling metrics (frequency, CPM, unique outbound CTR, first-frame retention, hook/thumbstop, hold, completion, shares), with definitions and the thresholds practitioners use.