Where Did My Clicks Go? How to Get ChatGPT and Perplexity to Recommend Your Products

The new question is not how to rank first, but how to become the answer.

If you run an online store and your organic clicks have been slipping over the past months, you are not imagining it. The game changed. For years, the obsession was simple: rank in the first position on Google, win the click, and turn that visit into a sale. Today, that journey is no longer linear.

More and more searches end without a visit to any website. The user asks, Google answers directly on the results page, or a tool like ChatGPT or Perplexity synthesizes the information and delivers a recommendation that is ready to act on. In other words, search stopped being only a gateway to websites and started becoming a destination of its own.

That shift completely changes how brands compete for visibility. It is no longer enough to “rank.” Now you need to be understandable to AI systems, trustworthy to outside sources, and useful enough to be cited or recommended when someone asks what to buy, which option is better, or which brand solves a specific problem best.

For many businesses, that sounds threatening. Fewer clicks usually get interpreted as fewer opportunities. But that reading is incomplete. What is happening is not just a loss of traffic. It is a redistribution of influence. Attention is moving from the traditional search results page into AI-generated answers. And that creates a new opportunity for brands that understand how to become part of those answers.

The good news is that ChatGPT and Perplexity do not recommend brands at random. They do not guess. They rely on clear signals, both from your own website and from the wider web. They evaluate how you describe your products, how easy they are to understand, how well you answer concrete questions, what reputation your brand has built across third parties, and what people say about you in reviews, forums, communities, and comparison content.

In this article, we will break that shift down in practical terms for business owners and eCommerce decision-makers. You will see why the zero-click trend is reshaping SEO, how AI brand recommendations actually work, what to fix on your product pages to gain visibility, and why AI traffic may bring fewer raw visits but often delivers visitors who are far more ready to buy.

If it feels like your clicks disappeared, what you are really seeing is the rise of a new product discovery channel. The right question is no longer “how do I recover lost traffic?” but “how do I get AI to choose me when the customer is ready to decide?”

The zero-click reality: when search ends before your site begins

Zero-click search has become one of the most important concepts in modern digital marketing. It refers to searches that are resolved directly on the search results page, without the user needing to click through to any website. This is not a technical nuance. It is a structural transformation.

Multiple market analyses show that a large share of queries now end without a click. Depending on the study, device, and query type, that share falls roughly in the 60% to 83% range. The message is clear: millions of people now get enough of an answer without ever visiting a site.

Why is this happening? Because platforms are reducing friction. Google answers with knowledge panels, featured snippets, rich results, merchant experiences, and AI Overviews. ChatGPT and Perplexity summarize sources, condense comparisons, and provide natural-language recommendations. From the user’s perspective, this is faster, easier, and often more satisfying.

From a brand’s perspective, it can feel like a direct hit to organic traffic. And in many cases, it is. Recent studies have found meaningful CTR declines for traditional organic listings when AI Overviews appear for certain searches. Even ranking first no longer guarantees the click volume that once justified years of SEO investment.

But here is the strategic twist many businesses still have not fully absorbed: in a zero-click environment, being present inside the answer can be more valuable than simply appearing at the top of a list of links.

If your brand is cited, summarized, or recommended by an AI answer, you are entering the exact moment when the buyer is forming an opinion. You are no longer competing only for traffic. You are competing for influence. And influence happens before the click.

Take a simple example. Someone searches for “best ergonomic chair for a small home office.” In the past, they might have opened six or seven tabs. Today, they might read an AI Overview or ask Perplexity and receive a short list with pros, cons, and suggested options. If your product appears there, you have already passed the hardest filter: initial consideration.

That means the value of visibility has changed. Success can no longer be measured only in sessions. Brands now need to ask:

• Does my brand appear in AI-generated answers?
• Can AI systems understand my products accurately?
• When an AI compares options, does it find enough context to choose me?
• Does my off-site reputation support that choice?

This shift also forces teams to rethink reporting. If your dashboards still judge performance only by organic clicks, you may conclude that the strategy is weakening when the brand may actually be gaining visibility in a newer, harder-to-measure discovery layer. Traditional attribution often misses what happens before the click.

In other words, disappearing clicks do not always mean disappearing demand. Sometimes they mean the buying decision started earlier, inside an interface that summarizes and recommends.

Why being cited inside an AI answer can matter more than ranking #1

For years, the number one organic position was the ultimate SEO trophy. It still matters, of course. But it no longer owns attention. For many informational and comparison-heavy searches, users now consume an AI-generated answer first and only then decide whether to click, which result to click, or whether they even need to leave the page.

That makes a citation inside an AI Overview or a recommendation inside a generative answer disproportionately valuable. Your brand is not just visible. It is being selected as a useful source or a trustworthy option at a moment of very high intent. That is fundamentally different from just showing up on a search results page.

The difference is significant.

A traditional organic listing competes visually with ads, modules, rich snippets, shopping elements, and other blue links. A brand mentioned inside an AI answer enters a more curated context. It often appears alongside attributes, comparisons, summaries, and a recommendation framework. That lowers research effort for the shopper and increases trust faster.

Recent studies also show that while there is considerable overlap between pages ranking well in Google and pages cited in AI Overviews, the relationship is not perfect. In large-scale analyses of AI Overview citations, a meaningful share came from pages outside the top 10 and even from pages not ranking in the top 100. That is an important lesson for mid-sized brands: you do not need to dominate every SERP to gain AI visibility, but you do need to be exceptionally clear, useful, and trustworthy.

That changes the eCommerce playbook.

It is no longer just about pushing category pages and product pages for transactional keywords. It is also about creating data and content that help an AI answer questions such as:

• Which product is best for a specific need?
• What option makes sense based on budget, materials, or compatibility?
• Which brand has a stronger reputation in a category?
• Which product solves a common problem most effectively?

When AI answers these questions, it often does not “read” your site like a human. It interprets your presence through entities, structure, relationships, third-party mentions, and trust signals. If your digital footprint is not built for that, you can lose visibility even if your classic SEO looks decent.

What used to be a ranking problem is now also a machine readability and distributed reputation problem.

How ChatGPT and Perplexity recommend brands

A common misconception is that tools like ChatGPT and Perplexity operate like a black box that recommends brands without any observable logic. That belief leads to paralysis. If recommendations were random, there would be nothing to optimize. But they are not random.

When these systems answer product and category questions, they tend to rely on two broad families of signals:

• On-site signals, meaning the quality and clarity of the information on your own website
• Off-site signals, meaning the reputation and validation your brand builds elsewhere on the web

In practical terms, a useful way to think about this is an approximate split: around 30% of your brand’s recommendability comes from your own digital assets, while roughly 70% depends on external reputation signals, discussion, and third-party validation.

That is not a rigid mathematical formula, but it is an excellent strategic model. Many brands invest almost everything in their own website and almost nothing in their broader footprint. Then they wonder why AI does not mention them even though their product pages look polished. Usually, the answer sits in that gap.

The 30% on-site: what your website teaches AI

Your site still matters a lot. It is where you officially define who you are, what you sell, who it is for, which attributes your products have, how they are used, what they are compatible with, which problem they solve, and which objections they remove.

If that information is incomplete, vague, or written only to sound promotional, AI will struggle to recommend you with confidence.

The on-site assets that usually matter most include:

• Well-structured product pages
• Specific titles and descriptions
• Helpful, conversational FAQs
• Complete technical attributes
• Product structured data
• Visible and credible reviews
• Educational and comparative content around the category

The implication is direct: your eCommerce store cannot keep writing product pages only to “look nice.” It needs to build pages that answer real customer questions in clear language.

The 70% off-site: what the internet says about you when you are not talking

This is the part most businesses underestimate. AI systems place enormous weight on signals distributed across the web. Why? Because models and retrieval systems need to compare what a brand claims with what third parties observe or report.

That is why sources such as these show up so often:

• Third-party reviews
• Product comparisons
• Editorial publications
• Specialist communities
• Forums
• Reddit threads
• Videos and user commentary
• Mentions in industry media and niche blogs

Not every mention carries the same weight. Quality, consistency, and specificity matter. A detailed discussion in which real users explain why they chose a product may matter more than ten shallow mentions. A review with real usage context may matter more than a purely promotional article.

With Perplexity, this logic is especially visible because it often cites and synthesizes accessible web sources in real time. With ChatGPT, depending on the experience and integrations involved, recommendations may draw from prior model knowledge, current retrieval, and web-wide signals that are richly represented across the internet. In both cases, brands with a stronger reputation footprint are generally easier to recommend.

Distributed reputation is the new digital shelf

In physical retail, visibility used to be won on the shelf: position, packaging, signage, and pricing. In AI-mediated commerce, the shelf is no longer just your category page or your Google ranking. It is the collection of evidence an AI can find when deciding whether your brand deserves to be mentioned.

Put differently, the entire internet has become your shelf.

If people on Reddit recommend your product for a specific use case, if an independent publisher compares your materials favorably, if reviews consistently mention durability, if your brand appears in specialized roundups, and if your website clearly describes specifications, then AI has enough pieces to build a stronger recommendation.

If it cannot find those signals, it will fill the gap with other brands that have left a clearer footprint.

This reality forces store owners to look beyond the website. The work no longer ends in Shopify, Magento, or WooCommerce. It also includes:

• Active review generation and management
• Relationships with relevant publishers and creators
• Presence in the right communities
• Useful resources that third parties want to cite
• Post-purchase programs that encourage detailed feedback

The lesson is simple: for AI to trust your brand, the web needs to trust your brand first.

The kinds of questions shoppers ask AI before buying

One of the biggest mistakes brands make when optimizing for ChatGPT and Perplexity is thinking only in short keywords. In reality, users talk to these tools in a much more natural, specific, and contextual way. They do not just type “best coffee maker.” They ask things like:

• What is the best coffee maker for a small kitchen and daily use?
• Which backpack works for a 16-inch laptop and short trips?
• What supplement offers the best value for beginner runners?
• Which ergonomic chair do you recommend if I am under 5'3"?
• What sheets are best for hot weather and sensitive skin?

Each of these questions combines intent, context, and constraints. That is exactly what AI is trying to solve. It is not only looking for a popular product. It is looking for the right product for a specific situation.

That is why your product pages must speak the language of use cases, not just product names.

When you describe an item only with generic manufacturer text, you waste the opportunity to match these richer, buyer-style questions. But when you explain who it is for, who it is not for, which problem it solves, what its limits are, and what it is compatible with, your chances of being selected increase dramatically.

How to write product FAQs that actually help ChatGPT and Perplexity

One of the most underused assets in eCommerce is the FAQ section on the product page. Many stores treat it as filler or skip it entirely. In the age of AI-powered discovery, that is a mistake.

Why? Because FAQs translate commercial information into direct answers. And direct answers are exactly what generative systems need in order to understand, summarize, and recommend.

What makes an FAQ valuable for AI

A good FAQ does not exist to fill empty space or restate the obvious. It exists to anticipate real buyer questions in everyday language.

It should be:

• Specific
• Conversational
• Clear
• Useful
• Honest
• Grounded in real use cases

Instead of asking “Is this a high-quality product?” which says very little, it is far better to answer concrete questions like:

• Does this product work well in small spaces?
• Is it compatible with a specific device, material, or system?
• Which size should I choose if I am between two sizes?
• Can it be washed, installed, or assembled without tools?
• How long does it last under normal use?
• What exactly is included in the box?
• What is the difference versus the previous model?

How to write them better

The key is to write the way customers talk, not the way catalogs talk.

Instead of:

• “Made with premium materials and innovative design”

Write:

• “Made from aluminum and high-resistance silicone, so it stands up to daily use and is easy to clean.”

Instead of:

• “Compatible with multiple devices”

Write:

• “Compatible with 13- to 16-inch laptops and tablets with slim cases. Not recommended for thicker devices or oversized bases.”

Notice the difference. The second version removes ambiguity, adds context, and helps an AI map the real usefulness of the product more accurately.

A practical framework for stronger product FAQs

Use this structure for every key product:

1. Ideal use case: who it is for and when it works best
2. Compatibility: sizes, systems, connections, fits, or formats
3. Limitations: when it is not the best choice
4. Materials and durability: what it is made from and how it performs over time
5. Setup or care: how easy it is to use, assemble, or maintain
6. Comparison: how it differs from other models in your catalog
7. Purchase details: what is included, delivery expectations, exchanges, or returns if relevant

When this information is written well, it does not just improve user experience. It creates reusable knowledge blocks that AI systems can work with more effectively.

The structured attributes that help AI understand your products

Now we get into a layer many businesses overlook because it is not visually exciting, but it is critical: product data structure.

Search engines and systems that consume web information need more than descriptive copy. They need clear, consistent attributes that identify what the product is and what properties it has.

Google has been explicit about this through Merchant Center and its structured data guidance: structured markup and product attributes help systems retrieve up-to-date information and understand offers directly from your website.

For eCommerce, that means strong copy alone is not enough. Your product also needs to be machine-readable.

Attributes that add the most value

Depending on your category, some of the most important include:

• Brand
• Model
• SKU
• GTIN when applicable
• Price
• Availability
• Color
• Size or dimensions
• Materials
• Weight
• Compatibility
• Capacity
• Included accessories
• Product condition
• Variants

These data points allow AI systems to make much more precise connections when a buyer asks for a product that fits specific needs.

For example:

• “I need a case compatible with iPhone 15 Pro Max and MagSafe”
• “I want a folding table under 90 cm wide”
• “I’m looking for percale cotton sheets for hot weather”

If your site does not express those properties clearly, the AI has to infer them. And when a system has to infer too much, the risk of omission goes up.

Structured data is not optional if you want visibility

Using schema markup and well-maintained product structured data is one of the clearest ways to tell search and commerce systems: this is what I sell, this is what it costs, this is how it varies, and this is what makes it relevant.

For businesses, the upside is twofold:

• Better understanding by search engines and shopping ecosystems
• More opportunities to surface in rich experiences, comparisons, and AI-generated answers

Consistency also matters. What appears on the page, what exists in the product feed, and what appears in structured data should align. When they do not, visibility suffers and trust erodes.

For larger catalogs, this is often a massive opportunity. Many stores have long descriptions but weak, incomplete, or inconsistent attributes across product variants. That kind of disorder silently costs visibility.

How to optimize product pages for conversational buying questions

A product page designed for AI does not look like a generic landing page overloaded with adjectives. It looks more like a smart combination of technical sheet, helpful salesperson, and category advisor.

That means rebuilding product content around a conversational answer framework.

Core elements of an AI-ready product page

1. A specific, descriptive title

Avoid empty naming. The title should include the elements a buyer would use to identify the correct option: category, material, capacity, fit, or core use case.

2. An opening summary that answers “what is it?” and “who is it for?”

The first lines should explain practical value quickly.

3. Visible attributes, not buried ones

Do not force users or AI to parse a paragraph to figure out dimensions or materials. Put the information in bullet lists, tables, or clearly labeled blocks.

4. FAQs aligned with real objections

Include useful answers, not recycled marketing language.

5. Reviews with context

The most valuable reviews are not just five-star ratings. They explain why the product worked, for what type of buyer, and under what conditions.

6. Comparative content within your catalog

Help users understand differences among models. AI systems value pages that support comparison and decision-making.

7. Coherent images and supporting assets

Multiple angles, clear dimensions, and consistency with the written description all improve understanding and trust.

Off-site reputation: Reddit, reviews, and communities really do matter

When someone asks AI which brand to buy, they are not only looking for specifications. They are looking for confidence. And confidence is rarely built from a brand’s official voice alone.

That is why off-site signals matter so much. Reddit, specialist forums, third-party reviews, comparison sites, marketplaces, and niche media all act as validators. These are places where the brand stops talking about itself and starts being evaluated by others.

This helps explain why technically solid products sometimes lose to competitors with a better external conversation. If the open web contains more evidence, more useful opinions, and more context about another brand, the AI has more material to support that option.

What to do without manipulating the conversation

The answer is not to flood the internet with artificial mentions. That is usually fragile and unconvincing. What does work is building a reputation that is both earned and visible.

A few practical moves:

• Encourage detailed post-purchase reviews
• Ask customers to describe real-world use cases
• Identify communities where your category is genuinely discussed
• Create buying guides and comparison resources worth citing
• Work with specialists who add honest evaluation and context
• Monitor recurring product and category questions across support, reviews, and forums

The goal is not to control the conversation. It is to participate in it with enough usefulness that other people amplify it.

Fewer clicks, better visits: the quality of AI traffic

This is one of the most reassuring parts for store owners: even though traffic from AI tools is still a relatively small share of total traffic, the evidence increasingly suggests that it is often higher quality than traditional organic search traffic.

Several signals point in the same direction.

On one side, recent research shows that AI referral traffic is still low in absolute volume, but growing quickly. On the other, different studies are reporting stronger conversion rates or stronger intent signals from visitors arriving through generative AI experiences.

One especially striking example found that, for a studied website, just 0.5% of visitors from AI search generated 12.1% of all sign-ups. In that same analysis, AI search visitors converted far above traditional organic search visitors. Broader benchmarks have also identified year-over-year growth in AI-referred conversion rates and behavior that looks increasingly similar to high-intent channels like paid search.

What does that mean for your business? You should not evaluate this channel on raw traffic volume alone.

A visitor coming from ChatGPT or Perplexity often has already gone through a layer of mental filtering. They have already seen a summary, comparison, or recommendation. By the time they land on your product page, they are not arriving to discover the category from scratch. They are arriving to validate, confirm compatibility, verify the final price, or complete the purchase.

That changes what the session is supposed to do.

With traditional search traffic, a large part of the page’s job is to educate from the beginning. With AI traffic, some of that education has already happened before the click. That is why these visitors often arrive more qualified.

What this behavior looks like in practice

AI-referred visitors often show patterns such as:

• Less random browsing
• Faster movement toward the right product
• Greater focus on attributes, social proof, and pricing
• Higher conversion likelihood if the page confirms what the AI answer implied

That last point is the crucial one. If the landing experience does not match the expectation created by the AI answer, trust breaks. If it does match, conversion can accelerate.

How to adapt your eCommerce store for higher-intent AI traffic

If we accept that AI traffic may be more qualified, then the operational question becomes obvious: what should your store do to convert that traffic better?

1. Reduce friction in validation

When users arrive, they need to confirm quickly:

• this is the right product
• it fits their use case
• price and availability are clear
• shipping and returns feel reasonable

Do not hide any of this.

2. Reinforce real-world context

Show clearly who the product is for, which problem it solves, and when it works best.

3. Make compatibility and limitations visible

This is essential for confidence and for reducing returns.

4. Prioritize useful social proof

An average star rating is not enough. Highlight reviews that explain sizing, outcomes, fit, use cases, and comparisons.

5. Design for decision, not only discovery

Your page should be ready to close a user who already arrives half convinced.

A 90-day action plan to start now

All of this sounds strategic, but it also needs to become operational. If you wanted to make your store more visible in ChatGPT, Perplexity, and similar experiences starting today, this would be a realistic plan.

Days 1 to 30: clean up the foundation

• Audit your best-selling product pages
• Rewrite titles and descriptions to be more specific
• Fill in missing product attributes
• Implement or fix product schema markup
• Add conversational FAQs to priority product pages
• Check consistency across PDPs, feeds, and Merchant Center

Days 31 to 60: strengthen trust signals

• Improve your review collection system
• Ask for more detailed post-purchase feedback
• Identify relevant publishers, creators, and communities in each category
• Publish useful buying guides and comparison content
• Turn recurring customer questions into content assets

Days 61 to 90: measure, refine, and scale

• Review pages receiving AI referral traffic
• Analyze their behavior and conversion rates
• Identify which products generate stronger engagement
• Repeat the model across additional categories
• Adjust messaging based on objections found in reviews, support conversations, and forums

This plan does not require waiting for a future AI revolution. The revolution is already here. What businesses need now is to adapt commercial operations and content strategy to the way purchase intent is being formed today.

What brands need to stop doing if they want AI visibility

Knowing what to do matters. Knowing what to stop doing matters just as much.

Stop writing generic product descriptions

If any competitor could copy and paste your product page without losing meaning, that page does not have strategic value.

Stop thinking only in short keywords

Search is now conversational, contextual, and comparison-driven.

Stop separating SEO, content, and reputation

In AI discovery, those functions merge. Visibility depends on their combination.

Stop measuring success only in clicks

Pre-click influence is now part of performance.

Stop ignoring communities

What people say about your brand in external spaces can directly affect your odds of being recommended.

The new SEO for eCommerce is also a recommendation strategy

For years, SEO for online stores focused on indexation, categories, titles, links, and a bit of transactional content. All of that still matters. But on its own, it is no longer enough.

Today, the real challenge is turning your catalog into a trustworthy, understandable, and citable source inside generative answer environments.

That requires an evolved mindset:

• From keywords to real customer questions
• From promotional product pages to useful ones
• From self-declared authority to distributed reputation
• From massive traffic to qualified traffic
• From pure ranking to algorithmic recommendation

The brands that understand this earlier will build an advantage that becomes very hard to catch later. Because this is not just about optimizing a page. It is about building a digital presence that both people and machines consider trustworthy.

Conclusion

If your clicks are down, that does not automatically mean your relevance is down. You may simply be seeing the effect of a new reality: shoppers are getting answers before they visit websites, comparing brands inside AI interfaces, and arriving later in the journey with stronger intent.

That changes the business question. It is no longer enough to want more traffic. Now you need your brand to be understood, cited, and recommended when someone asks ChatGPT, Perplexity, or Google’s AI-powered search experiences what to buy.

The opportunity is significant for stores that move quickly. The businesses that structure product information better, write genuinely helpful FAQs, strengthen off-site reputation, and design product pages for high-intent visitors will be in a much stronger position to win in this new discovery layer.

The future of eCommerce is not only about being visible. It is about being the right answer.

If you want to prepare your store for this new era of AI search, optimize your product pages, and build a strategy that turns visibility into sales, let’s talk.