The situation
A B2B SaaS company selling a workflow automation platform into mid-market buyers had a healthy demand gen motion on paper. ACV in the mid-five figures, a predictable pipeline from paid, and a four-person marketing team that knew LinkedIn and Google well. About 60% of their paid budget went into LinkedIn Sponsored Content targeting director-and-above titles, with the rest split between Google Search and Performance Max.
The primary conversion event was a booked demo that passed SQL criteria. Everything upstream of that — impression, click, form fill, calendar book — was measured and reported weekly. The conversion rate from ad click to booked demo had settled at roughly 2.3% and had stayed there, give or take, for two full quarters.
At the same time, LinkedIn CPMs were drifting upward. At a certain point the math stops working: a flat conversion rate against a rising impression cost means pipeline gets more expensive every quarter. The team needed to find lift somewhere in the funnel, and ad creative was the most controllable place to look.
The problem
When the team audited their own creative output, a cluster of issues stood out:
- LinkedIn CPMs had been climbing 12–15% quarter over quarter with no sign of slowing.
- Click-to-demo conversion flat at 2.3% for two straight quarters — not bad in absolute terms, but not defensible as LinkedIn got more expensive.
- Headline testing was ad-hoc. Designers wrote 2–3 variants per creative; nobody was systematically evaluating which headline was most visually dominant on-screen.
- CTAs were getting lost. In LinkedIn Feed, the team's existing "Book a demo" buttons had color contrast ratios around 2.8:1 — below the attention threshold for reliable registration in a fast scroll.
- Demand gen and design were running in parallel, not together. The brief stage didn't specify visual hierarchy constraints, so creative decisions happened too late to fix cheaply.
The approach
Rather than rebuild the creative function, the team added objective scoring at two points: a retrospective audit of existing creatives, and a pre-launch gate on new ones.
Audited the existing headline and CTA library
Scored the last 25 LinkedIn and Google Display creatives in GazeIQ. Headline salience scores averaged 58 across the library — with clear outliers at 70+ that, when cross-referenced to reporting, consistently delivered higher CTR. CTA visibility averaged 54, with most of the failures driven by low contrast on the primary "Book a demo" button.
Introduced stronger-contrast CTAs as a design system rule
The team updated their ad-specific design tokens: primary CTA button background contrast ratio must be ≥ 4.5:1 against its surrounding creative area. That single change moved average CTA visibility score from 54 to 78 across the portfolio within two weeks of production.
Tightened headlines for salience, not cleverness
Headlines were rewritten to a stricter rubric: one concrete outcome, one proof signal, max 7 words visible. Salient nouns (dollar amounts, time savings, specific integrations) moved to positions 1–3 in the headline. Score threshold: 70 minimum before launch.
Ran pre-launch scoring for every LinkedIn and Google Display variant
Everything ran through GazeIQ before production sign-off. Sub-70 creatives went back to the designer with the specific fix (move CTA, enlarge headline, increase contrast). Nothing below 70 got production approval, and nothing below 75 was allowed to scale past the initial test budget.
The results
Click-to-demo booking rate (+61% relative)
Cost per sales-qualified demo
Pipeline added in the subsequent quarter
LinkedIn CPC on newly-scored creatives
Click-to-demo booking rate moved from 2.3% to 3.7% — a 61% relative lift — over the course of one quarter as older creatives rotated out and only 75+ scored creatives ran. Cost-per-demo dropped 29% despite the underlying CPM still drifting upward, because the creative was doing enough additional work to offset the impression cost.
The most visible downstream impact came from pipeline. The subsequent quarter added 41% more pipeline from paid, driven largely by a higher volume of qualified demos at a stable demo show-up rate. LinkedIn CPC also dropped 18% on newly-scored creatives — a second-order effect of platforms rewarding higher engagement with better placement.
We were writing clever headlines nobody could read in under two seconds. Once we scored for salience first and cleverness second, the pipeline impact showed up within six weeks.
— Head of Demand Gen, B2B SaaS workflow platform
Quote reconstructed from customer feedback. Individual and brand identities withheld at customer request.
Key takeaways
On expensive channels like LinkedIn, creative quality matters more per impression than it does on cheaper channels — so pre-testing pays back faster.
CTA contrast is one of the highest-leverage fixes for B2B ads. Most B2B creative underweights it because the rest of the design is muted and "premium."
Headline salience scoring forced the team to separate clever copy from scannable copy. Clever copy still matters — but not at the expense of being legible in 1.5 seconds.
Pipeline lift trailed CTR and booking-rate improvements by roughly one sales cycle. Measure patiently.
The shared scoring rubric collapsed the gap between demand gen and creative. Both teams pointed at the same number instead of arguing about aesthetics.