09Meta Campaign Blueprint · Flagship Deliverable
The blueprint: structure, not budget
★ Included free in the Essential Audit
A ready-to-build Meta account architecture, produced by the First Steps paid media team as part of this audit — at no additional cost. Nordic Tigon is spending roughly SGD 65/day on Meta with no funnel architecture: cold-traffic purchase campaigns and profile-growth campaigns run side by side, with no remarketing layer to recover the intent it's already paying to generate.
The core problem
The result is a CPA of SGD 100–150 (some purchases near or above SGD 200) when the one well-built campaign proves SGD 80 is achievable. The fix is not more money — it's the right structure. With the same SGD 65/day, organized into a proper funnel, the CPA comes down and the wasted spend stops.
/The intent already being wasted
The account is generating real purchase intent and then letting it evaporate — none of it is being systematically brought back.
The single fastest win in the account
227 people started checkout and did not buy. 873 added to cart. None are being systematically brought back — and the traffic is already paid for. This is the cheapest revenue in the entire account.
/First — what an "objective" actually is
Before we touch structure, this one idea has to be clear, because it's the thing that quietly wastes the most money. Every Meta campaign asks you one question before it spends a cent: what should I go find? Whatever you answer, the algorithm spends your budget hunting for more of exactly that. Think of it like telling a pace-setter what pace to run — Meta will chase whatever number you give it, faithfully, whether or not that number wins the race. This single setting is called the optimization event, and it matters more than the budget, the audience, or even the creative.
The wording trap — this is the part everyone gets wrong
"Top of funnel" and "awareness" describe the CONTENT — the brand film, Moa's training clips, the story. They do NOT mean you set the campaign to a "traffic" or "awareness" objective. You can run pure brand-story creative and still tell Meta to optimize for Purchase — the creative and the objective are two separate dials. The confusion sounds like: "this is awareness content, so I'll pick the awareness/traffic goal." That's how budget ends up chasing clicks and video views that never become sales. (Note: where the Meta section earlier lists "video view rate" as a top-of-funnel KPI, that's a health-check on the creative — is the film holding attention? — not the campaign's optimization event. Different thing.)
/Two schools of thought — and why both are half-right
School A · "Feed the machine"
At low sales volume, optimize cold campaigns for a cheaper, more frequent event (Add to Cart) so Meta gets more data to learn from and you build retargeting pools faster.
Right about: at 2–3 sales a week, the algorithm is genuinely starving for signal on Purchase.
School B · "Just ask for buyers" (Hansell)
Keep Purchase as the event even on cold traffic — Meta finds buyers best when you point it at buyers, and optimizing for cheap events pulls low-quality traffic that never converts.
Right about: your one proven campaign is Purchase-optimized at SGD 80 CPA. Asking for clicks instead would make it worse.
The reconciliation — they're not opposites, they're two phases
Meta wants roughly 50 of your chosen event per week, per ad set, to "learn." Nordic Tigon generates 2–3 purchases a week for the whole account — so by that math no ad set will ever hit the target, which is School A's whole argument. But your proven campaign already converts on Purchase — so the real lesson isn't "stop optimizing for purchases," it's "stop spreading the budget so thin that nothing can learn." The genuine mistake would be launching the full three-campaign structure on SGD 65/day right now. So: don't break what works — but don't fragment it either. That's the phased plan below.
/The three building blocks
These are the three jobs a healthy account eventually does. At full scale you run all three as separate campaigns — but that's Phase B. Read them as the target, then see the lean version you actually run now.
01Prospecting — cold traffic
Optimization event: Purchase — keep it, it's your proven SGD 80 winner. Feeds the top of funnel and builds the retargeting pools.
- Audiences: lookalikes of purchasers, lookalikes of add-to-cart users, running/endurance interest stacks
- Segmentation: primary — women 34–44 (proven core buyer); secondary — men 55–64 (real response on select products). Don't over-fragment into tiny ad sets.
- Placements: the proven converters — FB Feed, IG Feed, IG Stories, FB & IG Reels.
- Creative: brand film + Moa's training-style content (the AI Video Bundle in production fills this).
- At scale (Phase B): add a second cold campaign optimized for Add to Cart to widen signal and grow the pools faster.
02Remarketing — warm audiences
Optimization event: Purchase. Recovers browsers and carters — the highest-efficiency spend.
- Segment by recency: 7-day, 14-day, 30-day windows
- Viewed product → "Still thinking about it?" + product benefit
- Added to cart → dynamic catalog ad showing the exact product + social proof
- Everyone → dynamic product catalog (DPA) for browsed/carted items
- Exclude: recent purchasers
03Cart & checkout recovery — highest intent
Optimization event: Purchase. Small audience, highest intent, best ROAS potential.
- Audience: initiated checkout, did not purchase (the 227) + add-to-cart abandoners
- Message: urgency + a closing incentive ONLY when margin allows (free-shipping threshold, limited-time discount, bundle). At 50%+ margins there's room to test a closing offer here.
- Exclude: recent purchasers
/The strategy: two phases, not three campaigns at once
NOWPhase A — consolidate (you're burning cash)
At SGD 65/day you can only afford to feed two engines properly. Two well-fed campaigns beat five starving ones — starving is exactly what produced the SGD 100–150 CPA. Run only:
- Prospecting — brand creative, Purchase-optimized (the proven winner)
- Retargeting — ONE campaign merging warm browsers + the 873 add-to-carts + 227 checkout abandoners, dynamic product ads + a closing offer, exclude buyers
- Kill profile-visit + geo-tests to fund it
LATERPhase B — expand at scale
Once you hit the trigger below, split into the full three campaigns and add the second cold campaign optimized for Add to Cart — your "feed the machine" idea, deployed as the scaling move, not the starting move.
- Prospecting (Purchase) + Prospecting (Add to Cart)
- Remarketing (its own campaign)
- Recovery (its own campaign)
Phase A budget — now, SGD 65/day
Phase B budget — at scale, ~SGD 100+/day
The trigger — when to move from Phase A to Phase B
Don't expand on a calendar — expand on evidence. Move to Phase B only when ALL of these hold true for 3–4 straight weeks:
① Spend sustained around SGD 100/day (your cap; currently at 65). ② Consolidated ROAS holding ≥ ~1.3x — so scaling multiplies profit, not loss. ③ Add-to-cart volume ~300+/month, enough to justify its own retargeting campaign.
Until all three are true, every extra campaign just re-splits a thin budget — the exact mistake that created the SGD 100–150 CPA in the first place.
/Audience & geography — probably where the budget bleeds
The structure above is how you spend; this is who you show it to. There are two very different meanings of "too broad," and only one of them is a problem — mixing them up is one of the most common and expensive mistakes.
Broad AUDIENCE — keep it
Few or no interest filters, inside one country. Modern Meta finds buyers better with a wide audience + strong creative than with narrow interest stacks. At 2–3 sales a week, narrow targeting starves the algorithm. Don't build dozens of tiny interest ad sets — that's fragmentation again.
Broad GEOGRAPHY — the actual leak
Many countries, or worldwide, on a tiny budget. It scatters SGD 65/day across wildly different ad prices and lets the algorithm chase the cheapest clicks, not the most profitable buyers. If the account is set to lots of countries, this is the first thing to fix.
The fix: concentrate the geography, keep the audience broad, and let the creative do the targeting. Women-focused creative naturally pulls the women 34–44 core; men 55–64 respond on select products — you don't need to hard-target them.
Which countries — in order, from her own data
The US is the tempting bet — and the wrong one right now
It feels like the big opportunity, but you already tested it: the New York campaign spent SGD 199 and returned zero. The US has the most expensive ad auction on earth — SGD 65/day buys almost nothing there, and the algorithm can't gather enough signal to learn. The 6 US orders are people finding you organically, not ads working. Revisit the US only when Singapore is reliably profitable and the budget is several times higher. Same logic kills random geo-testing (Sydney, New York already cost SGD 401 for nothing).
/Copy-paste audience library
Everything below is ready to build in Ads Manager. Build in this order: ① Custom Audiences first (they power both retargeting and lookalikes) → ② Lookalikes → ③ drop them into the ad sets. One caveat: type interest names into Meta's search box and pick whatever it actually surfaces — the exact list shifts over time and some smaller brands aren't targetable. And keep the golden rule: your primary prospecting ad set stays broad (no interests) — the interest stack is a single test ad set, not ten tiny ones.
① Custom audiences to create — from your own data
② Lookalike audiences — all Singapore
Lead with the Email-list and IG-engager lookalikes — they have the most data, so they're the strongest. The Purchasers lookalike is weak today (only ~114 buyers seed it) and gets better automatically as sales grow. Stack all four into one ad set and let Meta blend them.
③ Interest stack — for the one interest-test ad set
Running & endurance: Running · Marathon · Half marathon · Long-distance running · Trail running · Triathlon · Ironman Triathlon · Jogging · Physical fitness
Apps & communities: Strava · Nike Run Club · Garmin · Parkrun
Running & activewear brands: Nike · Adidas · ASICS · HOKA · On (running shoes) · Brooks Running · Salomon · Lululemon · Gymshark
Lifestyle: Athleisure · Sportswear · Standard Chartered Singapore Marathon
Put all of these into one broad interest ad set (Meta treats it as "anyone matching ANY of these"). Do NOT split them into separate ad sets — that's the fragmentation trap. Whatever Meta's search box doesn't offer, just skip.
④ The ad sets, ready to fill in — Phase A
Start with the Broad ad set alone. Add the Interest-test ad set only if Broad plateaus — and give each 3–5 days before judging. Layer the ② lookalikes into the Broad ad set once they're built.
⑤ Naming convention — so the account stays readable
Campaign: STAGE | Objective | Date
Ad set: Audience | Age-Gender
Ad: Creative | Format
Examples:
PROSPECT | Purchase | 2026-07 › Broad | 25-55-All › BrandFilm | Reel
RETARGET | Purchase | 2026-07 › ATC-90d | 25-55-All › DPA-CloseOffer | Catalog
/Turn off immediately
✕Profile-visit campaign
Spent ~SGD 385 for 233 followers, 179 site visits, and 1 purchase (~SGD 1.65/follower). Awareness, not a sales campaign — don't fund or judge it as one. Keep only a small controlled budget and feed those followers into remarketing later.
✕Dead geo-tests
Any campaign targeting a city outside Singapore right now (Sydney and New York already cost SGD 401 for nothing). One city at a time — Singapore only until it's profitable.
✕Low-intent placements
Inside purchase campaigns. Concentrate spend on the placements that actually convert.
The prerequisite — tracking
Before scaling any of this, the measurement foundation has to be sound, or every optimization decision runs on bad data. Meta Conversion API is implemented and Purchase match quality looks good, but Advanced Matching and automatic event tracking appear disabled, and there's no complete GA4 integration. Pixel/Shopify/CAPI deduplication must be verified to avoid double-counted conversions. This tracking setup is a separate scope of work that should be quoted (see Growth Opportunities → Services) — it is the foundation everything else depends on.
If you want this built for you
This blueprint is the "what." The Full Funnel Campaign Setup ($1,200) is the "we build it in your account" — Phase A live now, Phase B mapped for when you hit the trigger, all audiences, the dynamic catalog, and a walkthrough so Moa runs it herself. See Growth Opportunities → Next steps & services.
