How to Get Your Brand Recommended by ChatGPT

For most of the last two decades, one of the most valuable positions a company could own on the internet was the top of Google.

An enormous industry was built around that simple idea. Consumers had questions, Google organized the available information, and companies competed to appear at the exact moment someone was looking for an answer.

We are now watching the early stages of what could be a meaningful change in that behavior.

Consumers are increasingly asking ChatGPT, Claude, Gemini and other AI platforms questions that would have historically started with a Google search. They are asking which software to buy, which products are best for a particular use case, which companies they should consider and how one option compares with another.

The important difference is that AI doesn't simply give the consumer a list of links and ask them to figure it out. Increasingly, it does a meaningful portion of the research for them.

A consumer might have opened ten different websites to compare products a few years ago. Today, they can describe exactly what they need, provide context about their situation, ask follow-up questions and receive a relatively small set of recommendations in seconds.

That changes the economics of discovery in a way that I think most companies are still underestimating.

If thousands of potential customers are asking AI platforms questions about your category, but your company rarely appears in the answer, there is now an entire layer of demand that exists without you.

This is the opportunity behind Answer Engine Optimization, or AEO, sometimes referred to as Generative Engine Optimization, or GEO. The terminology is still evolving, but the objective is the same: increasing the likelihood that your company is surfaced, cited and recommended when consumers turn to AI for answers.

The way we think about AEO at FutureSource is actually fairly simple. We want to understand what potential customers are asking AI, determine why certain companies are consistently being recommended in response to those questions, identify the difference between those companies and yours, and then systematically close that gap.

There are four parts to that process.

STEP 01Understand what your customers are asking and how often you're part of the answer.

Most companies naturally start by thinking about their own website, their content or what they want an AI model to know about them.

We think that's the wrong place to start.

The more useful starting point is understanding the conversations customers are already having.

Take a project management company as an example. Someone might ask ChatGPT for the best project management software, but that's only one possible conversation. A creative agency might ask which platform is best for managing twenty client accounts. A startup might want something simpler than Asana. A marketing team might specifically need approval workflows. Another buyer might ask which platform integrates best with the tools they already use.

Taken together, these questions begin to form a map of the actual market.

Some prompts have enormous commercial value. Others represent relatively narrow use cases. Some happen early in the research process, while others are being asked by someone who may be minutes away from making a purchasing decision.

The first thing we want to understand is which of those conversations matter most to the company and what the recommendation landscape currently looks like across them.

We build a large set of prompts around the category, products, problems, use cases, competitors and customer personas, then study how ChatGPT, Claude, Gemini and other major AI platforms respond to them over time. The objective isn't simply to determine whether a company "ranks in ChatGPT." It's to develop a much more complete picture of how often that company enters the consideration set when AI is helping someone make a decision.

A company might have strong visibility around its broad category but almost no visibility for an important use case. It might perform well in ChatGPT but poorly in Claude or Gemini. It might consistently appear when someone asks for alternatives to a competitor while rarely appearing in more general category recommendations.

Those differences matter because they tell us where the opportunity actually exists.

Before trying to improve AI visibility, we want to establish a clear baseline for how much of the relevant conversation a company currently owns.

STEP 02Reverse engineer why certain companies are recommended and why you're not.

Once we understand where a company is and isn't appearing, the work becomes much more interesting.

The question is no longer simply, "How do we rank higher in ChatGPT?"

The better question is why the model appears to have more confidence recommending certain companies than it does recommending yours.

AI recommendations are the result of an enormous amount of information and context. A company's own website matters, but it is only one piece of a much larger information environment. Models encounter brands through editorial coverage, product comparisons, reviews, documentation, industry publications, forums, communities, customer discussions and the broader collection of information that exists about a company across the internet.

This means two companies with relatively similar products can have dramatically different visibility inside AI.

One might be consistently described across the web as the best option for a particular type of customer. Another might have significantly more independent reviews. A competitor might appear repeatedly in authoritative comparison articles or have a much stronger presence in the communities where people discuss the category. Sometimes the difference is even more fundamental: the market has developed a very clear understanding of what one company is best at, while the positioning of another remains ambiguous.

Our job is to understand those differences.

For every important group of prompts, we study the companies that appear disproportionately often and work backward from the recommendation. We look at how those companies are positioned, which sources discuss them, what attributes are repeatedly associated with them, where those sources overlap and how all of that differs from the company we're trying to improve.

Over enough prompts, the patterns become increasingly obvious.

That is where AEO starts becoming less about guessing what an algorithm wants and more about understanding the information advantage one company has built over another.

Once you understand that advantage, you can begin to quantify the gap.

STEP 03Fill the gap.

This is where most of the actual work happens.

If the research shows that a competitor owns an important use case because dozens of credible sources consistently associate that company with it, publishing another generic blog post probably isn't going to change much.

You need to build the missing evidence.

Depending on the company, that can mean making its positioning significantly clearer, restructuring important pages, creating resources around questions customers actually ask, earning inclusion in relevant third-party publications, improving comparison content, generating more credible customer validation or becoming part of the discussions taking place in the communities that influence how the category is understood.

In practice, it is rarely one thing.

What makes this approach different from traditional content production is that we aren't starting with a mandate to publish more content. We are starting with a specific deficiency in how the company is represented across the information environment AI uses to understand the category.

We know the questions customers are asking. We know which companies are disproportionately appearing in the answers. We have a hypothesis for why those companies are winning, and we can identify the signals that are missing from our own brand.

The work is then focused on changing those underlying conditions.

This is also why I think the best AEO strategies will ultimately look much broader than SEO. The company's website is obviously important, but so is the rest of the internet's understanding of the company.

You are not simply optimizing a page.

You are trying to improve the probability that when an AI system is asked to reason about the best companies for a particular customer, there is enough credible and consistent information available for your company to deserve inclusion in the answer.

STEP 04Keep optimizing as the models and the market change.

The final piece is the one I think will become increasingly important over the next several years.

There is unlikely to be a permanent version of winning in AI discovery.

The models are changing quickly. The information available to them changes every day. Competitors are producing new content, earning new coverage, accumulating new reviews and changing how they're positioned. Consumers themselves are also becoming more sophisticated in how they use these products, which means the questions they ask six months from now may look very different from the questions they're asking today.

A company can therefore improve its visibility significantly and still lose ground later.

We think this makes AEO much closer to an ongoing intelligence function than a traditional one-time optimization project.

The prompts that matter need to be monitored, changes in recommendation share need to be understood, and new competitors need to be identified. When a company suddenly begins appearing much more frequently for an important category of questions, we want to understand what changed and whether there is something the rest of the market can learn from it.

Over time, this creates a feedback loop. We observe how the models understand the market, form a hypothesis about why certain companies are winning, make changes based on that hypothesis and then measure whether the models begin responding differently.

The companies that become particularly good at this will develop something more valuable than a collection of AI-optimized pages. They'll develop an organizational capability for understanding how their market is represented inside AI and influencing that representation over time.

The next era of discovery is just beginning.

There was a period in the early days of Google when search demand was already becoming enormous, but relatively few companies understood how valuable it would eventually become to influence where they appeared.

By the time everyone understood the opportunity, competition had become an industry unto itself.

I think there is a reasonable argument that we are in a similarly unusual period with AI discovery today.

The consumer behavior is already here. People are asking AI which products to buy, which software to use, which companies to consider and how they should make increasingly complicated purchasing decisions.

The commercial competition around those answers is still surprisingly immature.

That creates a window that probably won't exist forever. There is already meaningful demand flowing through these platforms, while in many categories relatively few companies are systematically competing to become part of the answer.

At FutureSource, we're a team of marketing strategists, researchers and growth engineers focused entirely on understanding that opportunity.

We spend our time studying what consumers are asking AI, measuring which companies are being recommended, reverse engineering why those companies are winning and figuring out what would need to change for our clients to take their place.

There is a research component to that work, a marketing component and an increasingly technical component. That's why we deliberately built the team across all three disciplines.

We don't think anyone has completely figured out how AI discovery will work five years from now. In fact, the speed at which these systems are changing makes us skeptical of anyone who claims they have.