The Project THE X Blog

Field notes on running better Meta ads — without the agency markup.

Short, opinionated pieces on how lean media-buying teams actually win in 2026: what works in creative testing, how to read CPL signals without panicking, why lookalikes still earn their slot when you treat them as one signal among many, and what to do with the CRM data that's been sitting unused.

May 8 · 2026

AI-assisted media buying: what actually delivers ROI in 2026

Every Meta agency claims an "AI co-pilot" these days. We unpack which AI workflows are genuinely lifting performance (creative-variant generation, lead-quality scoring, micro-budget reallocation) versus which are dressed-up dashboards. With benchmarks from real EGP-denominated accounts.

AIWorkflowBenchmarks
May 9 · 2026

The new Meta lookalike playbook for thin-data accounts

When your source audience hovers around 1,000 hashed customers and Meta keeps rejecting lookalikes with #2654, what actually works? A field guide to per-country builds, ratio stacking, the "ladder" approach to scaling reach, and what to do when you literally can't get to 100 matched users per market.

AudiencesLookalikesMeta API
May 10 · 2026

Real-estate lead quality: from MQL fluff to compound-walking buyers

Why optimising Meta on raw lead count quietly destroys CPL economics in real estate — and the 4 lead-quality signals (call-back rate, site-visit booking, payment-plan inquiry depth, repeat name appearance) you should be feeding back into the campaign optimiser instead.

Real estateLead qualityCRM
May 11 · 2026

Creative testing that doesn't waste your week (or your client's money)

The 3-creative / 3-hypothesis matrix that gives you a statistically defensible read on what's working — and why the "test 12 variants at once" advice you read on LinkedIn doesn't survive a real budget. Includes the specific Meta delivery quirks that ruin amateur creative tests.

CreativeTestingStats
May 12 · 2026

Closing the loop: CRM feedback that actually moves CPL

Most accounts get Zoho or Sheets data into Meta as a custom audience and call it a day. The actual win is feeding back lead disposition — which leads turned into appointments, which got dispositioned as junk — so Meta's auction system optimises for quality, not volume. Here's the wiring.

CRMZohoOptimisation