Marketing Expert Analysis: 2026 ROAS Gains

In the dynamic realm of digital marketing, understanding what truly drives success requires meticulous expert analysis of campaigns. We’re not talking about glancing at a dashboard, but dissecting every layer, from initial concept to final conversion. How often do marketers genuinely understand the ‘why’ behind their results?

I’ve spent over a decade in this field, and one consistent truth I’ve observed is that surface-level reporting rarely tells the whole story. You can have fantastic impressions, but if your conversions are flat, what good is it? True expert analysis involves digging deep, identifying the levers, and understanding how they impact the entire customer journey. Let me walk you through a recent campaign teardown that illustrates this point vividly. We’ll call it the “Project Ascend” campaign.

Project Ascend: A B2B SaaS Lead Generation Campaign Teardown

Last year, my agency, Digital Vantage, partnered with a burgeoning B2B SaaS company specializing in AI-powered data analytics. Their goal was ambitious: generate high-quality leads for their enterprise solution within a six-month window. They needed to establish market presence and fill their sales pipeline. This wasn’t a small ask; the average contract value was substantial, meaning the cost per lead (CPL) could tolerate being higher than a typical B2C campaign, but efficiency was still paramount.

Strategy and Objectives

Our primary objective was to generate Marketing Qualified Leads (MQLs) who fit a very specific ICP (Ideal Customer Profile): decision-makers in Fortune 1000 companies within the finance and healthcare sectors. We aimed for 500 MQLs over six months with a target CPL of $300 and a minimum ROAS of 2:1 (Return on Ad Spend), factoring in a conservative sales cycle and conversion rate. The total budget allocated was $150,000.

Our strategy revolved around a multi-channel approach:

1. LinkedIn Ads: For precise professional targeting.
2. Google Search Ads: Capturing high-intent users searching for solutions.
3. Programmatic Display (DV360): For brand awareness and retargeting.

The core content offer was an exclusive whitepaper titled “The Future of Data Analytics in Enterprise,” backed by a series of webinars and case studies. We used HubSpot for CRM and marketing automation, which was crucial for lead nurturing and tracking.

Creative Approach: The Power of Specificity

For LinkedIn, our creative focused on problem/solution framing, highlighting specific pain points faced by data teams in large organizations. For example, one top-performing ad headline was “Struggling with Data Silos? See How AI Unifies Your Enterprise Analytics.” The visuals were clean, professional, and featured abstract data visualizations rather than stock photos of smiling businesspeople. We ran 10 different ad variations on LinkedIn to test different value propositions and calls to action (CTAs).

Google Search Ads were straightforward: highly relevant ad copy matching search queries like “AI data analytics for finance” or “enterprise data intelligence platforms.” We ensured our landing pages were meticulously optimized for these keywords, with clear forms and direct value propositions.

Programmatic display ads used a mix of static and HTML5 banner ads, primarily for retargeting website visitors who hadn’t converted. The messaging here was softer, focusing on thought leadership and reinforcing trust.

Targeting: Precision Over Volume

This is where we really leaned into the “expert” part of expert analysis. For LinkedIn, we used job title targeting (e.g., “Head of Data,” “Chief Analytics Officer”), industry (Finance, Healthcare), company size (5000+ employees), and seniority level (Director+). We also uploaded a list of target companies for account-based marketing (ABM) efforts.

For Google Search, we relied on exact and phrase match keywords, carefully excluding irrelevant terms. We employed audience targeting within Google Ads, focusing on “in-market” audiences for business software and data solutions.

DV360 allowed us to build custom intent audiences based on competitor website visits and specific content consumption patterns, further refining our retargeting segments.

Initial Performance (Months 1-2)

The initial two months were a learning curve, as they always are. Our budget allocation was roughly 60% LinkedIn, 30% Google Search, and 10% Programmatic.

What worked: Google Search Ads immediately delivered high-quality leads at a respectable CPL, clearly demonstrating strong intent. Our ad copy and landing page relevancy were spot on. The whitepaper was a strong offer. I’ve found that when you perfectly align search intent with a valuable resource, magic happens.

What didn’t work: LinkedIn’s CPL was far too high, significantly above our target. While the leads were generally high quality, the volume wasn’t justifying the cost. Programmatic also struggled with CPL, though its role was more awareness-driven. The LinkedIn creative, despite multiple variations, wasn’t resonating enough to drive down costs.

Optimization Steps (Months 3-6)

This is where the real work of expert analysis shines. We couldn’t just keep throwing money at underperforming channels. Here’s what we did:

1. LinkedIn Overhaul:
   • Creative Refresh: We shifted from problem/solution to direct “results-oriented” messaging, incorporating testimonials and specific ROI figures from early adopters. For example, “Cut Data Processing Time by 40% with Our AI Platform.”
   • Audience Refinement: We narrowed our target industries further, focusing only on companies with a known history of adopting new technologies. We also experimented with LinkedIn’s “Lookalike Audiences” based on our existing customer list, which proved incredibly effective. According to a LinkedIn Business report, lookalike audiences can increase conversion rates by up to 20%.
   • Offer Diversification: Instead of just the whitepaper, we introduced short, digestible video case studies as lead magnets for specific industry verticals.
   • Bid Strategy Adjustment: Switched from automated bidding to manual bidding with a focus on maximizing conversions within a set CPL.
2. Google Search Expansion:
   • Negative Keywords: Continuously added negative keywords to eliminate irrelevant clicks.
   • Ad Group Segmentation: Broke down broad ad groups into hyper-specific ones (e.g., “AI for financial risk analytics” vs. “AI for healthcare data security”) to improve ad relevance and quality scores.
   • Landing Page A/B Testing: Tested different hero images, CTA button colors, and form lengths. We found that a shorter form (3 fields vs. 5) increased conversion rates by 15%, though it sometimes led to slightly lower quality leads that required more nurturing. It’s a trade-off, and you have to decide which side of the fence you want to be on.
3. Programmatic Retargeting Optimization:
   • Frequency Capping: Reduced impression frequency to prevent ad fatigue, setting it at 3 impressions per user per day.
   • Dynamic Creative: Implemented dynamic creative optimization (DCO) to show different ad variations based on user behavior and stage in the funnel.
   • Exclusion Lists: Ensured converted users were immediately removed from retargeting pools.

Final Performance (Months 3-6)

The optimizations paid off significantly:

Overall Campaign Metrics (Total 6 Months):

• Total Impressions: 10,450,000
• Total Clicks: 139,500
• Overall CTR: 1.33%
• Total Conversions (MQLs): 700 (exceeding our 500 target!)
• Total Spend: $150,000
• Average CPL: $214 (well below our $300 target!)
• ROAS: Based on a 3% MQL to customer conversion rate and average contract value, we projected a ROAS of 3.5:1, significantly exceeding our 2:1 goal. That’s the kind of number that makes clients very happy.

Key Learnings and Editorial Aside

The most significant learning was the power of relentless iteration and the willingness to pivot. We could have stuck with the initial LinkedIn strategy, rationalizing its high CPL by arguing for “brand building,” but that would have been a disservice. Sometimes, you just have to admit something isn’t working and fix it. My advice? Never fall in love with your first idea; the data rarely does.

Another crucial insight was the synergistic effect of channels. While Google Search delivered the lowest CPL, LinkedIn and Programmatic played vital roles in creating awareness and nurturing leads higher up the funnel. Many of the Google Search converters had likely been exposed to our brand on LinkedIn or via a programmatic ad first. Without a robust attribution model, it’s easy to undervalue these touchpoints. I am a firm believer that multi-touch attribution is no longer optional; it’s essential.

Finally, the importance of first-party data cannot be overstated. Using the client’s existing customer list to create lookalike audiences on LinkedIn was a game-changer. It allowed us to bypass some of the broader targeting inefficiencies and directly reach users who shared characteristics with their most valuable customers. This is why I always tell clients: guard your data, clean your data, and use your data. It’s your most potent weapon.

The Project Ascend campaign demonstrated that even with a clear strategy, constant monitoring and agile optimization are what separate good results from truly exceptional ones. It’s not just about spending money; it’s about spending it intelligently, guided by insightful expert analysis.

Ultimately, successful marketing campaigns aren’t about magic formulas but about meticulous planning, continuous testing, and the courage to adapt based on real-world performance data. This campaign, with its clear objectives and iterative improvements, serves as a powerful reminder that even when things aren’t going perfectly, diligent analysis and strategic adjustments can turn the tide dramatically.

Donna Watson

Principal Marketing Scientist MBA, Marketing Science; Certified Marketing Analyst (CMA)

Donna Watson is a Principal Marketing Scientist at Aura Insights, specializing in predictive modeling and customer lifetime value (CLV) optimization. With 14 years of experience, he helps leading brands transform raw data into actionable strategies that drive measurable growth. His expertise lies in leveraging advanced statistical techniques to forecast market trends and personalize customer journeys. Donna is a frequent contributor to the Journal of Marketing Analytics and his groundbreaking work on multi-touch attribution models has been widely adopted across the industry

Credentials 14+ years experience