Guest Articles

Eventually, they return to your website and submit an enquiry.

Which channel generated that customer?

The honest answer is rarely as simple as our analytics reports suggest.

There has always been a grey area between discovery and decision. Marketing platforms might attribute the enquiry to Google, LinkedIn or direct traffic, but the final click only tells us a small part of the story.

Google describes this complicated stage of the customer journey as the “messy middle”: the space between the initial trigger and the final purchase decision.

Generative AI is making that middle even larger, more influential and more difficult to track.

Platforms such as ChatGPT, Claude, Gemini and Google’s AI-powered search features are becoming research and decision-making tools. Customers can use them to understand a problem, compare possible solutions, evaluate providers and decide what to do next, often before visiting a single company website.

For businesses, this creates a significant opportunity. AI search may generate fewer visible website visits than traditional search, but the people who do arrive can be better informed, more specific in their requirements and closer to making a decision.

The customer journey was already messy

The traditional marketing funnel suggests that customers move neatly from awareness to consideration and then conversion.

Real purchasing behaviour is rarely so linear.

Google’s original “messy middle” research identified two recurring mental modes: exploration and evaluation.

During exploration, customers expand their options. They search for information, discover providers and consider different ways of solving their problem.

During evaluation, they narrow those options. They compare services, read reviews, examine evidence and decide which providers appear credible.

Customers can move between these two modes repeatedly. They might shortlist three businesses, discover an alternative approach, reconsider their requirements and then begin comparing providers again.

This creates an attribution problem. Analytics software usually records individual visits and identifiable interactions. It cannot reliably capture private conversations, word-of-mouth recommendations, content viewed on other platforms or the thinking that took place between each touchpoint.

Even before generative AI, marketers could only see fragments of the complete journey.

AI search adds another influential layer. A customer can now complete a considerable amount of exploration and evaluation inside a private conversation with an AI assistant. Unless they click a cited source or visit a recommended business directly, much of that process remains invisible to the companies being considered.

From fragmented keywords to complete conversations

Traditional searches are often short and disconnected.

Someone might search for “accounting firm Dubai”, followed by “UK company expanding to UAE tax”, followed by “best business accountant Dubai reviews”.

The customer has to divide the problem into keywords, open multiple websites and assemble the relevant information themselves.

Generative AI allows them to describe the complete situation:

“I run a British-owned company that is expanding into Dubai. We expect to employ ten people during the first year and need support with corporate tax, VAT and monthly reporting. What type of adviser should we use, what should we expect to pay and how should we compare potential firms?”

The query is conversational, highly specific and rich in context.

An AI platform can respond by explaining the available services, identifying potential risks, suggesting questions to ask and comparing different approaches. It may also recommend relevant providers or reference external sources that support its answer.

The customer can then continue the conversation:

• Which option is best for a company of our size?
• What warning signs should we look for?
• Which firms have experience supporting British businesses?
• What information should we prepare before requesting a proposal?
• How do these two providers compare?

Exploration and evaluation begin to happen within the same interface.

AI search does not only help customers find companies. It can help them define their problem, understand the solution and establish the criteria they will use to choose a provider.

That gives AI platforms increasing influence over which businesses enter the customer’s consideration set.

Why AI-referred customers may be more valuable

Website traffic has traditionally been one of the most visible measures of SEO success. However, more traffic does not automatically mean more business.

A company might attract thousands of visitors who are researching a general subject but have no immediate requirement. Another page might attract only ten visitors, five of whom are actively looking for a provider.

AI-referred visitors can be particularly valuable because much of the initial research may already have taken place.

One Semrush analysis found that the average visitor referred by an AI search platform was worth 4.4 times more than a traditional organic search visitor when measured by conversion rate. Results will naturally vary by industry, business and method of measurement, but the reasoning behind the difference is important.

By the time an AI user reaches a website, they may already understand:

• The nature of their problem
• The type of service they require
• The available alternatives
• The factors that differentiate providers
• Why the recommended business may be suitable

In this sense, an AI-generated visit can behave more like a referral than a conventional search click.

People use assistants such as ChatGPT and Claude to interpret complicated information and make recommendations based on their specific circumstances. If the assistant introduces a company as a suitable option, some of that perceived trust can transfer to the recommended business.

AI should not literally be treated as a trusted friend. Its answers can be incomplete or inaccurate, and customers should still validate important information. Nevertheless, the interaction can feel more like receiving tailored guidance than viewing a list of advertisements or search results.

This is why visibility before the website visit matters so much. If a business is not present within the sources, discussions and evidence used during the evaluation process, it may never reach the customer’s shortlist.

How businesses become visible in AI search

Generative engine optimisation, or GEO, is sometimes presented as a replacement for traditional SEO.

In practice, the two are closely connected.

Search engines and AI platforms both need to understand what a business does, who it serves and why its information should be trusted. A technically inaccessible website with vague service descriptions and little supporting evidence is unlikely to perform well in either environment.

Clear positioning around specific problems

Businesses should explain their services in the language customers actually use.

A generic statement such as “we provide innovative solutions for modern businesses” gives an AI platform very little meaningful information.

A stronger website explains the specific problems the company solves, the types of customers it supports, the locations it serves and the circumstances in which each service is appropriate.

Clarity makes it easier for AI platforms to connect a business with a detailed customer requirement.

Evidence, authority and independent validation

Trust is particularly important when AI systems decide which sources to use and which businesses to mention.

Google often discusses quality through the principles of experience, expertise, authoritativeness and trustworthiness, commonly shortened to E-E-A-T.

These qualities can be demonstrated through:

• Detailed case studies
• Named and experienced authors
• Relevant qualifications and accreditations
• Genuine customer reviews
• Original research or professional insight
• Media coverage and industry contributions
• Membership of recognised organisations
• Consistent company information across the web

A business describing itself as an expert is not the same as independent sources demonstrating that expertise.

Useful content that supports buying decisions

Many businesses publish introductory content designed to attract broad search traffic but provide very little information for customers who are closer to purchasing.

Decision-stage content should address the questions people ask when comparing their options: 

• How much does the service cost?
• How long should it take?
• Which option is right for a company of our size?
• What are the risks?
• What should be included in a proposal?
• What alternatives should we consider?
• What separates a good provider from a poor one?

This content helps potential customers evaluate the business while also giving search engines and AI platforms useful information to reference.

Strong traditional SEO foundations

AI visibility does not remove the need for technical SEO.

Websites must still be accessible, logically structured and easy to understand. Important information should not be hidden inside images, inaccessible scripts or vague sales copy.

Clear page structures, descriptive headings, internal links, structured data and technically sound websites help search engines find and interpret information. These foundations can also make that information easier to retrieve and use within AI-generated answers.

Practical actions businesses can take now

The first step is to audit how the business is described online.

Search for the company using Google, ChatGPT, Claude and other relevant platforms. Ask the types of questions a potential customer might use and examine whether the business appears, how it is described and which competitors are recommended instead.

Next, review the website from the perspective of someone making a decision. Does it answer specific buying questions, explain the differences between services and provide credible evidence?

Businesses should also research the real discussions taking place around their services. Platforms such as Reddit, industry forums, customer reviews and community groups can reveal the questions, frustrations and concerns that polished keyword research often misses.

This does not mean entering communities to post disguised advertisements. It means listening to the language customers use and creating genuinely useful resources that address those concerns.

Businesses should then strengthen their independent authority. This might include contributing expert insight to reputable publications, developing case studies, earning relevant coverage, collecting authentic reviews and participating in respected professional communities.

Finally, measurement needs to evolve.

AI referral visits from platforms such as ChatGPT, Claude and Perplexity can often be identified in GA4 using traffic acquisition reports and customised channel groupings. However, this only captures users who click through directly. AI-influenced customers who later search for the brand or visit the website another way may appear as organic, direct or referral traffic.

Google has also introduced a dedicated generative AI performance report in Search Console, providing greater visibility into how websites appear within Google’s generative AI search features.

Businesses should combine this information with CRM data, enquiry quality, conversion rates and a simple question asking new customers how they discovered the company.

No individual report will provide the complete picture.

Search optimisation is moving beyond rankings

Rankings remain useful, but they no longer describe the whole search journey.

Businesses must now consider whether they are visible during both exploration and evaluation. That means being discoverable when customers identify a problem, credible when they compare solutions and persuasive when they decide which provider to contact.

Generative AI is expanding marketing’s messy middle while simultaneously helping customers move through it more quickly.

The companies that benefit will not necessarily be those producing the most content or chasing every new AI optimisation tactic. They will be the businesses with the clearest positioning, strongest evidence and most useful information.

Search visibility is no longer only about appearing when someone enters a keyword.

It is about becoming a business that search engines, AI platforms and customers can confidently understand, trust and recommend.

Author: Oscar Scolding, Founder of Eclypseo