Stop optimizing for Cost Per Lead. Apple's lesson for Facebook ads.

Apple sells fewer laptops than HP. Apple’s average sale price is $1,100 to $1,300. HP’s is around $600. Apple prints more money per customer because the audience Apple targets has higher willingness to pay.

This is the entire game of Pillar 2.

The CPL trap

Cost Per Lead is the metric agencies report because it’s easy to game. Lower the qualification bar, run a softer pain hook, and CPL drops. The dashboard looks great. The client renews. The bank account doesn’t move.

Lower CPL almost always means lower lead quality. Which means lower close rates. Which means lower ROAS. Which means the campaigns that “look like they’re working” are quietly eroding your business.

Why pain selection determines lead quality

Every service you sell solves multiple pains. Each pain attracts a different quality of lead.

CPA serving business owners:

• Hook the “tax problems” pain → low CPL, low quality. These are foot-in-the-door buyers. They want $200 of work and then ghost.
• Hook the “lack of vision for retirement” pain → high CPL, high quality. These are strategic, long-term buyers worth $20,000+ over the relationship.

Software (form builder):

• Hook the “cost of subscriptions” pain → low CPL, cost-sensitive customers. They churn fast, complain about price, leave bad reviews.
• Hook the “modularity / multi-step workflows” pain → high CPL, premium customers. They stick, they expand, they refer.

Same product. Different pain. Different customer.

The right metric

Optimize for ROAS. Optimize for LTV. Optimize for closed-won-per-dollar-spent. Do not optimize for CPL.

When you switch pain selection toward premium-buyer pains, expect your CPL to go up. That’s correct. CPL going up while ROAS goes up is the right outcome.

How to actually pick pains

Three steps:

1. List every pain your service solves. Not just the obvious ones. Dig.
2. Score each pain on buyer quality. Which pains attract people with budget? Which attract bargain hunters?
3. Test premium pains. Run two campaigns. Same offer, different pain. Measure ROAS at 30 days, not CPL at day one.

Where this fits in the 5-Pillar Client Machine

Pillar 2 only compounds when you have Pillar 1: Lead qualification filtering out the wrong audience, Pillar 3: CRM connected to Facebook telling the algorithm which leads actually closed, and Pillar 5: The Operator keeping the whole system in sync month after month. Without those, even a great pain selection sends Facebook noisy signals and burns out fast.

If your last “Facebook agency” only touched creative and bids, this is probably the second thing they broke. The first was Pillar 3. The third was Pillar 4. The fourth was Pillar 1. The fifth (the Operator) they never even knew existed.

That’s the pattern. Every. Single. Time.