Google has made some updates to Search profiles with improvements aimed at publishers, the search company has told us. Specifically, now you can manage multiple profiles from one login, there is an improved article displays and Google lowered the social/video follower requirement to 10,000 across platforms.

One login, multiple accounts. Google now allows you to claim and manage Search profiles for all of your sub-brands from a single login. This removes the old method of using multiple Google accounts to manage each Search profile. Many larger publications manage Search profiles for numerous brands; this should simplify account management.

Lower subscriber requirement. When Search profiles first launched, Google required you to have 100,000 subscribers on YouTube, Instagram, X, or TikTok. Now that threshold is down to 10,000 followers across YouTube, Instagram, X, or TikTok.

Google previously lowered that threshold to 35,000 followers about a month ago.

New design. Google is rolling out an update design for Search profiles, specifically the display of articles. Google said this was done to make it “easier for people to discover and dive deeper into your content.”

Now the image thumbnails are optimized to help your articles “look their best,” Google added. Gogle also increased the length of headlines. Here is what it looks like:

Why we care. Google continues to add features to help publishers get more attention from Search and Discover. Now Google is allowing more publishers to gain access to Search profiles with the lowered follower requirement. Plus, the design changes and login requirements can help publishers get the most out of Search profiles in Google Search.
]]> google-search-profile-design-update https://searchengineland.com/openai-is-turning-chatgpt-ads-into-conversations-488949 Wed, 16 Sep 2026 15:01:35 +0000 https://searchengineland.com/?p=488949 OpenAI is expanding ChatGPT Ads with conversational Sponsored Agents, AI-powered campaign tools and new 3rd party integrations.]]>

OpenAI is adding conversational ads, AI-powered campaign creation and new HubSpot and Shopify integrations as it builds out ChatGPT into a broader advertising platform.

What’s new. OpenAI is testing “Sponsored Agents,” which let users start a separate, clearly labeled conversation with a business-sponsored agent after seeing an ad in ChatGPT.

How it works. Someone who sees an ad for a product could ask the sponsored agent follow-up questions about features, sizing, compatibility or other considerations, then follow a link to the advertiser’s website when they’re ready to take the next step.

The key distinction. OpenAI says conversations with Sponsored Agents are separate from ChatGPT’s independent answers and from the user’s original ChatGPT conversation. The company has previously said advertising doesn’t influence ChatGPT’s answers and that advertisers don’t get access to users’ private conversations.

Where it stands. Sponsored Agents are being tested with select U.S. advertisers. OpenAI is also accepting interest from businesses looking to participate in future Sponsored Agent testing.

Meanwhile. OpenAI is bringing more of the campaign workflow into AI. Advertisers can use natural-language prompts in ChatGPT Work to create, update and analyze campaigns through an Ads Manager plugin — potentially turning a website or campaign brief into an advertising campaign without navigating a traditional ads interface.

Inside Ads Manager. OpenAI is also adding AI assistance that can suggest ad copy and imagery based on an advertiser’s landing page and campaign objective. Advertisers can review and edit the generated creative before deciding whether to use it.

Going further. Advertisers can opt into AI-powered text customization that adapts existing headlines and descriptions to the context of a conversation and translates copy into a user’s preferred language.

Why we care. The combination could compress several parts of the advertising workflow: campaign creation, creative development, optimization and analysis into conversational prompts, while opening a new ad format built around dialogue rather than a click alone.

And it’s coming to existing marketing stacks. OpenAI is integrating its advertising tools with HubSpot and Shopify, its first CRM and ecommerce partners for ChatGPT Ads, respectively.

For HubSpot users. Businesses can connect a ChatGPT Ads account with their customer context, bringing advertising workflows closer to CRM data and lead follow-up. OpenAI already offers a broader HubSpot plugin that can bring HubSpot context into ChatGPT workflows.

For Shopify merchants. U.S. merchants can use a ChatGPT Ads app to create and manage campaigns and monitor performance. Shopify product data is already integrated into ChatGPT through Shopify Catalog, which OpenAI says helps merchants’ products appear more accurately in relevant shopping conversations.

The bigger picture. OpenAI has been steadily building the infrastructure of a conventional advertising platform, adding self-service Ads Manager, CPC buying, conversion optimization, measurement and targeting while trying to make the actual ad experience native to conversational AI.

Dig deeper. Reimagining advertising with AI ]]> https://searchengineland.com/geo-pillars-488782 Wed, 16 Sep 2026 15:00:00 +0000 https://searchengineland.com/?p=488782 Your next search competitor may be an AI's shortlist. GEO addresses what it takes to get cited, recommended, or selected by AI systems. ]]>

In traditional search, visibility meant ranking. In AI search, it means getting cited, recommended, or selected.

GEO discussions often focus on whether it replaces traditional SEO, whether SEO is still relevant, or whether it’s really something new. In my view, they miss the more important question:

• What do you actually want to achieve with GEO?

This distinction between goals is often lacking. Not all GEO is created equal. Depending on the goal, the optimization approach, measures, and resource requirements change significantly.

The central goals are:

• Get your content cited: Design and position your content so it can be used, quoted, and referenced as a source by generative systems.
• Get your brand mentioned or recommended: Increase the likelihood that your brand will be mentioned, categorized, or recommended in AI-generated responses.
• Get your products selected: Position products and offers so they’re considered and specifically mentioned in relevant AI responses.

Three goals. Three different optimization approaches. Different measures, signals, and resource requirements.

GEO’s three pillars

GEO can be divided into three distinct areas:

• LLM readability optimization: How is your content understood by language models and cited as a source?
• Brand context optimization: How do you ensure your brand is mentioned and recommended in AI responses?
• Agentic commerce optimization (ACO): How are your products selected and processed by autonomous AI agents?

Each pillar requires different skills, resources, and responsibilities across SEO and content, PR and brand communications, ecommerce, and product data management.

Pillar 1: LLM readability optimization

LLM readability describes how efficiently large language models can process, understand, and use digital content as a source of answers. Unlike traditional readability, which focuses on the human reader, it focuses on how clearly and effectively content can be processed by a language model. That means using clear structure, precise language, and semantically unambiguous content to improve its citability within GEO.

Today, AI systems predominantly use retrieval-augmented generation (RAG). They search external sources for relevant text segments — so-called chunks or “nuggets” — and synthesize a response from them. Content with poor structure, ambiguous wording, or grammatical errors risks being misclassified in the AI’s vector space or ignored because of low confidence.

Traditional SEO remains the technical and semantic foundation that helps AI systems find, index, and consider content as a potential source. Without a properly optimized page, your content might not even enter the pool of documents considered during query fan-out, the process by which AI systems break down a user query into multiple subqueries and search indexes for relevant sources. If your content isn’t retrieved, it can’t be considered for the AI-generated answer.

Important classic SEO factors remain indispensable:

• Indexing: Content must be stored in the search engine index.
• Snippets: Meaningful titles and meta descriptions influence how AI systems summarize content.
• Technical SEO: Loading time, mobile optimization, clean HTML structure, XML, and HTML sitemaps.
• Internal linking and authority: Signals that help AI crawlers assess relevance and trustworthiness.

Once content can be found, LLM readability helps ensure it’s understood, extracted, and used as a quotation. The logical order is:

• Document retrieval first, LLM readability second.

Even the most brilliant, LLM-optimized text remains ineffective if it doesn’t reach the AI’s candidate pool. Conversely, a well-ranking text without LLM readability is no longer sufficient to be cited in AI responses. Only those who strategically combine both levels maximize their chances of being listed as a trusted source in AI Overviews, ChatGPT, Perplexity, Gemini, and AI Mode.

Traditional SEO isn’t replaced by GEO. It’s a prerequisite for it. LLM readability expands the existing SEO toolkit with a new, indispensable dimension.

Key factors of LLM readability

Natural voice quality

The foundation is flawless grammar and spelling, as well as clear, natural phrasing without keyword stuffing. LLMs are trained to recognize n-gram patterns in natural language — artificial keyword clusters disrupt semantic analysis and reduce the chance of being considered an authoritative source.

Chunk relevance: Paragraphs as independent ‘nuggets’

LLMs process text in chunks, or blocks of information. Each paragraph should therefore function as an independent “information nugget”: self-contained, fact-based, and with a clear thematic focus. Two practical rules:

• Short paragraphs of under 250 words with one clear idea per passage.
• Use the “W” questions (who, what, when, where, why) as subheadings so that the headline and content exactly match the potential user intent (chunk relevance).

The pyramid principle: Start with the main point

To avoid the “lost-in-the-middle” phenomenon — the well-known performance drop of LLMs when faced with information in the middle of a text — the most important statement is always placed at the beginning.

Reasoning-capable structure

Since AI responses are increasingly based on reasoning methodologies that go beyond query-passage matching, texts should also take this into account.

Direct answer → Explanation → Evidence → Context

This structure applies to the entire document as well as to each individual chapter and paragraph.

Entity focus and context management

LLMs understand information better when it’s embedded in a network of known entities: people, places, organizations, concepts, products, or events. Mentioning related entities (e.g., RAG, vector databases, tokenization in the context of LLM readability) sharpens the thematic framework for AI. A balanced context-to-information ratio prevents relevant facts from being lost in “filler text.”

Structuring and formatting

• Clear heading hierarchy (H1, H2, H3) with descriptive titles.
• Lists and tables for better extractability.
• Front-loading and 512-token rule: LLM agents often only read the first 350-400 words of long documents — the most important statement must be placed there.
• Statement-level attribution: Link evidence inline directly at the end of the sentence, not collected at the end of the page.
• Reasoning structures: LLMs prefer logical sequences of steps such as Premise → Comparison → Evaluation → Conclusion.

Information density and length

Ideally, the entire text shouldn’t exceed approximately 2,000 words. High information density within an appropriate length clearly trumps lengthy, context-poor texts — quality over quantity.

Chunk engineering in practice

Beyond the basic factors, other practical techniques can optimize machine processing:

• Semantic triplets: Simple subject-predicate-object sentences (“Paris is in France”) help LLMs clearly identify entities and relationships.
• Factual priming: Mentioning related facts in the same context (e.g., historical predecessors of a product) to support the model in internal knowledge retrieval.
• Consistent terminology: Use the same terms consistently for core concepts — different synonyms lead to tokens falling into different clusters and breaking n-gram patterns.
• Multimodal content: Combine text, images, videos, tables, and audio, as conversational search interfaces are multimodal. Videos should include transcripts.
• Multisensory metadata: Images and videos require semantically rich alt text and filenames in natural language — LLMs translate these into “Multimedia Content Tags.”
• Unique, exclusive insights: Proprietary data, original research, or expert opinions increase the likelihood of being cited as an authoritative source.

LLM readability is a fundamental technical requirement for your content to have a chance of being seen in AI-generated search results.

Clear structure, precise language, and well-organized paragraphs benefit both human readability and machine processing. LLM readability isn’t a departure from high-quality content, but its logical evolution as search becomes more AI-driven.

Dig deeper: Chunk, cite, clarify, build: A content framework for AI search

Pillar 2: Brand context optimization

Brand context optimization (BCO) is the discipline within GEO that aims to ensure your brand, company, or products are mentioned and recommended by name in AI-generated responses. Unlike optimizing for citability, the focus isn’t on linking to your own content, but on mentioning your brand as a suitable solution for a user need.

An example illustrates the difference: If a user asks ChatGPT, Perplexity, or Google AI Overviews, “Recommend a project management tool for remote teams,” the AI doesn’t deliver 10 blue links.

Instead, it provides a curated shortlist of two to five brands, often with a brief explanation for each recommendation. Those not on this list simply don’t exist for the user at that moment. There’s no “second page” to scroll through.

Brand context optimization ensures that your brand is one that AI identifies.

Why brand context optimization is so important

The importance of BCO stems from several fundamental changes in user behavior and the way modern search systems work:

• AI takes over the preselection: AI systems increasingly decide which brands and products make it into the user’s consideration set. Those that don’t appear lose relevance in the decision-making process.
• Early visibility in the decision-making process: Users often make their initial selection in dialogue with AI, long before they visit a website.
• Reputational risk: The conversational, authoritative tone of AI responses can make statements seem like facts. If LLMs rely on outdated, inaccurate, or selective information, this can cause lasting damage to brand perception.
• No advertising option available: Visibility in LLMs can’t be bought — there is no PPC equivalent.

How LLMs form brand associations

Large language models such as GPT-4, Claude, or Gemini represent concepts as vectors in a high-dimensional semantic space. Brands that frequently appear near certain terms in training and grounding data are mathematically closely linked to those terms.

The underlying principle is called co-occurrence optimization: The more consistently your brand appears alongside the right concepts, the more strongly the model learns this association.

Practical example: Notion. In hundreds of blog articles, YouTube transcripts, Reddit threads, and comparison articles, “Notion” consistently appears alongside terms like “all-in-one workspace,” “team collaboration,” “notes & docs,” and “productivity.” This semantic proximity leads LLMs to reliably recommend Notion when asked about productivity solutions.

For brand context optimization, this means it’s no longer about ranking for a keyword, but about how an LLM semantically characterizes your company.

Look at what works and transfer successful patterns to other subject areas.

Practical measures for more brand mentions

Place content and mentions on authoritative platforms

AI systems rely heavily on content from high-authority websites. Effective measures include:

• Guest posts in well-known specialist portals and blogs that are referenced in relevant fan-out queries.
• Place expert quotes and statements in topic-relevant articles.
• Collaborate with industry platforms on joint content (digital PR).
• Ensure the brand and products are mentioned positively in context.

Rankings and comparisons

AI chatbots particularly favor lists and comparisons like “Top 10 Providers” or “Best Tools” when dealing with product and brand inquiries. If your brand is mentioned in such posts, the likelihood of it being directly incorporated into AI responses increases dramatically.

• Actively advertise on relevant comparison sites and test portals.
• Initiate product tests in specialist media.
• Maintain consistent product data and user reviews.

Active presence in online communities

Analyses show that AI answers rely heavily on forum and community sources. Platforms like Reddit and Quora are particularly valuable data sources because they host authentic discussions.

• Answer questions actively, honestly, and helpfully, without a promotional tone.
• Mention the brand only where it truly fits the context.
• Build a genuine community reputation over the long term.

Dig deeper: How AI models ‘understand’ your brand

Pillar 3: Agentic commerce optimization

Agentic commerce describes a form of digital commerce in which autonomous AI agents independently research, compare, select, and increasingly purchase products on behalf of users. The user no longer delegates just the information search to AI, but also the purchasing decision and transaction.

Examples of this development are already visible:

• ChatGPT’s shopping features include direct product display and purchase options.
• Google’s integration of shopping results into AI Overviews and AI Mode.
• Anthropic and OpenAI’s efforts to build agents that can independently navigate websites and perform actions.

The key change: AI becomes the customer of your shop, at least during product discovery and preselection.

The consequences for online retailers and brands are significant:

• AI becomes a new target audience: Product pages, feeds, and structured data must be designed so AI agents can reliably read, understand, and process them.
• The consideration phase gets shorter: Users may see only the two to five products suggested by AI. Products that don’t make the shortlist are out of consideration.
• Trust signals matter more: AI agents rely on ratings, prices, availability, reputation, and other data to make recommendations. Data quality becomes a competitive advantage.
• Traditional marketing levers lose influence: Emotional advertising, visual campaigns, and store design have less impact when AI makes the initial selection. Facts and data become more important.
• New infrastructure is emerging: Protocols such as Anthropic’s Model Context Protocol (MCP) and OpenAI’s Function Calling create the technical basis for agent interactions with shops and services.

How AI agents select products

AI agents typically evaluate:

• Structured product data, including price, availability, attributes, and categories.
• Semantic fit to the user query.
• Trust signals, including ratings, reviews, return policies, and merchant reputation.
• Consistency across websites, marketplaces, and feeds.
• Availability through agent protocols, APIs, and MCP-compatible interfaces.

Retailers that are weak in these areas risk being overlooked by agents, even if their products are objectively the best on the market.

Practical measures for agentic commerce optimization

Consistently implement structured data

Schema.org markup is the lingua franca for AI agents. Particularly relevant are:

• Product with all attributes (name, brand, category, GTIN, color, size, etc.).
• Offer including price, currency, availability, and delivery time.
• Aggregate rating and review.
• Breadcrumb list for classification in the product range.
• FAQ page for product-related questions.

Structured data must exactly match the visible page content. Otherwise, AI systems may question its trustworthiness and quality.

Maintain and expand product data feeds

Feeds are often a primary data source for AI agents:

• Google Merchant Center feed.
• Meta/Facebook product catalog.
• Amazon listing data.
• Perplexity Shopping feed (where available).
• Bing Merchant Center.

Consistency between feeds and the website is essential. Inconsistencies can undermine trust in AI systems.

Rich, fact-based product descriptions

• Use clear, precise descriptions instead of marketing prose.
• Include specific technical specifications.
• Explicitly define use cases and target groups.
• Highlight comparisons and differentiating features.
• Answer frequently asked user questions directly in the description.

Professionalize review management

AI agents place a high value on authentic user reviews:

• Actively collect customer reviews (Trusted Shops, Trustpilot, Google, Amazon).
• Handle negative reviews transparently.
• Respond to reviews, as AI systems also read them.
• Maintain a consistent reputation across platforms.

Check new protocols and interfaces

• Model Context Protocol (MCP): Check if and how your shop system can provide MCP servers.
• OpenAI Operator / Anthropic Computer Use: Ensure your website supports agent-based navigation through clear buttons, semantic HTML, and accessible forms.
• API access: Provide access to price and availability data.

Strengthen trust and brand signals

Since agents react to reputation signals:

• Display certifications and quality seals prominently, including in structured data.
• Provide transparent information on shipping, returns, and warranty.
• Keep your legal notice, privacy policy, and terms and conditions up to date.
• Build positive mentions in trusted media (see brand context optimization).

Challenges and open questions

Agentic commerce also raises unresolved questions:

• Attribution and tracking: When an AI agent makes a purchase, which marketing channels receive credit? Classic attribution modeling has limited functionality.
• Legal aspects: Who is liable for incorrect purchases made by agents? How are rights of withdrawal exercised?
• Price differentiation: Agents compare with radical transparency, forcing brands to rethink pricing strategies.
• Bot traffic vs. real users: How do you distinguish desired agent access from unwanted scraper access?
• Platform monetization: It’s foreseeable that platforms such as OpenAI and Perplexity will develop monetization models for preferred product placements.

Agentic commerce represents the next step in the development of the previous two pillars. What LLM readability does for content and brand context optimization does for brand associations, agentic commerce optimization does for products and transactions.

AI agents represent a new target audience with specific requirements. Making product data, feeds, reviews, and technical interfaces agent-compatible will determine whether products make it into AI-driven consideration sets.

Dig deeper: Winning the AI decision layer: From AI discovery to agentic commerce

GEO is a foundation for AI-driven visibility

GEO brings SEO, PR, branding, content marketing, and product data management into an integrated approach to AI-driven visibility.

Machine-readable content, clear brand positioning, and agent-enabled product data can improve visibility across AI systems and prepare organizations for new standards, protocols, and forms of interaction.

Visibility in AI responses is the result of strategic optimization across the three pillars of GEO — making content, brands, and products more relevant to both AI systems and the people they serve.

Google Ads is giving advertisers a new way to see how their ad spending and performance stack up against businesses it considers similar to theirs.

What’s new. A Spend Benchmarks report appearing in the Google Ads account Overview compares an advertiser’s weekly spending and clicks with those of its peer group.

For example. One account was shown spending €284 per week compared with €268 among its peers. The same account generated 912 clicks, versus 765 for the peer benchmark.

Why we care. Advertisers have long had visibility into their own budgets and results, but benchmarking against comparable businesses could provide useful context for whether an account is spending and generating traffic, above or below its peers.

How it works. Google says the peer comparison takes into account factors including the advertiser’s industry and where it advertises.

Yes, but. “Similar” businesses aren’t necessarily comparable businesses. Two advertisers operating in the same industry and market could have significantly different margins, average order values, conversion rates, customer lifetime values and growth strategies.

The catch.The benchmark can also be accompanied by recommendations to increase spending to generate more results. That could make a peer-spending comparison feel like a target rather than simply a point of reference.

Reality check. Spending less than comparable advertisers doesn’t necessarily mean an account is underinvesting, just as spending more doesn’t necessarily mean it is overspending. The economics of the individual business still determine whether additional ad spend makes sense.

The bottom line. Google’s Spend Benchmarks could give advertisers a useful new reference point for understanding the competitive environment around their accounts. But peer spending is best treated as context — while profitability, incremental returns and business objectives determine whether there is actually room to increase the budget.

First spotted. This update was spotted by Paid Search expert Thomas Eccel who shared spotting the new feature on LinkedIn.

Google is rolling out a slate of agentic commerce updates aimed at helping retailers reach shoppers who are increasingly using AI to research, compare, and buy products.

What’s new. Google is expanding AI performance measurement, bringing its Business Agent to YouTube ads and adding capabilities to its Universal Commerce Protocol integration.

Getting visibility into AI discovery. Google’s AI performance insights in Merchant Center are now generally available to businesses in Australia, Canada, India, New Zealand, and the U.S. The feature shows retailers how their share of voice compares with other brands across Google experiences, including AI Mode and AI Overviews.

Coming to YouTube. Google is inviting eligible U.S. retailers to participate in a beta that embeds its Business Agent directly into YouTube ads, allowing viewers to ask detailed product questions and receive tailored answers without leaving YouTube.

The idea. A shopper who discovers a product in a video ad could move directly from inspiration to research, asking questions about features, fit or other details within the same experience. Google previously introduced Business Agent in Search as a way for brands to provide conversational shopping assistance.

Meanwhile. Google is expanding the capabilities available through its Universal Commerce Protocol integration hub in Merchant Center. UCP is designed to provide a standardized way for AI agents and commerce platforms to support shopping transactions.

UCP-powered checkout. Merchants using the hub for UCP-powered checkout can now enable features including cart transfers to their own websites and enhanced checkout-flow testing. Google says analytics and additional capabilities are coming soon.

Where it stands. The new UCP capabilities are gradually rolling out in the U.S., with Australia and Canada expected to follow early next year.

Tapestry, the owner of Coach and Kate Spade, used UCP-powered checkout to make its products available for purchase across Google Search — including AI Mode — and the Gemini app ahead of the holiday shopping season.

Don’t overlook the product feed. For all the attention on AI agents, Google says one of the fundamentals of e-commerce remains critical: accurate, detailed product data.

By the numbers. Google says merchants that adopt its core Merchant Center feed best practices see an average 5% increase in conversions the following month. During testing with Lululemon, Google says conversational attributes supplied by the retailer were incorporated 50% of the time in relevant product recommendations in AI Mode.

And there’s loyalty. Retailers can connect loyalty-program data to Merchant Center so Google can surface member-specific pricing and benefits. Members of Minted’s Minted More program, for example, can see their member pricing when searching for holiday cards.

Why we care. These updates give brands new ways to measure their visibility in AI results, answer product questions inside YouTube ads, and reduce friction between discovery and checkout. They also make product-feed quality increasingly important, as richer data can influence how products are surfaced and recommended in AI-powered shopping experiences.

The big picture. Google’s commerce strategy is expanding beyond getting products indexed and advertised. It increasingly wants retailers to make their products understandable, discoverable, and ultimately purchasable by AI agents.

What to watch. The holiday shopping season will provide a test of whether conversational commerce can move shoppers from product discovery to checkout — and whether retailers see enough incremental sales to justify investing in richer product data and agentic-commerce integrations.

Dig deeper. Boost your holiday sales with these agentic commerce updates ]]> Google expands agentic commerce tools ahead of the holiday shopping rush Google’s agentic commerce updates give advertisers new ways to boost AI visibility, engage shoppers and turn discovery into sales. News https://searchengineland.com/competitor-negative-reviews-improve-ppc-ads-488715 Wed, 16 Sep 2026 14:00:00 +0000 https://searchengineland.com/?p=488715 What customers complain about can reveal what they value. Use those insights to find stronger PPC messages to test.]]>

Most PPC competitor research focuses on the same things: keywords, ads, landing pages, and offers. But there’s another source of competitive intelligence that can tell you something those tools can’t: what their customers wish they did better.

I see every negative review as free customer research. Someone has already explained what they expected, what disappointed them, and, in many cases, what would have changed their experience.

If you read enough reviews, patterns start to appear. Maybe customers repeatedly complain about slow responses, unexpected fees, complicated booking processes, or not knowing what happens next.

Those complaints can help you understand what customers value before they click your ad and where your competitors may be leaving opportunities open. AI simply makes those patterns much faster to find.

Find recurring complaint themes with AI

Collect around 30 to 50 negative or mixed reviews from two or three of your main competitors. These could come from Google, Yelp, Tripadvisor, Booking.com, or another relevant review platform.

Simply copy and paste the reviews into your AI prompt. It can usually work with raw text.

I would focus primarily on one-, two-, and three-star reviews. You can use any AI tool for this. My go-to is Claude.

Try this prompt:

• “Group these reviews into recurring complaint themes. Return a table with the theme, number of mentions, percentage of reviews mentioning it, and one representative example.”

You might end up with:

• 12 complaints about slow response times.
• 9 complaints about unexpected costs.
• 7 complaints about poor communication.
• 6 complaints about a complicated booking process.
• 4 complaints about limited availability.

One complaint could be an isolated experience. If the same frustration appears repeatedly across several competitors, you may have found a broader customer expectation that the market isn’t meeting.

But don’t stop at identifying complaints. Ask what customers wanted instead.

Slow responses may mean customers value certainty. Unexpected costs may point to demand for transparent pricing. Poor communication may show that customers want to understand what happens next.

You don’t want to build ads around your competitors’ weaknesses. You want to build them around what customers value.

Identify the gap between complaints and your offer

Once you have the recurring themes, compare them with what your own business actually does.

Try this prompt:

• “Based on these reviews, what do customers consistently want that these businesses don’t appear to provide?”

Competitor reviews may repeatedly mention waiting days for a response, unexpected prices, or having to call multiple times for an update.

Now imagine your business already responds within 30 minutes, provides fixed-price quotes, and automatically keeps customers updated. Those operational details could be reasons to choose you.

And yet businesses often leave their strongest differentiators out of their PPC ads and fall back on the same generic messages as everyone else: “Trusted local experts,” “Quality service,” “Get a free quote.”

If your competitors could use the same message, it probably isn’t communicating much of a competitive advantage.

Don’t overlook your own reviews, either. One of my favorite prompts for this exercise is:

• “Here are my reviews and my competitors’ reviews. Compare them and identify my biggest competitive advantages and anything I’m not communicating clearly enough.”

You may discover that customers consistently praise something you barely mention in your ads. If competitors’ customers complain about that same thing, you’ve found a much more interesting message to test.

Dig deeper: How to write paid search ads that outperform your competitors

Audit your operations before you advertise

Warning: Never turn a pattern in competitor reviews into an ad claim your business can’t consistently deliver.

If competitor research reveals that fast replies are a clear market gap, don’t write “Replies within 30 minutes” until you’ve explicitly confirmed with your operations team that your actual lead response time supports it.

Making a promise your business can’t consistently keep will only create frustrated customers and potentially negative reviews of your own.

AI can identify the opportunity, but you need to make sure your business can actually deliver before turning it into PPC copy.

Turn customer pain points into PPC ad copy

Once you’ve validated a genuine operational advantage, test whether making it prominent improves PPC performance.

Instead of filling every headline in a Responsive Search Ad (RSA) with minor keyword variations, balance keyword relevance with competitor gap messaging.

If you want AI to help generate ad assets based on your research findings, try this prompt:

• “Based on the complaint themes identified, generate five RSA headline options (max. 30 characters) that highlight our validated strengths and address these customer frustrations.”

Dig deeper: The anatomy of compelling search ad copy

Match the message on your landing page

Changing the ad isn’t enough. If your ad promises transparent pricing but the landing page never explains how pricing works, you’ve created another gap.

Take the strongest themes from your research and ask whether visitors can immediately see the answer to those concerns.

If response time matters, explain when they’ll hear back. If hidden costs are a common frustration, explain your pricing process. If customers are worried about what happens after submitting a form, show the next steps. For example:

Get your quote today!

• Tell us what you need.
• We’ll review your request within 30 minutes.
• You’ll receive a fixed-price quote before any work begins.

You’ve now addressed three possible objections before the visitor even has to ask.

For lead generation campaigns, reducing uncertainty about what happens after someone submits a form may be just as important as improving the form itself.

Use review research to create PPC tests

The biggest mistake would be treating AI’s output as the final answer. Treat it as a source of hypotheses instead.

If reviews suggest customers care about price transparency, test price-transparency messaging. If they complain about response times, test speed. If communication is the recurring frustration, test messaging that explains the process.

Test 1: Ad messaging

• Control: “Trusted Local Plumbers | Get a Free Quote”
• Variant: “Fixed-Price Plumbing Quotes | No Surprise Charges”

Test 2: Landing page

• Control: Standard service page.
• Variant: Add pricing transparency and a clear “what happens next” section.

Test 3: Offer

• Control: “Request a quote.”
• Variant: “Get your fixed-price quote.”

Now you’re using competitor research to create testable PPC hypotheses rather than copying what competitors are already doing.

Give each variant enough time to collect a meaningful sample before judging results. CTR can tell you whether the message attracts attention. Conversion rate tells you whether it persuades. Qualified lead rate tells you whether it attracts the right customers.

A higher CTR isn’t necessarily a win if those clicks produce worse leads.

Dig deeper: Why PPC tests in 2026 call for nuance, not winners

Your next PPC idea may already be sitting in a review

Reviews show you what people expected, what frustrated them, and what they wish businesses in your market did differently.

AI makes it easier to turn those comments into patterns. The valuable part comes afterward: comparing them with what your business genuinely does better and turning those opportunities into PPC experiments.

Sometimes the biggest opportunity isn’t creating a new advantage. It’s realizing that you already have one and making sure your next customer can see it before they click.

Your schema tells search engines what your brand is. But that doesn’t mean Google’s systems understand it the same way.

That gap can reveal where your content strategy is falling short.

Our next SMX Now today (Sept. 16) at 1 p.m. ET will feature Ray Martinez, VP of SEO at Archer Education. He’ll show you how to measure the difference between the entities you define and the ones Google’s natural language processing (NLP) actually recognizes.

You’ll learn how to build an entity audit using schema.org markup, the Google Cloud Natural Language API, and an agentic coding tool such as Antigravity, Claude Code, or Codex. The process turns your existing schema into a queryable knowledge graph, then compares it with competitor content to uncover topics they cover, gaps they miss, and entities Google doesn’t yet connect with your brand.

Martinez will also show you how to turn those findings into action. You’ll learn how to create content around under-recognized entities, strengthen those connections with structured data and internal data, and make your pages more retrievable and citable across search and AI engines.

You’ll leave with a repeatable way to measure discoverability, find content opportunities, and track whether Google and AI systems are getting better at understanding what your brand actually does.

Save your spot

]]> https://searchengineland.com/ecommerce-search-trends-that-matter-488725 Wed, 16 Sep 2026 13:00:00 +0000 https://searchengineland.com/?p=488725 Some search shifts are already affecting ecommerce performance. Others remain mostly hype. Here's what the latest data reveals.]]>

The way people search for products is changing fast, and it’s not always clear which shifts actually matter. In product search, ecommerce brands are being pulled in a dozen directions at once, trying to figure out where to focus.

Some of these changes are already reshaping how customers find and buy products. Others are getting a lot of hype but haven’t moved the needle much yet.

Here are nine ecommerce SEO trends worth paying attention to in 2027, and what they could mean for your traffic and revenue.

1. Consumers are increasingly relying on Google search to discover products

There have been a lot of headlines that Google search and traditional SEO are dying.

To find out what’s actually happening in Google search, I analyzed the search volume data for 2,000 “transactional” keywords in Ahrefs. These are high-purchase-intent queries someone would search in Google when looking to buy a product (protein powder, air fryer, microwave oven, weighted blanket, etc.).

As you can see, Google Search continues to grow year after year. In the last 12 months alone, search volume is up 2.2%.

So it doesn’t appear that the rise of ChatGPT, Claude, and other AI search engines has slowed down the growth of Google search. People are still using Google to search for products at higher rates than ever before.

2. Traditional SEO is still a major revenue driver for ecommerce

But what about actual traffic and revenue? Yes, Google search volume is increasing, but aren’t click-through rates falling?

To answer this question, I pulled Google Analytics data from a panel of 75 ecommerce stores for the past 12 months. These are ecommerce stores mostly doing between $1 million and $20 million in annual revenue.

I then cross-referenced the organic search data in Google Analytics with Google Search Console to see the actual breakdown between branded and non-branded search queries.

Here’s the traffic and revenue across the main traffic channels for the past 12 months, sorted from highest revenue to lowest.

Based on the data, non-branded organic search is still one of the top-performing revenue channels for ecommerce, second only to paid search.

Non-branded organic search is directly impacted by traditional SEO efforts, indicating that SEO is still a major traffic and revenue driver for many ecommerce stores.

Dig deeper: 7 organic content investments that drive ecommerce ROI

3. Increasing AI Overviews for ecommerce queries

Google first added “AI Overviews” to the Google search results on May 14, 2024.

Initially, ecommerce brands were relatively immune to these AI Overviews since they were primarily appearing for informational queries, things like “how to clean a weighted blanket” or “weighted blanket benefits.”

But a lot has changed in the past few months, and these AI Overviews are now creeping into shopping searches.

• November 11, 2025 – AI Overviews appear in 1.2% of shopping queries.
• March 16, 2026 – AI Overviews appear in 14% of shopping queries.
• September 3, 2026 – AI Overviews appear in 24% of shopping queries (fresh data from Ahrefs).

Using Amazon.com as a proxy, we can see that the rapid rise in AI Overviews across shopping queries began in early November 2025, and then spiked up again in late May 2026.

However, the presence of AI Overviews has started to decline in the past few weeks, indicating that Google may be rolling back AI Overviews for some product search queries.

It’s safe to assume that every time an AI Overview is present in the search results, clicks to the “10 blue links” will likely decrease.

So if you’ve noticed a drop-off in organic search traffic despite growing keyword rankings, it’s very possible that Google has started rolling out AI Overviews for some of your top keywords.

4. AI Mode is everywhere in Google’s products, but still has very little adoption

You may have noticed that Google has been pushing AI Mode more and more over the past year or so.

As of now, there are buttons for AI Mode on both the Google homepage and Chrome search bar.

But with AI Mode becoming more prominent within Google products, what does adoption actually look like?

A Datos study of more than 10 million desktop users found that AI Mode adoption is only about 0.13% of web traffic, down from its peak of 0.16% in March. To put this in perspective, AI Mode has about 1.3% of the usage of the standard Google search experience.

AI Mode still has a way to go before it becomes a meaningful traffic source within search. The recent downtick indicates that Google users are likely relying more on AI Overviews within Google search results instead of using the dedicated “AI Mode” experience.

However, in the EU and UK, AI Mode has continued to grow at a steady rate since late 2025.

With AI Mode becoming more prominent within Google’s products, some people have been wondering if AI Mode will ever become the default search experience. Google added some clarity to this back in May, stating that there are no plans to make AI Mode the default experience and that the 10 blue links aren’t going away any time soon.

Rather than pushing AI Mode even further and making it the default search experience, the adoption data suggests that Google may instead focus more on rolling out AI Overviews across more search results. After all, it’s much easier to bring AI to where the user is already searching than to push the user to a new product like AI Mode.

For the foreseeable future, we can expect traditional Google Search to remain a major driver of traffic and revenue for most ecommerce websites. But this is one trend definitely worth keeping a close eye on since AI search is changing rapidly.

Dig deeper: Why ecommerce SEO audits fail – and what actually works in 30 days

5. Major AI assistants are gaining popularity, but the adoption rate is slowing down

There have been many headlines about AI assistants like ChatGPT disrupting Google Search. The same data study by Datos looked into this by examining the adoption rates of other popular AI assistants.

Here are a few interesting takeaways from the data.

• ChatGPT is the clear leader across all major AI assistants, but Gemini and Claude are catching up. ChatGPT had been steadily growing through Q4 2025, but usage seems to be flattening out, or possibly even declining.
• Claude adoption spiked in early 2026, just a few months after the release of Opus 4.5, and the growth has continued over the past few months.
• Gemini has been steadily gaining traction over the past few years, although growth may be slowing down as of mid-2026.

Other AI assistants, like Copilot and DeepSeek, see very little use in the United States. Their adoption seems to be declining, especially as Gemini and Claude gain in popularity.

Takeaway: Don’t worry about your presence on Claude, Perplexity, and other AI assistants when it comes to AI search. Realistically, Google and ChatGPT are the only search platforms with meaningful adoption.

6. Traffic and revenue from AI assistants are still very low

Across the same panel of 75 ecommerce sites, only about 1.6% of total search traffic and revenue is coming from AI assistants.

However, the big unknown is how much of the non-branded organic search is actually coming from AI Overviews instead of the “10 blue links” on page 1. Unfortunately, Google doesn’t provide a way to break this out in Google Analytics or Google Search Console, so it’s difficult to know for sure how much traffic and revenue are originating from AI Overviews.

In either case, it’s clear that organic Google search is still driving much more traffic and revenue than AI assistants like ChatGPT.

Dig deeper: How ecommerce brands actually get discovered in AI search

7. AI traffic converts much better than organic search

Using that same panel of 75 ecommerce sites, I also analyzed the conversion data of AI search vs. non-branded organic search.

Over the past six months, traffic from AI assistants has a 19% higher conversion rate than non-branded organic search (1.82% vs. 1.53%).

Even if AI search is sending less traffic than organic search, the traffic is more qualified and tends to convert better.

This is likely because AI assistants are very effective at condensing the research phase from hours to minutes. 

Rather than having to read comparison articles, product reviews, and feedback on Reddit, the AI assistants can prequalify the products with helpful product recommendations.

Dig deeper: How to make ecommerce product pages work in an AI-first world

8. Organic Google Shopping is a major revenue driver

Google Shopping is often overlooked as a traffic and revenue driver in ecommerce.

By separating organic Shopping from organic search, we can clearly see that it’s a major traffic and revenue driver for many ecommerce brands.

Organic Shopping sends a disproportionate amount of revenue when compared to its relative traffic. This is because organic Shopping typically has a much higher conversion rate than both non-branded organic search and AI referrals.

Our sample of 75 ecommerce brands showed that organic Shopping converted 60% higher than non-branded organic search.

• Non-branded organic search: 1.53% conversion rate
• Organic Shopping: 2.4% conversion rate

The massive increase in organic Shopping conversion rate suggests that many searchers may prefer purchasing products from the Google Shopping product grids instead of the standard “10 blue links” on page 1.

Don’t sleep on organic Shopping since it’s already a very strong revenue driver for many ecommerce brands. It can be helpful to understand the primary signals to better optimize your products for Google Shopping.

9. Agentic commerce hasn’t gained any real traction

There’s been a lot of talk about agentic shopping in the world of AI search, with mentions of technical protocols that allow for instant checkout within AI search. Two of them matter.

• Agentic Commerce Protocol (ACP): Co-developed by OpenAI and Stripe, ACP powered ChatGPT’s “Instant Checkout.” However, OpenAI ended Instant Checkout in March 2026 after only about 30 merchants ever went live. Now the official ACP documentation on OpenAI’s website has been reduced to guidance on traditional shopping feeds instead of the complete agentic checkouts and payments it once covered.
• Universal Commerce Protocol (UCP): Led by Google and Shopify to enable instant purchases within AI Mode and Gemini. Since February 2026, U.S. shoppers have been able to buy from a small group of retailers inside AI Mode and Gemini without leaving Google. So should you add “agentic shopping” to your to-do list?

Should you add “agentic shopping” to your to-do list and start adding these protocols to your ecommerce website? Probably not. Or at least not yet.

There have been only a few sightings of instant checkout buttons in AI assistants, and they’ve been limited to major online retailers like Wayfair, Etsy, Target, and Home Depot.

So far, I’ve yet to see or hear about instant checkout actually live for ecommerce brands on Shopify, Magento, BigCommerce, and other major ecommerce platforms.

Dig deeper: Why agentic commerce will matter more than ChatGPT ads

What these trends mean for 2027

Search platforms will keep changing, and there will always be hype around new trends that claim to change everything overnight. Most won’t, but a few will.

The ecommerce brands that come out ahead aren’t the ones that jump on every new trend or guess right every time. They’re the ones paying close enough attention to identify a real shift from a passing headline and refocusing only once it’s proven to have a meaningful impact on sales.

]]> image image image image image image image image image image image image image https://searchengineland.com/building-software-doing-marketing-488688 Wed, 16 Sep 2026 12:00:00 +0000 https://searchengineland.com/?p=488688 Your team doesn't need to build every AI workflow. Decide what to automate in-house, buy from a vendor, or outsource. ]]>

The most valuable 30 minutes in marketing right now are the 30 minutes you spend thinking before you start building any AI automation.

Last week, I made the case that AI does not delete marketing work, it moves it, and that the bill often gets paid out of the least attributable work your team does. Prompting, checking, editing, fixing, and maintaining are all real work hours, and they come from somewhere.

The question now is what to do about it.

All of the work your marketing team does with AI comes from three places:

• AI-powered software/tooling from a vendor.
• Expert customization and implementation from a consultant who has already built it.
• Time from your team to build it in-house.

Amanda and I work separate client rosters, but the same pattern is showing up in both: Teams aren’t sorting their AI work hours in a targeted manner. Instead, they’re going straight to automating as much as they can in-house.

View embedded content

Automating in-house is smart. But skipping the sorting step (which takes 30 minutes or less), is an expensive mistake. Thankfully, it’s an easy fix.

1. Buy the tool when the problem isn’t just yours

Rank tracking, citation monitoring, brand mention tracking, crawl diagnostics, and content scoring are not unique to you.

Despite $30-40 billion in enterprise investment into genAI, MIT’s review of enterprise AI projects reveals that 95% of organizations are getting zero return. From the report:

• “Strategic partnerships achieved a significantly higher share of successful deployments than internal development efforts. While we observed far more BUILD initiatives than BUY initiatives in our sample, with many more organizations exploring internal development… success rates favored external partnerships. Though we lack precise data on total initiative volumes, the pattern suggests that internal development efforts have substantially lower success rates despite being more commonly attempted.”

The missing 1% from the chart above is MIT’s third structure, hybrid build-buy, where an internal team co-develops with a vendor; the report had too little data to quantify it.

The report found that the main reason the projects fail is that the tools do not learn or integrate with how people already work. Making a tool fit is one of the most expensive parts, and a software vendor who’s already done that for 4,000 customers has done work you really don’t need to repeat.

When you’re deciding whether to build versus buy, you’re evaluating the software platform fee versus what an hour of your team’s time really costs, multiplied by the hours it takes to build and maintain the internal version. (Plus the hidden hours nobody is counting: brainstorming, meeting, testing, reiterating, retesting.)

Buy the tool when the problem is common. When you do, your team also gets two things you can’t build for yourself: the vendor maintains the software, and somebody on the other end fixes it when it breaks.

2. Buy the know-how when the workflow is yours alone

Some workflows really are yours, like who signs off on the work before it’s shipped, how your data is structured, when reporting goes out, how you gather SME input for content, or adding branded CTAs from a pre-approved copy bank.

This is where your team can accidentally burn the most hours, because “nobody else can build this” gets heard as “we have to build this ourselves.”

The workflow is yours, yes, but someone who has built 65 of these for 20 different teams already knows which steps usually break, which ones need input from your individual contributors, which ones are worth automating, and which ones look automatable but aren’t. 

Your team is finding all of that out for the first time, on the clock.

Here’s the test: If the hours you are about to spend produce knowledge you will not need weekly, contract out the knowledge instead.

We both sell consulting individually, so you can read that with the appropriate squint. The point stands true without hiring either of us: Buying the know-how looks like a template someone already validated, a contractor for 4 weeks, a consulting hour with a peer who has shipped the same thing, or a resource library built by people who made the mistakes first. 

What it avoids: Your team discovering the failure modes one at a time.

An AI tool you don’t have the experience to check will hand you wrong answers confidently

Only 13% of marketers fully trust AI output without a human reading it. While that often gets treated as a model maturity problem or AI adoption hurdle, it’s actually a staffing requirement. 

I (Kevin) ran into a situation where I was trusting my own AI workflow output slightly too much. When I paused and looked at it through a refreshed critical lens, I found a pile of things to correct and reoriented the workflow accordingly.  

And I (Amanda) had a case where my client generated a report from GSC data and sent it to me for feedback before bringing it into their weekly meeting. The model read a spike in a specific set of queries as proof of rising AI visibility and was glaringly wrong; the report was nearly on its way to the head of marketing to inform crucial decisions. We caught it. Now every report gets a human read before it leaves. 

The State of CRM Data Report 2026 found the following:

• “Nearly 78% of C-suite and 92% of SVP/VPs respondents say they have acted on an AI recommendation they later suspected was wrong because of bad underlying data, compared with 41% of individual contributors.”

So if the majority of inputs and outputs need review (and perhaps, even an intervention before an output lands in the CMO’s inbox), someone who knows the subject has to do the reviewing, which means the tool did not remove the skill required to do the work. Judging the work is the harder, less-replicable of the tasks.

The rule that comes out of both experiences: Never buy or build a tool for a job nobody on your team can verify accuracy for by hand. You won’t be able to tell when it breaks (and it will break confidently).

Swap out one step, not one job

With that in mind, this section is the actual method to use when making the build vs. buy decision, and it’s the part teams can get wrong even if they sort the work correctly.

Where failure can creep in at the start: A team names a slow process, decides AI should run the whole thing, and starts building something that does 15 steps. Six weeks later they have in-house software with no real tests, no documentation, and one person who understands how to maintain it, with maybe 2-3 people who know how to run it.

A job is a bundle of steps with judgment distributed across all of them. 

A step has one input, one output, and a check that takes seconds.

Part 1 made the point that every workflow you build becomes a small permanent job to maintain. That’s true. And it’s an argument for making the job small, not for skipping it.

A one-step swap has one input to validate, one output to eyeball, and one thing to fix when the model version changes, while your 15-step workflow has 15 places to break and no fast way to find out which one did.

How to find the step worth automating

Take the process your team wants to automate and write out what actually happens, in order. Then mark each step with three things:

• The input: What arrives, and where does it come from? A step whose input is “context from the last meeting” is not a step yet.
• The output: What leaves, in what format? If the answer is a paragraph of judgment, keep looking.
• The check: How does a person confirm the output is right, and how long does that take? The verification check is real work and takes real time; plan for it.

The step you want to automate is the one that is slow, repetitive, tightly defined, and checkable at a glance. 

Here’s an example: I built the new premium members’ area to include any of the new tools and skills we’re building, but I automated it to go back into the archives to locate each prior premium asset and build an easy-to-find page for it. 

• The input: Historical Growth Memo posts, specifically gated sections.
• The output: A simple, easily searchable page with a link to the asset and notes on how to use it, along with an area for members to leave feedback/comments for admin.
• The check: A 5-minute review before setting the page live.

Usually, an automatable step looks like a “small transformation,” raw data into a formatted table, a transcript into tagged quotes or a first draft, internal linking across content clusters, a spreadsheet into a brief skeleton, a GSC export into a list of pages to refresh.

What disqualifies a step from being automated

I (Kevin) would say there are three types of jobs that people do that people try to automate:

• Reporting (easy to automate).
• Synthesizing (hard to automate).
• Deciding (semi-hard to automate).

Anything requiring experienced taste, or any automated task where being wrong could go unnoticed for three months (like automating regular updates to meta titles and descriptions, checked quarterly), is usually an indicator that careful consideration is needed. 

Use these guidelines to disqualify an in-house build:

• Nobody on the team can do it “by hand,” so nobody can grade or verify the output.
• The check takes longer than doing the work (unless you’re producing long-form drafts, which naturally take time for review).
• The input changes often, which means you can’t easily maintain the workflow.
• It touches a system you do not control, where an update on a Tuesday could break the workflow on a Wednesday.

Experimenting earns its keep. But your common problems should go to a vendor; there’s a reason why they exist. Problems that are yours and that your team deeply understands can go to someone who has built something similar before. 

And the stuff that’s left… that’s the actual in-house AI workflow experimentation. Give it an owner, a kill date, and a sentence describing what winning looks like.

Everything else is software development, and, well, you didn’t hire your marketing team to write software.

This post first appeared on the author’s website and is republished here with permission. ]]> ai-work-sourcing-matrix Bought AI tools reached deployment twice as often as in-house builds Productivity gains from automation are real https://searchengineland.com/matt-van-wagner-shares-whats-changed-what-disappeared-and-what-ppc-still-gets-wrong-488765 Tue, 15 Sep 2026 17:52:13 +0000 https://searchengineland.com/?p=488765 Paid search veteran Matt Van Wagner reflects on 20 years of Google Ads, lost PPC tactics, rising automation and more]]>

The search marketing industry has gone from an entrepreneurial “gold rush” to an era dominated by automation, AI and increasingly consolidated platforms.

Looking back over that transformation, paid search veteran Matt Van Wagner reflected on the early days of Google Ads, the rise of Search Engine Land, tactics that thankfully disappeared and why advertisers still need to question what Google tells them.

Search marketing started with a gold-rush mentality

Van Wagner entered search in the early 2000s after working in technology sales and marketing, where traditional channels such as trade publications, conferences and direct mail were becoming increasingly ineffective.

His first search conference offered a stark contrast: an industry filled with entrepreneurial energy and people exploring an entirely new way for businesses to reach customers.

That experience ultimately convinced him to leave his existing career and start a search advertising agency — at a time when few businesses even knew what a search advertising agency was.

Selling Google Ads was the easy part

Once advertisers understood what paid search could do, Van Wagner said convincing companies to try Google advertising required little selling.

Demand was plentiful and the emerging agencies were generally collaborative because the market was expanding quickly enough for everyone.

The bigger mistake, he said, was that many early agencies simply didn’t charge enough for the value they were providing.

Search Engine Land helped the industry grow up

By the time Search Engine Land and SMX arrived, search marketing conferences were beginning to suffer from repetitive speakers and presentations.

SMX Advanced took a different approach: rather than repeatedly teaching the fundamentals, it started at an intermediate level and went deeper.

That attracted more sophisticated speakers and discussions, while conversations in bars, over coffee and at dinner became almost as valuable as the formal sessions themselves.

PPC tactics became far less granular

Some of the practices considered sophisticated in early paid search now look unnecessarily complicated.

Single keyword ad groups, duplicated match types and elaborate negative-keyword structures were routinely used to tightly sculpt traffic between campaigns.

Automation has fundamentally changed that approach, making many of those old structures largely obsolete.

Early Google felt more open

One thing Van Wagner misses about the early search industry is the accessibility of Google itself.

Events such as the Google Dance gave marketers opportunities to interact directly with people working on search, while figures such as Matt Cutts became a bridge between Google and the SEO community.

Part of that openness came from the fact that Google itself was still figuring search out. Advertisers, SEOs and search engines were effectively exploring the new ecosystem together.

Google’s definition of success isn’t necessarily yours

Advertisers want profitable sales and measurable business outcomes. Google also wants advertisers to spend money.

Van Wagner argued that marketers need to remember that distinction when evaluating recommendations, default settings and increasingly automated campaign types.

Google may genuinely want its technology to improve advertiser performance, but its commercial incentives and an advertiser’s objectives aren’t automatically identical.

Broad match defaults remain a major frustration

One longstanding example is Google’s use of broad match as a default.

Van Wagner argued that inexperienced advertisers can easily spend money learning lessons Google’s systems already have enough information to help them avoid.

His broader concern is that advertisers increasingly have to spend their own money teaching automated systems things they already know about their businesses.

Automation still needs advertiser knowledge

Van Wagner accepts that automation is moving in one direction and advertisers won’t regain every control they once had.

But that doesn’t mean Google should ignore knowledge advertisers already possess.

If an advertiser already knows particular queries are irrelevant, for example, forcing an AI system to rediscover that through wasted spend makes little sense. Automation should be able to learn while still incorporating the advertiser’s existing expertise.

Ginny Marvin helped bridge Google and advertisers

Van Wagner highlighted Google Ads Liaison Ginny Marvin as a positive influence on the relationship between Google and the advertising community.

Having worked on the advertiser side before joining Google, Marvin understands many of the frustrations marketers experience and, he argued, has been able to advocate for advertisers while still representing Google.

Bidding your way to No. 1 was an expensive mistake

One misconception from early PPC was that advertisers simply needed to outbid everyone to secure the top position.

Van Wagner recalled campaigns where advertisers were willing to pay enormous CPCs purely because they wanted to own the No. 1 spot.

The better philosophy, he argued, was often to compete for second place and focus on economics rather than prestige.

Google Instant showed the power of product publicity

One Search Engine Land article Van Wagner particularly remembers writing examined Google Instant, which began predicting and displaying search results as users typed.

His argument was that Google generated enormous media attention from the product change while Microsoft was spending heavily on conventional advertising to promote Bing.

The episode demonstrated just how powerful Google’s ability to turn product changes into mainstream news had become.

Microsoft realized it couldn’t reinvent search behavior

Microsoft faced a different problem: users weren’t actively deciding which search engine to use every morning.

Van Wagner believes Microsoft therefore made a sensible decision to make its advertising platform familiar to Google advertisers rather than forcing them to learn an entirely different system.

That reduced friction for marketers running campaigns across both platforms, even if it made differentiation more difficult.

Google’s Florida update changed SEO overnight

For Van Wagner, Google’s Florida update was the moment that exposed the risks of relying too heavily on organic search.

The update hit shortly before the holiday season and wiped out rankings for businesses and agencies that had built their strategies around SEO techniques that previously worked.

Some of those techniques were aggressive, but the episode demonstrated how quickly Google could change the rules — and how much revenue could disappear with them.

Paid search became an SEO insurance policy

Florida also helped reinforce the value of paid search.

Organic visibility could disappear following an algorithm update, while PPC gave advertisers another way to maintain traffic and revenue.

For Van Wagner, that uncertainty was one reason he ultimately specialized in paid search rather than continuing to build an SEO business.

Search grew through collaboration, not secrecy

The early search community was unusually willing to share information.

Conference speakers made themselves available after sessions, practitioners exchanged observations when platforms behaved unexpectedly, and competitors frequently helped one another.

Van Wagner attributes much of that generosity to the rapidly expanding market: there was enough opportunity that practitioners didn’t feel they needed to guard every insight from competitors.

Conferences needed to make room for new voices

As search matured, its conference circuit risked becoming dominated by the same established names.

When Van Wagner became involved with speaker selection for SMX Advanced, he tried to make the process more focused on ideas by removing identifying details such as names, genders and companies from submissions during the initial evaluation.

The goal was to judge the strength of an idea before the reputation of the person presenting it — helping new speakers break into an industry that had become increasingly difficult to enter.

Independent search journalism still matters

Asked what would disappear without Search Engine Land, Van Wagner pointed to its independent editorial voice.

For practitioners trying to understand constant platform changes, there remains a difference between advertiser-driven messaging and reporting that can independently question what Google, Microsoft and other platforms are doing.

That editorial independence, he argued, is what has made industry publications valuable over the long term.

PPC professionals need to admit mistakes faster

Van Wagner also acknowledged one uncomfortable PPC habit: finding creative explanations when something goes wrong.

He recalled accidentally applying “loan” as a broad-match negative in an expensive financial campaign and burning through roughly $7,000 in around four hours.

His lesson was that practitioners are often better served by simply acknowledging a mistake, explaining what happened and learning from it rather than looking for something — or someone — else to blame.

Never assume Google is acting in your best interests

One myth Van Wagner wants advertisers to abandon is the assumption that Google is automatically acting in their best interests.

That doesn’t mean Google representatives are dishonest. It means Google is a business with its own incentives, and advertisers need to independently evaluate recommendations rather than treating them as objective instructions.

The industry may have changed dramatically over two decades, but that principle has become more important as advertisers surrender more campaign decisions to algorithms.