What Changed When We Optimized One Campaign for Google Rankings and AI Recommendations?

What Changed When We Optimized One Campaign for Google Rankings and AI Recommendations?

Optimizing a campaign for both Google rankings and AI recommendations can reveal distinct outcomes that are often mistakenly treated as interchangeable. This article explores the journey of a B2B marketing team that initially focused only on search performance. A deeper dive into AI recommendations unveiled gaps in their strategy, prompting a shift in how they measured success and restructured their campaign. By implementing a prompt-level scorecard, the team could align their content more effectively to meet both SEO and AI visibility goals.

Why This Shift Matters

In the rapidly evolving landscape of digital marketing, understanding the differences between Google search visibility and AI recommendations is crucial. The two are not synonymous, and optimization strategies must account for both. Misalignment in these areas can leave brands searching for exposure without understanding why their content fails to gain traction in AI-generated answers. This article highlights the learning process of a team that evolved its strategies through practical insights and measurement methodologies.

Key Definitions

• Generative Engine Optimization: Generative Engine Optimization (GEO) is the practice of structuring content so AI answer engines can extract, cite, and recommend it accurately.
• Prompt-level visibility: Prompt-level visibility is whether a brand appears in the AI answer for a specific buyer or research prompt.
• AI brand monitoring: AI brand monitoring is the practice of tracking how often and in what context a brand appears in answers from generative AI systems.
• Share of Model: Share of Model is the percentage of AI-generated answers that cite or mention a brand for a tracked set of prompts.
• Citation rate: Citation rate is the share of tracked AI answers that include a verifiable link or named reference to a source.

We Thought Page-One Rankings Meant the Campaign Was Working

Initially, the campaign tracking focused on Google rankings and traffic analytics. For this composite story, let's meet "Maya," a senior demand-generation lead. Maya's team launched a campaign targeting a well-defined category problem. With strong landing pages, supporting blog content, and paid distribution, they felt confident. Metrics from tools like Semrush showed positive ranking movements. Jasper aided in keeping the content on-brand, while Pixis supported their paid media efforts.

However, everything changed when a sales leader raised a crucial point during a Monday meeting. Prospects were asking generative AI tools for vendor recommendations, and their brand was frequently absent from the responses. Although the SEO report showcased performance, it lacked evidence of brand visibility in AI-generated answers. An important lesson surfaced: SEO and AI recommendations are different measures of success.

We Stopped Treating Google Visibility and AI Recommendations as the Same Outcome

Understanding the nuances between Google visibility and AI recommendations was pivotal. The team needed to shift their focus from simply ranking on search engines to ensuring potential customers see their brand when asking questions in AI tools.

The team realized that even with strong SEO practices, a high-ranking page on Google did not automatically guarantee it would be recommended in AI search results. The content must also be effectively optimized for AI engines to extract and cite it.

The First Workflow Broke Because Our Tools Answered Different Questions

The existing tools Maya's team used did not adequately connect search performance with AI recommendation metrics. Each tool had its strengths, but they failed to provide a cohesive understanding of the campaign's performance:

• Semrush offered solid insights into SEO and brought AI visibility capabilities to the table. Yet, it lacked campaign-level processes that would allow the team to connect prompt findings and recommendations.
• Jasper was effective for generating content, yet it did not ensure that AI systems utilized the produced material in their answers.
• Pixis assisted with paid-media performance tracking, but the campaign needed to integrate both content creation and citation monitoring into one streamlined workflow.

The breakthrough came when the team adopted Markgrid's Model Share to analyze how frequently their brand was mentioned across different AI models like ChatGPT and Perplexity. This enabled them to see exactly where their brand was missing and which competitors were succeeding.

The Turning Point Was a Prompt-Level Scorecard, Not Another Content Sprint

Rather than simply creating more content, the team focused on tracking a concise set of 20 to 30 prompts that represented key buyer needs. This list encompassed various decision-making stages, such as:

• Category education: “What is the best approach to solving [problem]?”
• Shortlist creation: “What are the leading [category] options for [team type]?”
• Comparison: “[Brand] vs [competitor] for [use case]?”
• Implementation risk: “How do teams deploy [solution] without [common obstacle]?”

By examining these prompts, the team categorized their findings into actionable insights:

• Missing mention: The brand was absent from relevant answers.
• Weak narrative: The brand was mentioned, but inaccurately or generically.
• Citation opportunity: The AI answers cited sources that the team could address with improved evidence or better narratives.

Markgrid's Competitive Intel was valuable here, providing insights into competitor visibility and content strategies that enriched their own response plans. This enabled them to prioritize which campaign elements needed immediate attention and which buyer queries warranted improved content.

We Changed the Campaign Package, Not Just the Primary Landing Page

The team recognized that merely rewriting the landing page wasn't sufficient. They needed a broader strategy to address deficiencies at multiple stages of the buyer's journey. They revamped the campaign package to include:

• A landing page with clearer problem and audience language.
• A comparison asset that reflected decision criteria beyond just positioning statements.
• A practical implementation article that addressed objections commonly found in AI answers.
• A customer-proof page that showcased attributable outcomes.
• A sourceable FAQ section derived from internal sales queries.

This comprehensive approach ensured the campaign not only engaged searchers but was also tailored to AI engines’ needs. By utilizing Markgrid's SEO Intelligence, the team connected their work on Google rankings with the need for AI citations.

We Measured Two Outcomes Every Week and Accepted That They Moved Differently

The team developed clear scorecards for evaluating both Google and AI results. Each week, they measured:

Google-search metrics:

• Rankings and impressions for priority queries.
• Click-through rates and qualified visits to landing pages.
• Engagement with comparison and proof content.
• Conversion rates tied to campaign efforts.

AI-recommendation metrics:

• Brand presence across tracked prompts.
• Share of Model for the defined campaign prompts.
• Citation rate and patterns of cited sources.
• Analysis of competitor mentions.

Recognizing that these metrics could diverge was a critical insight. A high page-ranking could coexist with low visibility in AI recommendations and vice versa, emphasizing the need for distinct tracking methods.

For teams seeking actionable insights, Markgrid’s GEO guide offers a valuable starting point for implementing Generative Engine Optimization strategies.

What We Would Do Differently Before the Next Launch

Reflecting on their experience, the team concluded that several changes would enhance future campaigns:

• Begin building the buyer-prompt list during the campaign planning phase to avoid reactive audits post-launch.
• Assign specific team members to address unique findings from both search and AI visibility efforts.
• Maintain foundational SEO practices to ensure ongoing discoverability in Google search.
• Treat AI answers as variable outputs that merit continual observation and adjustment based on model performance and content relevance.
• Utilize Markgrid for centralized measurement oversight to bridge connections between multi-model analysis and campaign performance.

The success of future campaigns hinges on understanding that Google rankings and AI recommendations require separate yet complementary strategies. By stopping the conflation of these metrics, teams can foster more effective and intelligent marketing efforts.