From large-scale model exposure to the first order, how can overseas marketing measure the return on AI?

Chinese companies going global are shifting from selling products to building brands.

According to data from the General Administration of Customs, my country's exports of high-tech products and domestically branded products increased by 39% and 25.4% respectively in the first half of 2026. While export volume and product capabilities continue to improve, this does not necessarily mean that these brands have entered the stable selection range of overseas consumers.

How to be discovered, understood, and trusted remains a crucial step for Chinese companies as they move from cross-border sales to global operations.

The widespread adoption of generative AI has introduced new variables to this issue. Consumers are beginning to use AI to compare products, research brands, and obtain purchasing advice, thus adjusting the marketing chain that was previously built around search, clicks, and conversions.

On September 3rd, impact.com held the iPX 2026 China Overseas Marketing Summit in Shenzhen. Regarding the changes brought to the industry by AI, impact.com summarized it as a shift from the "search era" to the "answer era."

Its Chief Revenue Officer Justin Morrison and Greater China President Jennifer Zhang also shared their views on industry changes after the meeting.

In their view, businesses are shifting their focus from AI functionality to market opportunities and business results. AI can help brands find partners and analyze large-scale model exposure, but securing sustained budgets still depends on metrics such as new customers, revenue, and retention.

AI has increased the speed of information generation and resource matching, but it has not simultaneously shortened the cycle of product refinement, localization, and brand trust accumulation.

01 AI is changing the way brands are interpreted

In the past, consumers still had to open multiple pages and compare information themselves after entering keywords; now, AI can first complete the filtering and summarizing, putting a limited number of brands and reasons into the same answer.

Justin Morrison summarizes this as a shift from "active search" to "being recommended".

This change doesn't mean traditional search will disappear anytime soon, but rather that an information intermediary has been added between brands and consumers. Brands may still complete transactions through search, e-commerce platforms, or their official websites, but whether a product or service appears in the candidate pool is increasingly influenced by AI answers at a much earlier stage.

Public data reveals this coexistence. Adobe's statistics, based on over 1 trillion visits to US retail websites, show that in the first quarter of 2026, visits from generative AI sources increased by 393% year-on-year; in March 2026, the conversion rate of these visits was 42% higher than that of non-AI traffic.

This indicates that AI-driven traffic is still in a phase of rapid growth, but the year-on-year growth rate itself cannot prove that its traffic volume has exceeded that of mature channels.

An analysis of 68,900 searches by Pew Research Center showed that 8% of users clicked on traditional search results when AI summaries were available, compared to 15% when no summaries were available. Meanwhile, Google's search-related revenue grew by 17% year-over-year in Q2 2026.

The more accurate assessment at this stage is not that AI is replacing search, but rather that search interfaces and traffic allocation methods are being transformed by AI.

For the marketing industry, the key change is occurring at the level of where the answers are sourced.

Justin Morrison emphasized that a significant portion of the content cited by AI is created by partners. Jennifer Zhang further mentioned that the tools can trace which content sources are being indexed by AI and thereby identify influential collaborators.

Therefore, creators, content publishers, and user communities, who previously mainly served the functions of seeding interest and driving traffic, may now also influence whether a brand can enter the AI's recommendation scope.

The collaboration between impact.com and generative search optimization company Evertune follows this logic. The company attempts to identify creators and publishers frequently cited by AI answers and then integrate these entities into its partner recruitment and management process.

Evertune revealed that among the 10,000 most cited sources in its analysis of AI, over 40% of the content contained affiliate links or sponsored attributes.

This data comes from our partner's own research and cannot be used to conclude that paid content will necessarily lead to AI recommendations. AI answers are also affected by question wording, region, time, model version, and retrieval mechanism; brand mentions are not consistent.

For businesses, generative search optimization is closer to a content and data capability that is still developing than a new advertising product where placement can be bought out.

02 Enterprises are beginning to reassess the return on their AI investments

After the change in entry points, the real problem facing enterprises is not how many more AI functions they can add, but rather what business opportunities these capabilities correspond to, whether they can serve the company's operational priorities, and what metrics should be used to judge whether the investment is effective.

“The core of most conversations is no longer just about solving pain points, but about how to seize opportunities,” Justin Morrison told Wall Street Insights.

According to Justin Morrison's observations, when corporate executives evaluate marketing technology service providers, they not only ask which AI capabilities their products currently use, but also assess their understanding of industry changes, future product directions, and whether these capabilities align with the company's core priorities and performance indicators.

AI has thus evolved from a single function into a factor for enterprises to judge the long-term capabilities of service providers.

Justin Morrison believes AI can improve the efficiency of finding and matching. Jennifer Zhang mentioned that brands also hope to use AI to analyze their exposure in large model answers and to identify potentially influential partners based on the cited content. These applications address the problems of information overload and selection efficiency.

But identifying potential opportunities is only the beginning of the growth chain. In an environment where capital costs are high and marketing budgets place greater emphasis on measurable results, brands still need to determine whether a partnership can simultaneously achieve brand awareness building and revenue conversion.

AI mention rate, answer ranking, content reach, and number of partners can reflect process changes, but they cannot replace operational metrics such as new customers, sales revenue, customer acquisition cost, and customer retention.

Especially when AI-generated answers and reviews influence consumers, but the final transaction occurs on the official website or e-commerce platform, a single last click is unlikely to fully explain the contribution of the front-end content.

When evaluating the return on investment in AI, companies also need to clarify the time frame for observing the results.

Platform integration, partner recruitment, securing the first order, and establishing stable market influence are all at different stages. If measured by the same timeframe, the speed of technology deployment, short-term conversion rates, and long-term brand building can easily be confused. Justin Morrison stated that it typically takes several weeks, and at most several months, from project launch to securing the first customer or order.

Jennifer Zhang explained that large brands with existing brand awareness and reputation may see initial results within one or two months; however, for smaller brands with lower brand awareness, it may take six months, a year, or even longer to build partner response and market influence.

Therefore, AI can improve the efficiency of information analysis and partner matching, but it cannot compress platform launch, first conversion, and brand building into the same cycle.

For businesses, it is necessary to observe both short-term orders and assess whether partner networks and brand awareness can be continuously built up in order to evaluate whether the opportunities brought by AI can truly be transformed into business returns.

03 From clicks to recommendations, attribution shifts are forcing organizational restructuring

When a consumer decision is influenced by AI answers, creator content, affiliate links, and user recommendations, the attribution problem further transforms into an organizational problem: which team manages the same partner, and should their contribution be counted as brand exposure, content marketing, or sales conversion?

Jennifer Zhang noted that the boundaries between different types of partners are becoming blurred. Influencers can participate in affiliate marketing simultaneously, and media partners may also create content and manage social media accounts.

If companies still set up teams entirely based on a single channel, the same partner may be contacted repeatedly by multiple departments, and the relevant data will be difficult to compare under a unified standard.

This issue has now entered into the realm of internal resource allocation within the company.

Jennifer Zhang stated that many companies still have separate teams for email, influencers, and search, and some haven't even truly started KOC (Key Opinion Consumer) and consumer referral programs. This fragmentation between teams can lead to overlapping partner resources and unreasonable KPI settings.

Jennifer Zhang summarizes this phenomenon as some companies "setting up teams for a model, rather than for a way to actually get results." The adjustments she refers to are not just merging departments, but re-evaluating the organization, data, and budgets in terms of results such as customer acquisition, retention, and revenue growth.

Regarding senior management responsibilities, Justin Morrison believes that the core tasks of any company revolve around acquiring customers, retaining customers, and achieving revenue growth. Traditionally, customer acquisition is typically the responsibility of the Chief Marketing Officer (CMO), while customer retention and revenue growth fall more under the purview of the Chief Revenue Officer (CRO). However, the specific roles depend on the company's customer type and business priorities.

Jennifer Zhang predicts that as marketing departments are required to have more business and sales thinking, some senior positions may be merged in the next three to five years. She further stated that if she were the CEO of a company, she would integrate the CMO into the CRO, or have the CMO report to the CRO.

04 From exporting products to localizing operations, AI has not shortened the trust cycle.

For Chinese companies, the change in AI entry points occurs on top of another transformation: many companies already have the ability to produce products, supply chains, and rapid iteration, but from completing exports to operating overseas brands, they still need to solve the problems of local expression, community relations, and management authorization.

Justin Morrison believes that Chinese brands have an advantage in product innovation and rapid market launch, but to achieve scale in the global market, they still need to build brand trust and deeper customer relationships.

Jennifer Zhang summarized the current stage as a start-up and exploration period. New brands need to build brand awareness, and even established companies with years of experience in supply chain and production need to adjust their brand messaging according to the target market. This means that products and marketing methods validated domestically cannot be directly replicated overseas.

Jennifer Zhang cites the example of a tattoo equipment company entering the US market. The company later discovered that its target users were more concentrated among tattoo artists, music events, bars, and street culture communities, rather than traditional industry trade shows.

This case doesn't reflect channel selection techniques, but rather the prerequisite for localization: companies must first understand users' lifestyles and community structures before they can determine who to collaborate with and what content to use for communication. AI can assist in organizing information and finding targets, but it cannot replace the judgment of those on the front lines of the market.

Another hurdle for local operations lies within the organization itself.

Jennifer Zhang observed that while some Chinese companies recruit employees overseas, they rarely bring local talent into core management positions; other companies, due to differences in work pace and management culture, have returned business to their domestic teams.

Whether overseas teams can participate in decision-making, access resources, and be accountable for results directly impacts a company's responsiveness to changes in the local market. Simply deploying local personnel at the execution level may not achieve true organizational localization.

This set of judgments points to a change in the competitive foundation of Chinese companies going global.

By leveraging product efficiency and channel deployment, companies can quickly enter the market; once they reach the stage of large-scale operation, competition begins to extend to brand expression, third-party content, community relationships, and user retention.

AI has changed how consumers obtain information, and it has also made information gaps and misinterpretations in overseas markets more likely to directly affect whether a brand is included in the recommendation scope.

In the longer term, AI has not bridged the gap between product globalization and brand globalization. It can reduce the time required for content production, information analysis, and resource allocation, but it cannot simultaneously reduce the processes of cultural understanding, product validation, and building user trust.

For Chinese companies, the key in the next stage is not just selling more products overseas, but whether they can transform their supply chain and innovation advantages into stable brand recognition and local operating capabilities.

Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.