Ahmed ElflalMost B2B operators in the Gulf treat Meta like a billboard.
They boost a post.
They pick a dozen interests.
Then they wonder why every lead is a tire-kicker.
The platform gets blamed. The platform is rarely the problem.
I have spent more than $10M in managed spend across the UAE, Saudi Arabia, and the wider GCC.
And the single biggest thing that changed how this works has a name: Andromeda. If your playbook is from before it, your playbook is wrong.
Andromeda changed the game: creative is the new targeting
Here is what happened.
For a decade, you picked the audience.
Interests, behaviors, lookalikes.
You controlled who saw the ad.
Andromeda, Meta's AI ad system that fully rolled out by late 2025, flipped that.
Now Meta picks the audience.
Your creative tells it who to pick.
Meta's own data team says creative quality drives roughly 56% of campaign performance now. More than targeting, budget, placement, and timing combined.
So the lever moved.
It is no longer "find the right audience."
It is "feed the machine the right creative, and let it find the audience for you."
Go broad, because targeting is Meta's job now
Stacking fifteen interests is not just unnecessary anymore.
It actively fights the algorithm.
Run broad. Let the creative and the offer signal who it is for.
Exclude your existing leads and customers.
Once you have conversion data, build lookalikes off your best closed deals, not page likes.
But broad is table stakes now. The real work moved to the creative.
Keep the structure simple, feed it a big creative library
The structure stayed simple. The creative count did not.
One campaign, budget set at the campaign level (CBO, or Advantage+).
One or two broad ad sets.
That part is the same as it always was, and Andromeda rewards it even more, because consolidating lets the algorithm learn in one place instead of spreading thin across silos.
Here is what changed: the old advice of "one or two creatives" is dead.
Andromeda wants a chunky creative library.
10 to 20 unique ads, live at once.
And not 20 versions of the same ad. That is the trap most people fall into.
Variation, not iteration
There is a difference, and it is the whole game now.
Iteration is changing the hook on the same concept.
Variation is changing the entire concept.
Andromeda rewards real variation. Different angles. Different tones. Different creative archetypes.
A founder talking to camera.
A text-on-screen problem callout.
A before-and-after dashboard.
A customer story.
A bold claim.
A contrarian take.
You want both, iteration and variation, but concept-level variation is what moves the needle. Give the machine genuinely different ideas and it finds pockets of buyers you could never have targeted by hand.
The offer is still 80% of the result
None of this saves a weak offer.
Before you touch the ad, you need an offer a real buyer wants. Not a cold "book a call." Something that solves a real sliver of the problem on its own: an audit, a checklist, a calculator.
The B2B Funnel Audit Checklist is a good example. Someone who wants to score their funnel cares about pipeline. The offer self-selects quality before the form asks a question.
If you want the full method, that is the 5Ts offer framework.
Instant forms vs landing pages
Instant forms are cheap and fast but lower intent, so add one or two qualifying questions.
Landing pages cost more per lead but raise intent, better for higher-ticket offers.
Low-ticket or high-volume goes to an instant form with qualifiers.
High-ticket goes to a landing page with a booking step.
More on that in how I keep leads qualified.
Scale with more accounts and more angles, not more complexity
When you pass $200 a day, you do not add complexity. You add accounts and angles.
Run several ad accounts in parallel, each with the same simple structure, each driving to the same funnel.
Two accounts at $100 a day each, testing different creative directions side by side.
This validates winners faster and builds redundancy, so one account flagged for a policy issue does not sink your pipeline.
The metrics that actually matter
Cost per lead is the most over-watched number in the Gulf.
A low CPL feels like winning. But if those leads never convert, you found a cheap way to waste money.
Judge against real benchmarks:
- CPL (opt-in): roughly $50 to $60, industry-dependent.
- Cost per booked appointment: around $100.
- Outbound CTR: 2% to 2.5% or higher. Under 1.5% is a red flag on the creative.
- CPM: $65 to $75 in a competitive industry.
- ROAS: 9 to 11x is exceptional; 4 to 5x is a realistic worst case for profitable scaling.
The golden rule: if some metrics look noisy but your cost per lead and cost per appointment are on target, do not touch it. Let it run. Decisions come from benchmarks, never from emotion.
The part everyone skips: speed-to-lead
You can do all of this perfectly and still lose, if you are slow.
A Meta lead is cold and shopping. Call within five minutes, not five hours.
Automate the instant touch, in the Gulf, WhatsApp is often your highest-leverage channel, then put a human on fast, then run a real follow-up sequence.
The ad bought you the conversation.
Whether it becomes pipeline is decided by how fast you show up after the click.