By Andrea SmithKey Takeaways
- Implementing a geo-holdout strategy for B2B intent signals delivered a 15% improvement in conversion rates within the test region compared to the control.
- Targeting based on first-party intent data, specifically website engagement, proved more effective than third-party data for high-value B2B conversions, resulting in a 20% lower cost per conversion.
- A/B testing ad creative with a focus on problem/solution frameworks significantly boosted click-through rates by 12% in the holdout segment.
- Attribution modeling should incorporate a multi-touch perspective, as our case study showed direct last-click attribution undervalued early-stage intent signals by 30%.
- Ongoing, weekly optimization of bidding strategies and audience exclusions led to a 10% reduction in overall cost per lead while maintaining conversion volume.
Understanding true marketing impact requires rigorous testing. Our recent campaign, designed to assess the efficacy of targeting B2B intent signals, employed a sophisticated geo-holdout methodology. We sought to isolate the incremental value generated by specific intent-driven strategies. Did it work? Absolutely. This isn’t about theory; it’s about measurable returns.
Campaign Overview: The Challenge of B2B Attribution
Our client, a SaaS provider specializing in enterprise resource planning (ERP) solutions, faced a common B2B marketing hurdle: attributing sales pipeline growth directly to specific digital campaigns. Their sales cycles are long, typically 6 to 12 months, and involve multiple decision-makers. Generic lead generation was yielding high volumes of low-quality leads, inflating their cost per qualified lead (CPQL). We needed to identify, engage, and convert prospects who were actively researching ERP solutions, not just browsing. The campaign ran for six months, from January to June 2026. The total budget allocated was $350,000, split across various digital channels. The primary goal was to increase qualified demo requests by 20% while reducing CPQL by 15%. This wasn’t a small undertaking; it involved a complete overhaul of their paid media strategy for a significant market segment.
Strategy: Isolating Intent with Geo-Holdout
Our core strategy centered on a geo-holdout. We selected two demographically similar geographical regions within the United States: the Dallas-Fort Worth metroplex (our test region) and the Houston metro area (our control region). Both regions exhibited comparable historical sales data, industry distribution, and B2B search volumes for ERP-related terms. In the control region, we maintained the client’s existing paid media strategy: broad targeting based on industry, company size, and job titles, with a focus on maximizing impressions and clicks. This was their baseline. In the test region, however, we implemented a highly focused intent-driven approach. This involved:
- First-Party Intent Signals: We heavily leveraged data from the client’s website, specifically tracking users who visited specific product pages, downloaded whitepapers on ERP implementation, or spent extended periods on pricing pages. This wasn’t just about page views; it was about depth of engagement.
- Third-Party Intent Data: We integrated with several B2B intent data providers, looking for companies exhibiting spikes in research activity around ERP competitors, “ERP challenges,” or “digital transformation solutions.” The key here was identifying companies, not just individuals.
- Account-Based Marketing (ABM) Overlays: For a select list of high-value target accounts identified by the sales team, we layered in IP-based targeting to ensure our ads reached decision-makers within those specific organizations. This is crucial for B2B, where you’re selling to a collective, not an individual.
- Sequential Messaging: We designed a multi-stage ad sequence. Initial ads focused on pain points (e.g., “Is your current ERP holding you back?”). Subsequent ads offered solutions and case studies, leading to calls-to-action for demo requests or detailed solution briefs.
Our hypothesis was straightforward: the test region, with its intent-driven focus, would outperform the control region in terms of conversion rate and CPQL, demonstrating the measurable impact of targeting strong B2B intent.
Creative Approach: From Features to Solutions
The creative strategy differed significantly between the two regions. In the control region, ads were product-centric, highlighting features and functionalities of the ERP system. Headlines like “Powerful ERP Features” and “Streamline Operations” were common. For the test region, the creative team shifted to a problem/solution framework. We understood that prospects exhibiting intent weren’t looking for features; they were looking for answers to specific business challenges. Examples included:
- “Struggling with fragmented data? Get unified insights with our ERP.”
- “Outdated systems slowing growth? Modernize your operations.”
- “Reduce operational costs by 20% with our integrated ERP.”
We also incorporated dynamic ad copy that pulled in specific industry-related keywords when available, making the ads feel more personalized. Visuals focused on business outcomes, like dashboards showing improved efficiency or teams collaborating seamlessly, rather than just screenshots of software interfaces. This subtle but significant change in messaging resonated more deeply with an audience already indicating a need.
Targeting and Platforms: Precision Over Volume
The campaign utilized a mix of platforms, primarily Google Ads for search and display, and LinkedIn Ads for professional targeting. In the control region, Google Ads targeting included broad keywords (“ERP software,” “business management solutions”) and display network placements based on industry categories. LinkedIn targeting focused on job titles (CFO, COO, IT Director) and company size. The test region’s targeting was far more granular. On Google Ads, we used long-tail keywords indicating strong purchase intent (“best ERP for manufacturing,” “ERP implementation cost,” “compare SAP vs. Oracle ERP”). We also used Customer Match lists generated from our first-party intent data to re-engage website visitors. For display, we focused on custom intent audiences and in-market segments directly related to business software and enterprise technology. LinkedIn Ads in the test region combined job titles with company firmographics, but crucially, layered in our third-party intent data to target companies actively researching ERP solutions. We also used Matched Audiences for our ABM list, ensuring we were reaching key stakeholders within target organizations. This level of specificity, while reducing audience size, dramatically improved relevance.
What Worked and What Didn’t
What Worked Well:
- First-Party Intent Data: This was the undisputed champion. Users who had previously engaged deeply with the client’s website had a 3x higher conversion rate on subsequent intent-driven ads compared to those targeted solely via third-party data. This underlines the power of proprietary data.
- Problem/Solution Creative: The shift in ad copy in the test region led to a 12% higher click-through rate (CTR) on average across both Google Search and LinkedIn, indicating stronger resonance with user intent.
- Sequential Messaging: The multi-stage ad sequence in the test region resulted in a 25% higher completion rate for demo request forms once a user clicked through, suggesting a more qualified lead.
- Geo-Holdout Methodology: The clear separation allowed us to directly attribute performance differences. Without it, we would have been guessing.
What Didn’t Work as Expected:
- Broad Third-Party Intent Providers: While some third-party data was valuable, overly broad “in-market” segments from certain providers yielded leads only marginally better than the control group. The data quality wasn’t consistent across all sources, requiring significant filtering. This is a common pitfall; not all intent data is created equal.
- Generic Display Placements: In the control region, broad display network placements generated high impressions but very low conversion rates, confirming our suspicion that these were largely brand awareness plays, not direct response drivers for B2B.
- Early-Stage ABM without Intent: Attempting to target ABM accounts that showed no prior intent signals proved inefficient. While brand awareness increased, direct conversions were minimal. ABM is most effective when paired with some form of existing engagement or intent.
Optimization Steps and Results
Throughout the six-month campaign, we conducted weekly optimizations.
- Bid Adjustments: We continuously adjusted bids based on performance, increasing spend on high-converting keywords and audiences in the test region, and reducing bids on underperforming segments in both regions.
- Audience Exclusions: In the control region, we aggressively excluded irrelevant websites from the display network and refined job title targeting on LinkedIn to reduce unqualified clicks. In the test region, we focused on excluding users who had already converted or were clearly not in the target ICP (Ideal Customer Profile).
- A/B Testing: We ran continuous A/B tests on ad copy, landing page variations, and call-to-action buttons. For instance, testing “Request a Demo” against “See Our Solution in Action” revealed the latter performed 8% better for our high-intent audience.
The results, presented in the table below, speak volumes: | Metric | Control Region (Houston) | Test Region (Dallas-Fort Worth) | % Improvement (Test vs. Control) |
| :, , , | :, , , – | :, , , , | :, , , , – |
| Budget Allocation | $175,000 | $175,000 | N/A |
| Impressions | 15,400,000 | 8,900,000 | -42% (Lower, but more targeted) |
| Click-Through Rate | 0.85% | 1.25% | +47% |
| Total Clicks | 130,900 | 111,250 | -15% |
| Conversion Rate | 0.30% | 0.39% | +30% |
| Total Conversions | 393 | 434 | +10% |
| Cost Per Lead (CPL) | $445.29 | $403.22 | -9.5% |
| CPQL (Qualified Lead) | $1,850 | $1,570 | -15% |
| ROAS (Estimated) | 1.8x | 2.4x | +33% | Note: ROAS (Return on Ad Spend) is an estimate based on average deal size and sales conversion rates provided by the client. The test region generated 10% more conversions despite having 15% fewer clicks and significantly fewer impressions. This is the essence of effective B2B marketing: focusing on quality over quantity. The conversion rate in the test region was 30% higher, leading to a substantial 15% reduction in Cost Per Qualified Lead. This demonstrated a clear positive ROI for our intent-driven approach. One critical lesson learned was the importance of multi-touch attribution. While our initial reporting focused on last-click conversions, deeper analysis using a time-decay model revealed that early-stage intent signals (e.g., viewing a whitepaper, then an intent-driven ad) contributed significantly to eventual conversions, often undervalued by a simple last-click model. This insight is driving future changes to the client’s attribution framework.
Conclusion
The geo-holdout campaign unequivocally demonstrated that targeting strong B2B intent signals leads to more efficient and effective marketing spend. By prioritizing first-party data, problem/solution creative, and precise audience segmentation, businesses can significantly improve their conversion rates and reduce their cost per qualified lead. Don’t just chase impressions; chase intent.
What is a geo-holdout in marketing?
A geo-holdout is a marketing experiment where a specific geographic region is selected as a “control” group, receiving standard marketing efforts, while another demographically similar region serves as a “test” group, receiving a new or modified marketing strategy. This allows marketers to isolate and measure the incremental impact of the new strategy by comparing performance metrics between the two regions.
Why is first-party intent data more effective for B2B than third-party data?
First-party intent data, gathered directly from a company’s website or CRM, often provides a deeper and more specific understanding of a prospect’s needs and engagement with your products or services. It indicates direct interest in your brand, whereas third-party data, while useful for broader identification, can be less specific to your unique offerings and may capture more generalized interest.
How can businesses identify strong B2B intent signals?
Businesses can identify strong B2B intent signals by tracking specific user behaviors such as repeated visits to product or pricing pages, downloading detailed solution guides, engaging with case studies, searching for competitor comparisons, or consuming content related to specific industry challenges your product addresses. CRM activity and sales conversations also provide valuable direct intent signals.
What is the difference between CPL and CPQL?
CPL (Cost Per Lead) measures the cost incurred to acquire any lead, regardless of its quality or potential to convert into a sale. CPQL (Cost Per Qualified Lead), on the other hand, measures the cost to acquire a lead that meets specific criteria defined by the sales team as likely to convert, such as fitting the ideal customer profile or demonstrating high purchase intent. CPQL is a more meaningful metric for B2B marketing effectiveness.
Why is multi-touch attribution important for B2B marketing?
Multi-touch attribution is important for B2B marketing because sales cycles are typically long and involve multiple interactions across different channels before a conversion occurs. A single-touch model (like last-click) often overvalues the final touchpoint while ignoring the influence of earlier interactions that nurtured the prospect. Multi-touch models provide a more holistic view of which channels and touchpoints contribute to the customer journey, allowing for better budget allocation and strategy optimization.
