LLM Industry Benchmark 2026: Which Sources Shape Marketing & Advertising Recommendations?

When potential customers ask ChatGPT about advertising platforms, marketing tools, or campaign options, the answer rarely comes from first-hand experience. AI systems assemble it from a mix of sources, and that source mix is what this article examines. Peak Ace’s LLM Industry Benchmark analysed 1,000 standardised prompts across ten industries in a fixed German-market snapshot. This article looks at the sources behind marketing and advertising answers and what they may mean for content and Digital PR.

Note: This article is an excerpt from the LLM Industry Benchmark. The complete analysis, including the source mix for all ten industries, is available in the benchmark itself.

In short:

• In our benchmark, product pages lead the source types for recommendation and pricing questions in marketing and advertising, while comparison pages lead across all industries.
• Comparison formats are accessed particularly often for direct comparisons, and how-to guides come first for criteria and advice questions.
• Mentions in this industry are moderately to highly concentrated within the top 20 entities.
• All findings refer to this prompt set, this time period, and this setup.

Methodology and data corpus

The benchmark is based on a clearly defined data corpus. The following sections describe its structure and the key figures derived from it.

Prompt set and structure

The prompt set comprises 1,000 standardised, bottom-of-funnel prompts: 100 commercially relevant prompts per industry, distributed across ten topic clusters. The prompts are geared towards purchasing decisions, i.e., questions people ask shortly before making a purchase or booking. All prompts are brand-neutral, so no brand names appear in the questions.

Search intent

Each industry covers four search intents, with 25 prompts per intent:

• Best/Top/Test Winner: recommendation questions, such as “Which tool is best for…?”
• Comparisons: contrasts, such as “Which is better, A or B?”
• Prices/Where to buy: questions about costs and sources, such as “Where can I book and test…?”
• Criteria/Advice: questions about selection criteria, such as “What should I pay attention to when…?”

Market, language and time period

The prompts are in German and geared towards the German market. Answers in German were observed from 1–31 August 2026.

Systems and repetitions

Gemini, ChatGPT, and Google AI Overviews were observed. Each prompt was executed 15 times, five times per system. Individual responses were treated as samples, which allows a better classification of fluctuations.

Database

The data comes from Peec AI exports. Fifty brands per industry were monitored, and the report focuses on the top 20 brands by visibility.

Key figures

• Visibility: how frequently an entity appears in the responses to the prompt set.
• Share of Voice (SOV): an entity’s share of total visibility within the top 20, normalised to 100 percent.
• Sentiment: the tone of the responses, rated on a scale of 0 to 100.
• Retrieval: a source has been accessed as an information base. It does not automatically mean the source is cited in the answer.

Sources mix and industry context

Industry is the lens for this analysis, not a scoreboard. The question is which source types AI systems draw on for each search intent, and how that differs between industries.

Industry averages

In the benchmark, Marketing & Advertising reaches a visibility of 6.85 percent, an SOV of 4.15 percent, and a sentiment of 52.8. Across all ten industries, the averages are 6.94 percent, 3.72 percent, and 54.58. These are indicators for this setup, not market shares or quality ratings.

Concentration within the top 20

For Marketing & Advertising, the HHI within the top 20 is 1,200, and the top 3 share is 51 percent. About half of the SOV in this group falls on three entities. This describes how mentions are distributed in this dataset. It does not show who leads the market, nor why an individual brand was selected.

Leading source types by intent

Source mixes differ by intent. The table compares the leading source type in Marketing & Advertising with the cross-industry picture:

Four patterns stand out in the data:

• Product pages lead in this industry. For recommendations and pricing, product pages come first in marketing and advertising, while comparisons lead across all industries. This may indicate that platforms and tools are often described most directly on their own pages.
• Comparisons remain central. Comparison formats lead direct comparisons in both views, and they also feature in recommendation and pricing lists.
• Advice changes the mix. For criteria and advice, how-to guides lead, followed by product pages and articles. How-to guides account for 22.33 percent of retrieval across industries, compared with 5.81 percent for recommendations.
• Product pages appear across all intents. They rank second for comparisons and advice, which suggests that product information plays a role at several stages of the decision.

Retrieval means that a source has been accessed. This does not automatically imply that it will be cited or that the brand will appear in the final answer. Source types describe page formats, not ownership or authority. “Other” is a residual category and not a specific format.

Most frequently mentioned entities

In our Marketing & Advertising prompt set, LinkedIn reached 21 percent visibility, while Google Ads and Meta each reached 19 percent. These platforms are themselves products covered by the commercial questions. For purchase-near questions about platforms or tools, the platform is often the direct subject of the answer. Their visibility should therefore not be read as evidence of superior content performance or overall brand strength.

What this means for SEO teams and marketers

The data reveals a pattern: depending on the intent, answers draw on different types of sources. This can help you position your content more deliberately. The goal is not to be visible everywhere. Instead, identify what you offer, where you have well-founded answers, and where visibility actually matters. Our results suggest the following practical implications:

• Handle answer types separately. A comparison should explain alternatives, advantages, and disadvantages. A guide should clarify requirements or support a task. A general article rarely fulfils both tasks equally well.
• Keep product information consistent. Features, pricing, availability, and terms should match on product pages, in comparisons, and in third-party sources. Contradictions increase the risk of inaccurate statements in generated answers.
• Check how you are described where users compare. Verify that pricing, features, integrations, and use cases are represented correctly in relevant comparison and listicle sources.
• Explain criteria and link them to offers. As how-to guides lead for advice questions, explain how your product addresses a specific task, and connect these explanations to suitable products so that guidance and offer do not drift apart.
• Prioritise your Digital PR. Identify the most important third-party sources per intent and check what they say about your brand. Not every publication is equally relevant, so use the source mix as a starting point for a prioritised list.
• Treat monitoring as an ongoing task. Track brand mentions and accessed sources with a consistent prompt set, analysed by category, intent, and system.

Expert tip: Gordon Herenz, Team Lead Content Strategy & Digital PR at Peak Ace, recommends building comparison and list formats where they make sense, seeking placement in relevant rankings and listicles, and strengthening citable authority on independent sources through consistent Digital PR.

Limits of expressive power

The results are valid only for this specific prompt set, time period, and setup. They do not represent a complete measurement of an industry’s or brand’s visibility in AI systems. In particular, please note:

• The prompt set shapes the outcome. Different questions or wording can lead to different answers.
• Individual responses are samples. Small differences should not be overinterpreted.
• Three systems are not one system. Gemini, ChatGPT, and Google AI Overviews cannot simply be equated, and patterns may differ by engine.
• Retrieval is not citation. Source fetches do not establish causal mechanisms, and the mix of sources cannot be read as proof of impact.
• Dialogue and personalisation are missing. Typical purchasing situations cannot be reliably described on this basis.
• Sentiment is not a quality assessment. It describes the tone of the responses in the dataset, not actual customer satisfaction with a brand.

Read the full benchmark

This article provides an excerpt. The complete LLM Industry Benchmark includes, among other things:

• The source mix for each of the ten industries and all four intent types
• Visibility, SOV, and sentiment within the benchmark
• HHI and top 3 share within the top 20 of each industry
• The leading entities observed per industry
• The cross-industry shares of the source types
• The industry scorecards and the complete methodology

Download the full LLM Industry Benchmark.

Looking to strengthen your brand’s presence in AI answers? Get in touch with our LLMO and GEO team.

FAQ: Marketing & Advertising in LLM answers

Which source types do LLMs use for marketing and advertising questions?

In our benchmark, product pages lead for recommendation and pricing questions, comparisons lead for direct comparisons, and how-to guides lead for criteria and advice questions. This applies only to the examined prompt set and time period.

Are the most frequently mentioned brands the market leaders?

No. The figures show which entities were mentioned most often in this dataset during the observation period. They are not market shares, and the platforms at the top are themselves products covered by the prompts.

Does AI visibility depend solely on one’s own website?

No. Comparisons, listicles, guides, and other third-party formats play a significant role. It is therefore worth checking specific URLs to compare your own content with relevant third-party sources.

How can we test whether measures are effective?

In three steps: audit your sources and content, check the timeliness and consistency of your information, and then re-measure with the same test questions under the same conditions.

What does the benchmark measure?

It captures visibility, SOV, sentiment, concentration, and retrieval patterns for a defined prompt set, market, and time period. Results are aggregated across Gemini, ChatGPT, and Google AI Overviews.