Helping your business grow through SEO & PPC.

Sacha Redler

Malte Landwehr is the CPO & CMO of Peec AI, the leading solution for AI Search Visibility. Peec AI has recognized by Business Insider as one of the fastest growing companies in Europe.

Malte is a former VP SEO, VP Product, and management consultant. He once built and ran one of Germany's 40 largest blogs.

Malte Landwehr has 20+ years of SEO experience and is a regular conference speaker and podcast guest. He has been recognized as a Leader (Venturi's Voice, Econsultancy), Visionary (Adello), and expert in Enterprise SEO & AEO.

Malte Landwehr

I spent the last two days at the OMT Summit. Here is what I learned:

1️⃣ Traditional SEO
• SEO as a marketing channel is not dead, yet.
• Glossary content is dead.
• SEO/GEO goals have to be aligned with company goals.
• ChatGPT recommendations are a major lead source for SEO agencies.
• Many SEO agencies are struggling. Customers replace them with vibe-coded solutions. Especially when it comes to content creation.
• The placement of your content on category pages can have a huge impact on your Google rankings.
• Google has a limit of 400 billion indexed documents.
• The longer a document has not been crawled, the less likely it is to be indexed.
• After 128 days without crawling, a document drops from the Google index.
• If documents have little information gain or are outdated, update them.
• The Panda diet is still working well.
• Google cannot measure the quality or content. They use human behaviour as a proxy for content quality.
• Both Google and Bing reward recent content. Even just updating the publishing date can work.
• Adding 6m in revenue with ecommerce SEO is still possible in 2026

2️⃣ Human e-mails
• 77% or B2B decision makers want to be contacted via email. 86% expect personalisation.
• Turn your traditional email newsletter into a human mail. You can expect +50% open rate and +100% CVR.
• The best human mails are 150 words or shorter and contain only a single link.
• If you send out all your newsletters as human mails, people will get used to it and the impact will decrease.

3️⃣ GEO
• LLMs love facts and simple language.
• LLMs are immune to traditional clickbait.
• Keep paragraphs as short as possible.
• Each paragraph should be written around a single thought.
• Write precisely and with high entity-density. Write about relationships between entities. Highlight these relationships with schema.
• LLMs ignore hreflang.
• Since fanout queries are often in English, not having a strong presence in the English web is a huge disadvantage.
• Use html instead of JavaScript and keep the TTFB below 300ms.
• AI visibility is the responsibility of the whole company, not just the marketing team.

4️⃣ Product Grid SEO
• 50% of ecommerce searches in Google have a product grid.
• Outside of the EU, Google search result pages are replacing category pages of online shops.
• Past: User > PLP > PDP > Buy
• Now: User > Product Grid / AI > PDP > Buy
• Soon: User > Product Grid / AI > Buy
• Success in product grids: PDPs, feeds, structured data, variants, item group ID (Google Merchant Center), images, 
• Best product titles: brand + product title + product type + attributes 
• Use the full 150 characters for the product title - only the first 70 are displayed. It does not have to be the same title as on the PDP.

6️⃣ CTR/Click uplift from Product Grid SEO:
• Titles: +54%
• Split SEO and SEA product feeds: +60%
• Optimising PDPs: +100% 
• Images: +115%
• Variants and item group ID: +250%
• Product attributes: +205% revenue

Did you know the shopping widget in ChatGPT is powered by Google Shopping?

You heard that right. No ACP, UCP, Agentic Checkout, or direct feed integrations.

OpenAI is simply scraping Google Shopping.

My Peec AI colleague Tom Wells even reverse engineered which scraping provider OpenAI is using.

Actions you should take:
• Check your attribution. Some of your free Merchant Center clicks might come from ChatGPT.
• Optimize your ranking in Google Shopping / product grids. It will also boost your presence in ChatGPT!

By the way, if you want to see the exact shopping fanout queries used by ChatGPT, you can find those in Peec AI.

Big pricing and feature updates just landed at Peec AI.

We made all plans cheaper and included more prompts. We also made everything more flexible: models, projects, and tracking frequency. Pick and change everything freely!

Of course, all existing customers will be migrated to the new pricing automatically. Many can add another model or more prompts - for free.

On the feature sidem, we added:
• Share of Voice (SoV) in addition to Visibility.
• Much better prompt suggestions. Personalized to your brand and automatically tagged.
• More recommendations in the Actions section. Recurring themes on source URLs and continuous robots.txt checks.
• Shopping fanout queries in ChatGPT. Did you know they are different from the regular fanout queries? We now display them for every prompt you track.
• We made the app much faster. You can now look at 12+ months of historical data, switch between daily/weekly/monthly aggregation, or change the line chart to a bar chart.

ChatGPT regularly cites deleted Wikipedia articles. And marketers can abuse it!

I have repeatedly seen deleted Wikipedia articles in ChatGPT citations.

Even 4 months after they have been deleted!

The articles do not rank in any search engine (Bing, Google, DDG, Brave).

No other LLM (AI Overviews, AI Mode, Perplexity, Gemini, Copilot) cites them.

An yet the get cited by ChatGPT over and over again.

Below is a chart that show how often an article deleted in November shows up as a source in ChatGPT for a set of prompts I am monitoring with Peec AI.

As you can see, the deletion had no impact on how often I encountered the article.

This suggests to me that OpenAI has a cache of certain domains - at least Wikipedia. And they do not properly remove documents from it.

SEOs have always spammed Wikipedia. I fear this observation will increase such behaviour.

Get a Wikipedia article for your brand. Lose it after a few days. Profit from additional AI search visibility for months.

I am on my way to SMX Munich to talk about AEO/GEO. And to do something Peec AI has never done before.

For the first time ever, we will have a Peec AI booth! If you are at SMX, please come by and say hello.

If you don't care about visibility in LLM-based search and answer engines, we also have a barista.

How many searches is ChatGPT seeing (relative to Google)?

To the best of my knowledge, around 10%.

Below, is all the research on the topic I could find.

Good news for me: everyone arrived at the same range (4% to 12%) that I originally calculated. 🤗 

The latest numbers are from Rand and on the low end. But they are based on Desktop. So I am not surprised they are lower than what most people arrived at. Desktop is boomer-skewed and office/work-skewed.

Not getting cited by LLMs? You can debug that!

Here is the exact framework I use to find the gap - and fix it. Tried and tested across dozens of domains.

Let’s assume I have a website and want to be the source for a set of prompts. Each prompt generates some fanout queries. I have already gathered these.

With this framework, we will iterate through all prompts and all fanout queries

1️⃣ Is your website directly cited in the answer?

Yes? Perfect. Your work is done and you should probably focus on other prompts where you are not cited, yet.

No? Go to step 2.

2️⃣ Is your domain used as a source but not explicitly cited?

Yes? Make it more citeable / quotable. Add a summary with the relevant key statement. Keep it short. Write in clear and simple language. Keep the entity density high. If relevant, explicitly express relationships between entities. If you use technical terms or abbreviations, shortly explain them. Use an authoritative, declarative voice.

The summary does not have to be for the whole article. You can also summarize a long paragraph, a table, or a chart/diagram.

It a summary does not fit, you can use question-answer style content blocks.

No? Go to step 3.

3️⃣ Does your domain rank for the fan-out query but not get used as a source?

Yes? Improve the content so it is more source-worthy. Make sure you agree with the broad consensus. Add unique information no other source has. Strengthen your EEAT signals. Add quotes from trustworthy sources. Explicitly name your sources.

Ideally, expand your document to cover multiple fanout queries.

No? Go to step 4.

4️⃣ Do you already have a relevant page for the fan-out query that does not rank yet?

Yes? Make sure it is indexed. Build internal and external links. Improve the content. Make sure there are no crawling or rendering issues. In short, do SEO.

No? Create that page.

No alternative text description for this image

A few thoughts on how to get started with prompt tracking.

In traditional SEO, the playbook was straightforward. SEO tools showed that a keyword like “project management software” receives around 10,000 search

Open article: How to Choose the Right Prompts for LLM Tracking by Malte Landwehr • 4 min read

How to Choose the Right Prompts for LLM Tracking

Episode 2 of my podcast series with Niklas Buschner is out. We covered a lot of ground: 

1️⃣ Perplexity is done. Even Grok overtook Perplexity. Grok!

2️⃣ Bing Webmaster Tools show some fanout query data. SEOs are excited about it. But the source and exact scope of the data are unclear.

3️⃣ I went on an unplanned 10-minute rant about Google Search Console data and why I believe using the built-in filtering in GSC should be a fireable offense.

4️⃣ ChatGPT almost always runs at least one fanout query in English. Even if users prompt in French, German, or Spanish. This has implication on content creation for many brands!

Search for #MonthlyLandwehr on Spotify or Youtube to find the episode.

What should we cover in the next episode?

I recently joined Torsten Schwarz’s market podcast.

We talked about:
1️⃣ AI Search
• Thinking in clicks, keywords, search volume, and URLs will limit your ability to understand and measure AI search.
• The market shares of LLMs are highly dynamic. Gemini surged to second place. Perplexity disappeared.
• Many marketing agencies are facing difficulties because clients have already priced in the expected time-saving efficiencies of AI tools.
• LLM answers are non-deterministic and highly personalized. You need to measure the probability of showing up, not rank.
• AI search is about entities not URLs. 

2️⃣ Measurement and Data Tracking
• The most effective way to measure AI search success currently is self-reported attribution - asking new leads where they heard about you.
In the B2B SaaS industry, 20% - 30% of leads attribute their discovery to LLMs.
• Measure AI search like a brand marketer - not like a performance marketer
• Look at aggregated data (topic, persona, funnel stage) not individual chats
• Analyze internal search, support tickets, and sales call transcripts to understand language of your audience.

3️⃣ Content Strategy and Digital PR
• Digital PR is experiencing a major resurgence because LLM visibility is heavily dependent on brand mentions within large media websites.
• LLMs have a recency bias. You need an always-on digital PR campaign.
• Content format matters a lot. A prompt asking for a car's traits will pull from brand profiles, while a prompt asking for the fastest cars will pull from listicles.
• LLMs execute background "fanout queries" that often use English search terms - even if the user's prompt is in German or Spanish - which gives a significant advantage to US-based sources.
• Publishing specific, targeted content in English can be a strategic way for non-US companies to become more visible in such LLM background searches.
• During quick searches, LLMs do not watch videos or "look" at complex diagrams. 
• To ensure LLMs can understand tables or graphics on a website, brands should add a 1-to-2 sentence text summary above or below the visual element.

4️⃣ Backlinks & Sources
• Backlinks are not dead. They still influence the traditional search engines (like Google & Bing) that LLMs use to retrieve their source documents.
• For the representation of your brand in LLMs, building links to your own website is now less important than ensuring your brand is mentioned positively on authoritative third-party websites.
• Grokipedia's utility as an LLM source is currently dropping rapidly because it is losing its Google rankings.
• YouTube is frequently used as a source by Perplexity and Google AI Overviews, but very rarely by ChatGPT.

5️⃣ My predictions for the next 12 months:
• Google's standard search layout will look differently, likely integrating more AI Overviews or defaulting to AI Mode for certain queries.
• Many players will continue to push agentic shopping (AP2, ACP, UCP, etc.).

SEOs are best equipped to lead AI search efforts. This is the topic I discussed with Jordan Koene on the Voices of Search podcast.

We covered:
• The shift from rankings and clicks to brand visibility.
• Why a vacation in South Korea made me join Peec AI.
• SEO is the foundation for AI visibility.
•  SEOs are best equipped to understand AI search.
• The shift from click-based measurement to brand visibility.
• Prompt volume is not a great metric.
• AI search is very personalized.
• ChatGPT often cites publishers or forums instead of brand websites.
• Google is the biggest player in AI search. Because of AI Overviews. Number 2 is ChatGPT with 10% of Google's size. 
• In markets like China, DeepSeek and Qwen have significant market share.
• LLMs sometimes expand the original prompt with fanout queries.
• Try to influence the grounding layer - not the training data.
• LLMs use individual passages not whole documents for grounding.
• Different LLMs use different search engines for grounding: Google, Bing, Brave, DuckDuckGo.
• You must survive multiple stages: crawling, indexing, ranking for fanout queries, source selection, citation, and positive brand framing.
• A page can rank for relevant fanout queries and still fail to become a source.
• Advice for small brands: describe yourself consistently online.
• Ads in ChatGPT are good for GEO. They validate the commercial value of showing up in ChatGPT. And the post valuable users pay to not see ads.
• ChatGPT shopping currently relies heavily on scraped Google Shopping results.
• Merchants should maximize reach by distributing product data broadly rather than trying to withhold it.
• Many online shops risk becoming commodity fulfillment layers with weaker differentiation.
• The long-term direction points toward fewer online shops and more direct connections between suppliers and AI-driven buying interfaces.
• Real-world examples: click-based attribution showed 0% ChatGPT influence while customer surveys showed 22% ChatGPT influence.
• A single prompt run once is too noisy to interpret reliably.
• Visibility should be aggregated across many prompts and many days to get a meaningful directional signal.
• Citation share is a key supporting metric - it measures how often a domain appears among all citations versus competitors.
• Source coverage is another key supporting metric - it measures how often a brand is mentioned across the source documents used by LLMs.
• Improving citation share and source coverage tends to lift overall visibility.
• Gains in AI visibility usually correlate with increases in sales, traffic, and revenue.
• Automated large-scale content creation can work briefly but often ends in ranking loss or penalties over time.
• Prompt sets should be built around topics, personas, funnel stages, and intent types rather than isolated keywords.

Most SEO agencies are now GEO agencies.

iBusiness recently published a list of the top 100 SEO agencies in Germany.

The ranking included - self-reported - focus areas.

➡️ 84% of SEO agencies now offer GEO services.

➡️ For 30% GEO is one of their focus areas.

➡️ 7% primarily offer GEO.

Do you think this is a realistic representation of the current agency market?

taplio

The Battle

Sacha Redler

Malte Landwehr

Followers

Posts

Engagements

Likes

Performance ScoreMalte Landwehr's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Sacha Redler

Malte Landwehr

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Malte Landwehr's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreMalte Landwehr's LinkedIn Performance