If your ads feel more expensive in 2026, you're not imagining it. Meta's Q1 earnings just confirmed what we've been watching across the accounts we manage all quarter. Here's the breakdown of what's shifting, how we're handling it, and what realistic returns actually look like in this market.
Meta is selling more ad space at higher prices. Google Search CPCs climbed 5–9% year-over-year. AI Overviews, Performance Max, and Gemini ad placements are expanding inventory — but more inventory doesn't equal cheaper clicks. It means a more competitive auction, more often.
We're stress-testing every account we run against this new cost environment — auditing creative fatigue, audience overlap, and bid strategy. The fix for inflated CPCs is rarely "spend more." It's usually creative refresh and structural cleanup, which is built into our monthly scope.
Here's where most people get tripped up: they compare cold prospecting numbers to retargeting and assume something's broken. Nothing's broken. They're different jobs, and they should perform differently. Here's what's realistic across each.
The bottom line: Cross-industry median ROAS sits around 1.93x. If your cold campaigns are hitting 2x+ on prospecting alone, they're working. Anyone telling you cold ROAS should be 5x is selling you remarketing data with a cold label on it.
ROAS works for ecommerce because the revenue is immediate, attributable, and clean. When your "conversion" is a discovery session, a free trial signup, or a camp registration that closes weeks later, ROAS lies. It under-reports your real return by a wide margin.
The right metric for service businesses, programs, and franchise locations: Cost Per Lead, paired with what that lead is actually worth.
Our work lives at the top of the funnel, and our reporting reflects that. Here's what we actually track in your account:
What happens after a lead hits your CRM is your sales process. Lead-to-trial rate, trial-to-paid rate, and customer LTV are numbers we can't see from the ad account — but they decide whether the campaigns we're running are actually profitable for you.
Once we know your LTV, the math gets simple: your max profitable CPA is LTV × target margin. Until then, we're optimizing toward CPL — which works, but it's not the same as optimizing toward profit.
Retargeting only works if there are new people to retarget. Kill cold prospecting for 60 days and your warm pool collapses. Your retargeting audiences age out. Then your "good" campaigns stop working too — and most owners blame the algorithm.
If retargeting is hitting 8x ROAS, the cold campaign feeding it at 1.5x is doing its job. Fund the funnel, not just the finish line.
The 40/30/20 split assumes you're running the entire funnel through paid ads. Most of the accounts we manage aren't — and don't need to. The warm and hot tiers don't have to be retargeting campaigns on Meta. They can be (and usually should be) email and SMS.
For most of our clients, the funnel runs across three channels, not one: paid ads to bring people in, email to nurture them, SMS to close the loop. Each channel does the work it's best at — which is why your ad budget doesn't have to carry the whole funnel.
You're paying to fill a leaky bucket. The leads come in, the algorithm does its job — and then nothing happens to them. The biggest lift our new clients see is usually in the first 60 days, when we plug those gaps.
Spend varies widely across the accounts we manage — from multi-location operators running $200+/day down to single-location programs running closer to $20–30/day. The 40/30/20 split above works once daily spend clears roughly $50/day. Below that, the math breaks: there isn't enough volume to feed three campaigns, and Meta's algorithm needs roughly 50 conversions per ad set per week to optimize properly. At $20/day with a $40 CPL, an ad set gets 3–4 per week. That's the algorithm running blind, and it shows up as inconsistent results, rising costs, and creative that "stops working" for no clear reason.
How we run your account:
Realistic monthly output at $20/day:
At lower spend levels, the trade-off is real: this is steady, slow signal — not rapid testing. The right benchmark isn't a brand spending $200/day. It's the same account's previous month.
If you're not sure whether your campaigns are healthy, underperforming, or being judged by the wrong metric — book a call. We'll pull the numbers together and tell you the truth, whether you work with us or not.
Book a call to review →Sources: Meta Q1 2026 earnings, Triple Whale 2025 benchmarks, WordStream 2026 PPC data, Trendtrack ROAS analysis