Across 56 LinkedIn Ads accounts totaling $9.4M in spend, audits found 32% of budget went to non-buyers. That's roughly $53,600 per account lighting up dashboards with impressions, clicks, and leads that never had a shot at becoming pipeline. On Google Ads, audits of 104 enterprise B2B SaaS accounts showed 34% waste ($26.5M of $78M), driven mostly by infrastructure gaps: broad match without negatives, Performance Max campaigns without offline conversion tracking.
Spending more didn't fix any of it. Spending better did.
The correlation problem nobody talks about
Here's the data that should change how you run your next QBR. A pipeline correlation study across B2B SaaS campaigns found CTR correlates with pipeline at r=0.09. CPL? Slightly better at r=0.23. Neither number would survive a stats class, let alone a board meeting.
Cost per SQL, on the other hand, correlates at r=0.71. ICP-fit score comes in at r=0.66. The metrics most teams optimize toward (clicks, cost per lead) are almost noise when it comes to predicting actual revenue. The metrics that matter (cost per SQL, ICP-fit) rarely show up in platform dashboards at all.
This isn't an academic distinction. Non-branded search CPC jumped 29% in 2025 while CTR dropped 26%. SaaS/tech CPC climbed 15–18% year over year. Every dollar of waste now costs more than it did last year. And if you're optimizing to CTR or CPL, you're scaling the wrong signal.
Why budgets keep growing while performance stays flat
Marketing budgets have dropped to 7.7% of company revenue, down from 11% in 2021 according to Gartner. The mandate from finance is clear: prove business outcomes, not spend levels. Yet the reflex persists. Pipeline stalls, so someone suggests increasing budget. CPL looks efficient, so the team scales that campaign. LinkedIn captures 41% of B2B ad budgets and shows aggregate ROAS of 121%, so it must be working.
Maybe. But those same LinkedIn accounts showed 32% waste in the audits. Aggregate ROAS masks account-level dysfunction. A channel can be structurally sound and still bleed money when targeting is wrong, audience suppression is absent, and closed-loop measurement doesn't exist.
As Andrew Goodman wrote in MarTech, advertisers who overspend early in pursuit of hypergrowth "often flame out and lose stakeholder buy-in." Jim Collins's "bullets before cannonballs" framework applies here: validate before you scale. Most campaigns aren't ready for a cannonball on day one. Algorithms are still learning, quality scores haven't matured, and you don't yet know which audiences or creative will perform.
The 90-day fix that outperformed a budget increase
One documented remediation program took Google Ads waste from 34% down to 11% in 90 days. The result: $17.9M recovered and cost per SQL reduced by 62%. The fixes weren't creative refreshes or bidding strategy changes. They were infrastructure: implementing offline conversion tracking, suppressing non-ICP audiences, and aligning platform signals with CRM outcomes.
That's the uncomfortable truth. The biggest performance lever for most B2B SaaS paid programs isn't more budget or better ads. It's measurement plumbing. Connecting ad exposure to CRM to SQL to closed-won revenue. Without that loop, you're optimizing in the dark and calling it strategy.
What to measure (and what not to over-interpret)
Platform metrics like CTR and CPL still have operational value for in-channel diagnostics. Treat them as inputs, not success criteria. The metrics that belong in your exec reporting:
- Cost per SQL (pipeline correlation r=0.71)
- ICP-fit score (pipeline correlation r=0.66)
- Blended ROAS tied to revenue, not platform-reported conversions
For measurement architecture, the expert consensus points toward triangulation: media mix modeling (MMM) for macro budget allocation, multi-touch attribution (MTA) for day-to-day optimization where signals are strong, and incrementality testing (holdouts) to prove causality. Not every team has the data maturity for all three on day one. Start with offline conversion tracking and CRM alignment. That alone closes the biggest gaps.
When this is wrong: If you're in a land-grab market with a short window (think a new category with two or three competitors and a clear ICP), aggressive early spend with looser controls can be the right trade-off. But even then, you need closed-loop measurement to know whether the land-grab is working or just expensive.
The real diagnostic
Before your next budget conversation, answer one question: can you trace a dollar of ad spend to a closed-won deal in your CRM? If the answer is no, more budget won't help. It'll just make the waste more expensive. The teams that win the 7.7%-of-revenue era aren't the ones spending the most. They're the ones who can prove what their spend actually produced.
