The biggest thing to understand about AI search is this: AI doesn’t return pages. AI doesn’t drive clicks like normal search engines. What AI does is answer questions.

While you’re reading this, a potential customer could be asking AI about a topic related to your product. They’re trying to understand a problem. They’re learning about potential solutions. They’re comparing options and building a shortlist. Eventually, they might make a purchasing decision.

But you can no longer rely on potential customers visiting your website for answers. The click is redundant when AI has already given them detailed answers containing all the information they need.

As a result, the entire customer journey can happen in an environment that generates no analytics, no keyword data and virtually no referral traffic. But while you can’t track or even see these conversations, they’re directly shaping your commercial outcomes.

How certain are you that your brand would come up in these AI-generated answers? And if it does, are you confident those responses will accurately represent your ideal messaging?

If you’re thinking that AI isn’t really a problem or priority for your brand right now, consider that a recent McKinsey survey found that around 50% of consumers now use AI tools or features when conducting research and making decisions.

Of those who do use AI-powered search, 73% ask questions to learn about a category, brand, product, or service – typical top-of-funnel behaviour – while more than half continue to use AI throughout the entire decision journey.

There’s no doubt that the customer journey is changing, and rapidly. If your content isn’t built around answering the questions customers are asking – meaning the questions they’re really asking, not what you’d prefer they asked – AI may simply leave your brand out of the conversation. And if you’re not part of the conversation, you can’t influence the decision.

The questions your content doesn’t answer

I still regularly come across content strategies that start in the wrong place, focusing on what a brand wants to say, or the keywords most likely to generate traffic, rather than the questions potential customers are actively trying to answer.

While considering keywords can still boost your brand’s visibility in organic search, they won’t get your brand cited in AI-generated answers. AI cares more about topical relevance and depth than it does matching strings of letters.

For example, a shoe retailer might publish a comprehensive guide to running shoes, briefly covering all the different types – trail, track, road, and so on. It ranks well for a bunch of useful short-tail keywords, pulling in lots of top-of-funnel traffic. And because each section of the article links to the corresponding product page at the bottom of the funnel, those pages may get traffic that can convert. Classic SEO in action.

But look at it from the perspective of someone using AI to research running shoes.

If the retailer is very lucky, AI might draw on information from its handy primer to formulate an answer to the user’s initial question. If they’re even luckier, the response might explicitly mention the brand or include a link.

However, once the user has a grasp on the different kinds of running shoes, they’ll probably have further questions. As they dig further to learn more and refine their options, each new answer prompts yet more questions. What is the difference between a shoe built for performance and one built for comfort? Which features should they prioritise to get the most out of their budget? Do shoes from some brands accommodate orthotics better than others? The answers are likely to have a big influence on their final purchase decision.

But our fictional shoe retailer’s initial article doesn’t answer any of these follow-up questions. Its relevance to the AI conversation has expired just as the conversation began. That’s because the content was never designed to provide detailed answers. It was designed to rank for certain keywords, to attract website visitors who may or may not convert via internal links and CTAs. Take the web traffic away and the article no longer drives the customer journey forward.

Meanwhile, if none of the retailer’s other content contains the necessary information, the LLM will extract what it needs from other websites, other sources, other brands. The shoe retailer’s content may have pointed the customer in the right direction, but the lack of any other answer-based content with sufficient depth effectively handed over the rest of the customer journey to their competitors.

The customer questions that really matter

Content strategies need to shift from keyword-first to question-first thinking. Your content needs to be built for asking, not just ranking, in ways which support grounding and fan-out queries.

While the old model broke the customer journey into funnel stages, the new model focuses on decision questions. These serve as touchpoints that, with the right answer, help the customer to move forward in their thinking and closer to a decision.

Customer questions typically fall into one of five broad types:

1. Problem framing

The potential customer tries to understand the problem and whether it’s worth addressing. Is it really a problem? What’s the cause? Can I live with it? What are the risks? How urgently do I need to act? Most SEO content strategies start here, building awareness of a need.

2. Solution exploration

Now that the customer understands the problem, they begin looking for ways to fix it. What options are open to me? Which is the best approach? Are there any trade-offs? This is where AI often comes into its own, collating information from multiple sources in seconds that previously would have taken the customer hours.

3. Evaluation

Armed with a list of workable options, the customer begins to whittle them down by comparing specifics. How does A stack up against B for this or that criteria? Which features are essential, nice to have, or irrelevant?

4. Objections and risk

Before settling on a solution, the customer might stress-test their decision. What could go wrong? Is this worth the investment? What’s missing? Many brands avoid answering uncomfortable questions that might reveal shortcomings or competitive weak spots. But if your content doesn’t address them, AI will source the information from somewhere else, such as online reviews or a competitor’s comparison page – and you no longer get to shape and influence this pivotal stage of the conversation.

5. Use case validation

Probably most importantly of all, the customer wants to know whether their preferred option will work for them specifically – not in general. Is this the best fit for our particular situation, in our industry, with these constraints, budget, and other criteria? Generic “one-size-fits-all” content is extremely unlikely to satisfy these kinds of questions. For AI to answer these questions effectively, it will seek out content with much greater depth and granularity.

A good answer starts with listening

While most content strategies typically address problem framing questions reasonably well, the other four question types are where the decision actually gets made. Right now, they’re also where most brands are effectively absent.

But identifying and mapping all the potential decision questions in sufficient detail for each of your personas will take more than pulling your team into a room with a whiteboard for an afternoon. In the absence of clear data from these AI conversations, simply guessing what customers might ask won’t do.

Instead, find ways to talk to your customers; surveys, interviews, focus groups. Also, talk to those teams who interact with your customers every day: like sales and support. What questions do people ask? What problems do they struggle with? Which features do they value most, or least, and in what circumstances? Which messaging helps to close a deal or resolve an issue?

Only then can you start crafting strategic content with enough depth and specificity to be genuinely useful. Breadth of coverage isn’t the goal. A thousand words that comprehensively answers just one decision-critical question will do more for your visibility in AI than ten generic overviews that say nothing new.

You can’t win a question you’re not answering

Right now, you have a huge opportunity to effectively “own” the questions shaping customer decisions by ensuring your brand shows up consistently and usefully in those AI-generated answers.

Don’t wait for all the analytics and marketing platforms to catch up with AI search. Your best chance of gaining the competitive advantage is to move now, before mapping content to AI questions becomes so easy anyone can do it. There are still plenty of old-school methods you can use to find out what your customers are asking. Like, well, talking to people.

If you don’t own the question, if you’re not part of the answer, you don’t get considered. When customers can research, shortlist and make a purchasing decision before visiting your website even once – never mind being captured as a lead – getting considered is no longer a sales problem. It’s a content strategy problem.

Need help identifying the decision-critical questions your audience is asking? Want to build a content strategy where your brand becomes part of AI’s answer? Get in touch to find out how we can help: [email protected]