Video summary

The BEST Facebook Ad Campaign Structure for 2026

Main summary

Key takeaways

Business

Updated Meta (Facebook/Instagram) Ad Campaign Structure: Core Shifts (2026)

1) Stop separating “cold” vs “warm” audiences by default

  • Old approach: Build separate ad sets/campaigns for cold vs warm (e.g., website visitors, video viewers, email list, IG followers) with different budget splits and messaging.
  • New approach: Rarely separate cold and warm, because Meta’s targeting system increasingly treats both as suggestions and expands delivery to reach likely converters.

Why it matters (mechanism):

  • In Meta’s UI, “Suggest an audience” effectively makes custom/warm audiences (website/email/followers) inputs, not hard limits.
  • If you create one ad set targeting “warm” and another targeting “cold,” you may still see similar conversion/impression splits, suggesting Meta is delivering to overlapping user sets anyway.

Action / setup

  • Define:
    • Engaged audience (interacted, but not yet customers)
    • Existing customers
  • Do this in Advertising Settings → Audience segments so Meta can report spend/results across:
    • New audience
    • Engaged audience
    • Existing customers

Example result snapshot mentioned

  • From 93 website purchases:
    • 9 from engaged audience
    • 59 from existing customers
    • 20 from new audience
    • Remaining: unknown/uncategorized (tracking limitations)

2) Consolidate: use fewer campaigns and ad sets (often 1 campaign + 1 ad set)

Primary reasons

  • More conversion data per unit → better optimization
    • Example logic: If you get ~70 conversions/week split across 10 ad sets, Meta optimizes poorly due to limited data per ad set.
    • If those conversions consolidate into 1 ad set, Meta can learn better:
      • best time/day
      • optimal frequency delivery windows
      • which users respond to which ads
      • overall to maximize the objective (sales/leads/etc.)
  • Reduce “auction overlap” complexity
    • Not about bidding against yourself.
    • It’s about Meta coordinating delivery cadence (e.g., 4–5 impressions to a user over 24–48 hours) being harder when multiple campaigns/ad sets fragment delivery.

Default recommendation

  • Combine cold + warm into the same campaign/ad set.
  • Use as few campaigns/ad sets as possible, unless a clear exception applies.

3) Test less at the targeting (ad set/campaign) level—except for location

Stop frequent tests like:

  • interest vs lookalike vs open targeting
  • separate ad sets for those variants

Rationale:

  • Those targeting types often fall under Meta’s “suggested audience” behavior (Meta expands beyond your intended targeting).

Exception: location-based targeting

  • Location behaves more like a hard control boundary, so Meta won’t expand beyond the location as freely.
  • Still run separate ad sets/campaigns for:
    • country/state comparisons (e.g., US vs UK vs Australia)
    • local multi-location businesses (different cities → separate performance tracking)

4) Run many creatives inside one ad set (creative diversity is the new “testing surface”)

  • Meta can handle far more ads per ad set than before.
  • New guidance:
    • Often 20+ ads in an ad set (no strict “upper limit” beyond quality and available resources).

Creative mix examples explicitly mentioned

  • formats: images + video
  • styles: UGC, founder-led, product demos, customer testimonials

Why this helps

  • Meta can personalize delivery so different people receive different creatives that match likely response.

5) Test ad copy variations within the same ad using built-in “variations”

Instead of creating separate ads for every headline/primary text test:

  • Use variation options:
    • Up to 5 primary text variations
    • Up to 5 headline variations
    • Up to 5 description variations

Key principle

  • Separate ads should focus more on creative types (UGC vs demo vs testimonial), not headline-level A/B tests.
  • Doing too much headline testing via separate ads creates “silly numbers of ads” and operational overload.

Funnel Strategy without Separate Funnel Campaigns

(Top/Mid/Bottom still works, but inside one ad set)

  • Old structure: Separate campaigns for top-of-funnel, middle-of-funnel, bottom-of-funnel.

  • New structure: Keep the marketing logic, but implement it as creative types within one consolidated campaign/ad set:

    • Top funnel: founder story / “why we exist”
    • Mid funnel: product demonstrations
    • Bottom funnel: customer testimonials + conversion CTAs

Automatic ad sequencing (possible)

  • Meta may show some ads that spend without conversions (top-of-funnel role),
  • while other ads convert later (retargeting/conversion role).

Operational warning

  • Don’t turn off “top funnel” ads just because they have weak direct conversion metrics—conversions may depend on them earlier in the journey.

When to NOT use the Default (Exceptions)

  1. Different product/service ranges → separate campaigns

    • Don’t split “red hats vs blue hats” (variations), but do split hats vs shoes (different optimization needs and convincing levels).
    • Also helps with operational control:
      • go out of stock / capacity control by turning off specific offers.

  2. General Dynamic Catalog campaigns

    • Common in retail/product-style businesses.
    • Can span multiple product ranges and lean more toward retargeting-heavy delivery.

  3. Separate offers that require different optimization and capacity rules

    • Example: a company offers:
      • Done-for-you services
      • Done-with-you mentorship/program
    • Put these in separate campaigns because:
      • target segments differ
      • budget behavior differs
      • capacity toggling differs (turn off/on)

  4. Omnipresent content strategy

    • Requires different structure (multiple ad sets).
    • Suited to high-ticket / involved decisions (e.g., $10k–$20k services).

  5. Creative testing becomes “stuck”

    • If you run 1 campaign / 1 ad set and Meta keeps spending on older winners while refusing to spend on new creatives (creative fatigue problem), you may need a different campaign structure to force/enable new-creative testing.

Metrics / KPI Focus Mentioned

Attribution examples

  • Example: £96,000 generated, but £58,000 not reported by Meta (tracking gap)
  • Purchase split example (website purchases):
    • 93 total purchases → 9 engaged, 59 existing customers, 20 new, remainder unknown/uncategorized

Implied KPI philosophy

  • Optimize toward true business outcomes (revenue/qualified actions), not just Meta-reported initial conversion signals—especially for recurring billing models.

Tracking / Attribution Recommendation (Business Execution Impact)

  • Tool/source mentioned: Hyros (positioned as best-in-class tracking and attribution for Meta ads).
  • Business argument:
    • Meta may show an initial transaction, but miss later recurring payments.
    • Hyros provides more complete downstream revenue visibility.
  • Example:
    • In an extreme scenario, Meta under-reports by £58,000—so relying only on Meta’s dashboard can lead to incorrect optimization decisions.

Presenters / Sources

  • Presenter: Not explicitly named in the subtitles (spoken first-person).
  • Sources/tools mentioned:
    • Meta (Facebook/Instagram Ads platform) changes and capabilities (including “Andromeda” and suggested audience behavior)
    • Hyros (tracking/attribution software)
    • Omnipresent content campaign strategy (referenced as the presenter’s methodology)

Original video