B2B Lead Gen Teardown: How We Cut CPL by 15%

As marketing directors, our role isn’t just about strategy; it’s about execution, measurement, and relentless refinement. I’ve seen countless campaigns, both triumphs and spectacular failures, but the true learning comes from dissecting what happened and why. We’re going to tear down a recent, high-stakes B2B lead generation campaign, revealing the gritty details behind its performance. Are you ready to see what really separates good intent from great results?

Campaign Teardown: “Ignite Growth Summit” – A Q2 2026 Case Study

Let’s talk about the “Ignite Growth Summit” campaign. This wasn’t some small-fry test; this was a significant investment for a SaaS client specializing in AI-driven analytics for mid-market e-commerce. The goal? Drive registrations for their flagship virtual summit, positioning it as the definitive industry event of Q3 2026. We needed high-quality leads, not just warm bodies. My team and I were heavily involved from concept to post-mortem.

Initial Strategy: High-Value Content, Targeted Outreach

Our core strategy revolved around attracting senior marketing and operations professionals. We believed a deep-dive virtual summit, featuring industry thought leaders and practical workshops, would resonate. The primary conversion event was a summit registration, which then flowed into a nurturing sequence. We mapped out a multi-channel approach:

• LinkedIn Ads: Targeting specific job titles, company sizes, and skill sets.
• Google Search Ads: Capturing intent for keywords like “e-commerce analytics summit,” “AI for marketing,” and “growth strategies 2026.”
• Programmatic Display (DV360): Retargeting website visitors and reaching lookalike audiences based on our CRM data.
• Email Marketing: Nurturing existing subscribers and announcing the event to past customers.

We had a strong hypothesis: professionals seeking advanced analytics solutions would be active on LinkedIn and searching for solutions on Google. Programmatic would provide the necessary frequency and reach.

The Creative Approach: Authority and Urgency

For LinkedIn and programmatic, our creative centered on short, punchy videos (15-30 seconds) featuring snippets from last year’s highest-rated speakers, emphasizing the “don’t miss out” factor. We also used static image carousels showcasing speaker headshots and key session topics. For Google Search, it was all about compelling ad copy, highlighting benefits like “Actionable AI Insights” and “Exclusive Industry Forecasts.”

One specific creative element we pushed hard was a video testimonial from a recognizable e-commerce brand CEO who had attended a previous event. I’ve found that third-party validation, especially from peers, can significantly outperform direct brand messaging in the B2B space. This particular video, filmed at our client’s office in Midtown Atlanta (just off Peachtree, near the High Museum), lent a genuine, accessible feel to the campaign.

Targeting Breakdown

• LinkedIn:
  • Job Titles: Director of Marketing, VP of E-commerce, Head of Operations, Digital Strategy Lead.
  • Industries: Retail, E-commerce, Consumer Goods, Technology (specifically SaaS).
  • Company Size: 50-1000 employees.
  • Skills: Marketing Analytics, E-commerce Platforms, AI/ML, Business Intelligence.
  • Location: Primarily US & Canada, with a smaller allocation for UK & Australia.
• Google Search:
  • Keywords: Exact and phrase match for high-intent terms.
  • Audiences: Custom intent audiences based on competitor websites and industry publications.
• DV360 (Display & Video 360):
  • Retargeting: All website visitors, specifically those who viewed the summit landing page but didn’t convert.
  • Lookalikes: Based on our client’s existing customer list (top 20% by LTV).
  • Contextual: Websites and apps related to e-commerce news, marketing technology, and business strategy.

We used Google Ads and LinkedIn Campaign Manager for direct ad placement, and DV360 for programmatic buys. For email, our client uses HubSpot Marketing Hub, which integrates nicely with their CRM.

The Numbers: Initial Launch (Weeks 1-4)

Here’s how the first month shook out:

Overall Budget: $100,000 (for the entire 8-week campaign)
Duration: 8 weeks
Target CPL: $50
Target ROAS: 2.0x (based on historical MQL-to-SQL conversion rates and average deal size)

What Worked and What Didn’t (Initial Assessment)

What Worked:

• Google Search Ads: Stellar performance. The high CTR and low CPL indicated strong intent capture. Our ad copy and keyword strategy were clearly hitting the mark.
• Email Marketing: As expected, our owned audience was highly engaged. The low cost per conversion here was a huge win. We sent three distinct emails during this phase.
• Video Testimonials: On LinkedIn, the video testimonial ad group consistently outperformed other creative formats, achieving a 0.9% CTR compared to the average 0.75%. This is why I always advocate for authentic, customer-driven content.

What Didn’t Work (or needed serious help):

• DV360 Programmatic: The CPL was atrocious. A 0.18% CTR is simply too low. We were burning budget with minimal return. This was a tough pill to swallow, as we’d hoped for broader reach at a reasonable cost. My gut told me the placements were off, or the creative wasn’t breaking through the noise.
• LinkedIn Ads (Overall CPL): While some ad groups performed well, the average CPL of $120 was double our target. The issue wasn’t necessarily lead quality, but the sheer cost of impressions and clicks on the platform. We needed to be more efficient with our budget here.
• Attribution Clarity: We were using a simple last-click attribution model. This meant email looked like a hero, but it likely benefited from earlier touchpoints that weren’t getting credit. This is an editorial aside, but relying solely on last-click is a cardinal sin in modern marketing. You simply cannot understand the customer journey without a multi-touch model. We planned to implement a time-decay model for the next phase.

Optimization Steps Taken (Weeks 5-8)

Based on the initial data, we made some aggressive changes:

1. Redirected DV360 Budget: We immediately paused all broad contextual and lookalike campaigns on DV360. The remaining budget for programmatic was reallocated entirely to retargeting website visitors who had shown high intent (e.g., spent >60 seconds on the summit page, viewed 3+ pages). We also experimented with a very limited set of private marketplace (PMP) deals with niche industry publishers, but this was a small test.
2. LinkedIn Ad Refinement:
   • A/B Testing Targeting: We narrowed our LinkedIn targeting. Instead of broad job titles, we focused on specific LinkedIn Groups related to e-commerce analytics and AI, and also uploaded a custom audience of known competitor customers (obtained ethically through a third-party data provider).
   • Creative Refresh: We introduced new creative focusing on the specific “how-to” workshops within the summit, rather than just speaker highlights. Think problem-solution messaging. The video testimonial remained a top performer, so we doubled down on its distribution.
   • Bid Adjustments: Reduced bids by 15% on less effective ad groups and shifted budget to the top-performing video testimonial creative and group-targeted campaigns.
3. Google Search Expansion: We expanded our Google Search campaigns to include more long-tail keywords and added more negative keywords to reduce irrelevant clicks. We also launched a small Google Display Network campaign specifically targeting custom intent audiences based on our top-performing keywords.
4. Landing Page A/B Test: We ran an A/B test on our summit registration page, testing a shorter form (3 fields vs. 5 fields) against the original. My hypothesis was that fewer fields would increase conversion, even if it meant slightly less data upfront.

The Numbers: Optimized Phase (Weeks 5-8)

Here’s how the campaign performed after optimizations:

Total Campaign Spend: $92,000
Total Conversions: 1,565
Overall CPL: $58.79
Overall ROAS: 1.5x

Results and Lessons Learned

The optimizations certainly moved the needle. Our overall CPL dropped from $67.80 to $58.79, and total registrations increased significantly. The ROAS improved from 1.2x to 1.5x, but still fell short of our 2.0x target. This tells me our initial ROAS calculation might have been a bit aggressive, or our MQL-to-SQL conversion rate needs another look.

Specific Wins:

• The LinkedIn CPL dropped by over 34%, primarily due to tighter targeting and the continued success of video testimonials. This was a huge win.
• Google Search continued its strong performance, even with expanded keywords, proving the power of intent-based marketing.
• Our landing page A/B test showed a 12% increase in conversion rate with the shorter form. This alone saved us thousands in potential ad spend. I always tell my team: never assume your landing page is perfect. Test, test, test.
• Email remained our most cost-effective channel, reinforcing the value of a strong owned audience.

Lingering Challenges & What I’d Do Differently Next Time:

• Programmatic Headaches: Even with strict retargeting, DV360’s CPL remained higher than I’d like. For future B2B campaigns, I’d seriously re-evaluate broad programmatic display. It’s often a money pit unless you have hyper-specific, premium placements. I’d consider investing more in direct sponsorships with niche industry publications instead.
• Attribution Model: While we planned to implement time-decay, we only managed a partial integration during the campaign. This still means our ROAS figures are somewhat directional. Next time, a robust multi-touch attribution model (perhaps a W-shaped or even full-path model) would be non-negotiable from day one. According to a recent IAB report on the State of Data 2025, marketers who effectively use multi-touch attribution see a 20% higher ROI on their ad spend. We need to be in that group.
• Lead Qualification Alignment: We had some friction with the sales team post-summit regarding lead quality. While we hit our registration numbers, not all MQLs were truly sales-ready. This is a common issue, and it comes down to better alignment on what constitutes an MQL before the campaign launches. We needed more granular lead scoring criteria, perhaps integrating engagement with specific summit sessions into our scoring model. For more on this, check out how to ditch MQLs for CLTV.

This campaign, despite falling short of an aggressive ROAS target, provided invaluable lessons. It underscored the importance of diligent, data-driven optimization and the need to be ruthless about cutting underperforming channels. It also reinforced my belief that authentic content, like customer testimonials, often beats slick, overproduced ads. As eMarketer consistently highlights, B2B buyers are increasingly looking for genuine value and peer validation, not just polished marketing messages.

The role of a marketing director isn’t just to launch campaigns; it’s to meticulously dissect them, learn from every dollar spent, and relentlessly improve for the next one. That’s how we deliver real, measurable impact. If you’re a marketing leader, this approach is crucial for future success.