Meta Ads — Facebook plus Instagram plus Messenger plus WhatsApp — is still the largest performance advertising surface outside of Google. A well-run account is one of the highest-leverage marketing assets a mid-market business can have. A badly-run account is the fastest way to burn $30K in a month. Here's the current operator playbook.

Account structure in 2023

The old Meta playbook was surgical targeting and tight budgets. The new playbook is broad targeting and creative volume. The algorithm has improved fast enough that the optimal structure for most accounts is:

  • 1–3 prospecting campaigns with broad audiences (country-level, minimal interest stacking). Let the algorithm find the buyers.
  • 1 retargeting campaign for pixel events, video views, and email lists.
  • 1 testing campaign for creative iterations, graduating winners into prospecting.

Consolidation is the theme. Ten campaigns with fragmented learning beats three campaigns with concentrated signal — every time.

The measurement stack

iOS 14 and subsequent privacy changes gutted pixel-only tracking. If you're still running Meta Ads with only a browser-side pixel, you're flying half-blind. Minimum viable measurement stack in 2023:

  1. Meta Pixel (browser-side, still useful for cross-session attribution).
  2. Conversions API (CAPI) — server-side event tracking. Non-negotiable.
  3. Aggregated Event Measurement (AEM) properly configured with 8 prioritized events.
  4. Advanced Matching — email, phone, and external IDs sent with events for matching precision.
  5. UTM discipline — every ad URL tagged, measurement reconciled weekly against GA4 and your CRM.

Advantage+ — use it, don't fight it

Meta's Advantage+ placements, audiences, and budget allocation are no longer toys. On most accounts, Advantage+ Shopping Campaigns (ASC) outperform manually-structured campaigns by 15–30% in our tests. The instinct to control every placement and every audience is an old instinct. Let the algorithm do the allocation. Your job is to ship the creative.

Creative velocity

Creative is where accounts are won or lost. The best-performing accounts we run ship 15–30 new creative variants per month, kill losers within seven days, and graduate 1–3 winners per month into scale. Patterns that still work:

  • UGC-style video. Low-production, talking head or hands-on, 15–30 seconds.
  • Before / after transformations. Works for services, DTC, home improvement, fitness.
  • Founder-led video. Especially for B2B — a founder explaining the problem in their own voice converts.
  • Static ads that look native. Meme-format, screenshot-format, native-feed-aesthetic. Not over-designed brand graphics.
If your account is spending over $50K per month and you're shipping fewer than 15 new creatives per month, you're paying the agency to run the account, not grow it. — The velocity threshold

The incrementality check

Meta's in-platform ROAS is generous. It attributes conversions that would have happened anyway: organic buyers, repeat customers, brand-search traffic. The question that matters is incremental ROAS: what lift would you have seen without the ads?

Running an incrementality test is uncomfortable — it requires pausing ads in a test geography or holdout audience and measuring the delta. Most teams skip it. The ones that run it quarterly are the ones whose CFOs trust the marketing number. Start with a lightweight holdout: 10% of your audience, four weeks, measure revenue delta.

Budget allocation

Working budget pattern for a mature account:

  • 70% prospecting. Finding new buyers.
  • 20% retargeting. Closing warm audiences.
  • 10% testing. New creative, new audiences, new formats.

Under-invest in testing and you stop finding new winners. Over-invest in retargeting and you cannibalize organic conversions.

Operator read

The accounts that scale sustainably increase spend in 20% increments, weekly, against a CPA target — not 200% increments monthly when the boss asks. The algorithm needs time to adjust; aggressive scaling resets the learning phase.

What good looks like

Three indicators. In-platform CPA stable or declining over 90 days. Incremental ROAS above 1.5x (for most DTC) or above 3x (for most B2B). And creative velocity of 15+ variants per month. If all three are true, the program is healthy and scaling is a math problem, not a creative one.