Qianfeng Interview with Sun Licui of Xinggu Cloud: AI Marketing Turns Brand Investment into Business Opportunities, Enabling Stores to Continuously Convert Leads

【Ebrun Original】Sun Licui is a partner at Shanghai Xinggu Information Technology Co., Ltd. (Xinggu Cloud). Built on nearly 20 years of B2B overseas expansion industry know-how, Xinggu Cloud focuses on automobiles, two-wheelers, energy storage, robotics, AI computing power, smart manufacturing, and consumer brands, creating a one-stop overseas AI marketing and sales intelligence platform. Its business focuses on brand building, global customer acquisition, sales conversion, and customer operations. In this Qianfeng Award interview, Sun Licui shared Xinggu Cloud's practices and judgments in the AI overseas marketing direction, centered on the modal evolution brought by AI and overseas expansion.

Ebrun: This year's Qianfeng Award takes "Modal Evolution Brought by AI and Overseas Expansion" as its core theme, focusing on the deep restructuring of industrial capabilities. In the context of the AI overseas marketing industry, how do you understand this round of industrial modal evolution jointly driven by AI and globalization? What are Xinggu Cloud's core strategic layouts in AI capability building and global overseas expansion? Could you specifically introduce the company's positioning and business model?

Sun Licui: We judge that the competition in overseas expansion of China's advantageous industries is moving from product and channel layout further toward competition in global operating capability. In industries such as automobiles, two-wheelers, and energy storage, many companies have already established overseas brand and channel systems. The next key is to turn brand investment into effective business opportunities, enable sales and stores to continuously convert those opportunities, and let customer relationships accumulate as long-term operating assets. AI provides a new path for the coordinated upgrading of these links.

As a leader in the AI overseas marketing field, Xinggu Cloud, built on nearly 20 years of B2B overseas expansion industry know-how, focuses on automobiles, two-wheelers, energy storage, robotics, AI computing power, smart manufacturing, and consumer brands, creating a one-stop overseas AI marketing and sales intelligence platform. Our business focus is brand building, global customer acquisition, sales conversion, and customer operations, and we drive these links to collaborate with headquarters, regional teams, and channel partners.

Our strategic focus is to promote the deep integration of enterprise knowledge, industry data, and AI into core marketing and sales scenarios. Through platform products, industry solutions, and professional services, we transform enterprise product knowledge, sales methods, and market feedback into executable and reusable organizational capabilities, helping companies build AI Native marketing and sales organizations.

Ebrun: You mentioned that the overseas expansion of China's advantageous industries is moving from "product and channel layout" to competition in "global operating capability." What does this judgment mean for Xinggu Cloud's product layout? Why focus on industries such as automobiles, two-wheelers, energy storage, robotics, and AI computing power?

Sun Licui: A common point of these industries is that many companies have already established overseas brand and channel systems. The next bottleneck is not "whether there are channels," but "whether brand investment can form effective business opportunities, whether business opportunities can be continuously converted by sales and stores, and whether customer relationships can accumulate as long-term operating assets." Therefore, Xinggu Cloud's product is not a single-point customer acquisition tool, but a one-stop platform covering brand building, global customer acquisition, sales conversion, and customer operations, driving these links to collaborate with headquarters, regional teams, and channel partners.

Ebrun: Currently, in which core business links has your company implemented AI/large model technology? What kind of distinctive AI or digital-intelligence competitiveness system has been formed?

Sun Licui: We apply AI throughout "brand expression—customer acquisition—sales conversion—customer operations": content production and marketing operations help companies convey product value; inquiry reception, demand identification, and product recommendations support business opportunity advancement; continuous follow-up and human-machine handover connect subsequent conversion. Enterprise knowledge, task planning, and tool invocation support AI execution within authorized scope, enabling these links to form a continuous workflow.

Industry depth is reflected in the understanding of specific operating logic: automobiles and two-wheelers value dealer empowerment, store visit invitations, and terminal sales; energy storage values project demand identification and channel cooperation; robotics, AI computing power, and smart manufacturing value technical value expression and complex demand communication; consumer brands value localized communication and customer operations. We translate these differences into knowledge, processes, and execution strategies, allowing technology to adapt to industries.

Ebrun: Could you share a most representative "AI value chain" innovation practice case? What common industry pain points did it solve, and what perceptible and quantifiable value did it create?

Sun Licui: Ofero is a new two-wheeler brand established in 2023, developing electric two-wheeler business for emerging markets, with nearly 400 overseas stores. Through insights into local consumer groups, usage scenarios, and climate characteristics, it advances product localization and expands overseas retail networks. As brand and channel scale expand, the company needs to simultaneously improve coordination among headquarters, online marketing, and local stores. In the early stage of our cooperative project, based on a store network of 260 stores in Indonesia, we focused on solving problems such as scattered traffic, high dependence on manual reception, difficulty covering nighttime and weekend inquiries, and easily broken links from online customer acquisition to in-store conversion.

Together with Ofero, we built a full-chain AI growth system around brand knowledge, marketing, sales, and customer management: brand and product information, frequently asked questions, and daily communication samples are accumulated into a knowledge base; AI shopping guides continuously receive inquiries, communicate needs, recommend products and stores, and advance test drive invitations; local stores handle offline experience and conversion. Customers who have not yet visited a store or converted are continuously nurtured, and conversations and customer information accumulate as enterprise assets. Marketing, AI reception, and store teams collaborate around the same customer path, allowing online traffic and offline operations to connect.

According to same-caliber results disclosed in the first phase of the project, the cost from business opportunity to conversion dropped from a historical baseline of about 1,000 yuan to about 200 yuan, a decrease of about 80%. Ofero's shared materials also show that inquiries outside working hours account for more than 50%, making the value of continuous AI service more concrete: helping the brand handle nighttime and weekend demand and enabling stores to understand prior communication background when continuing with customers. The above cost change is a phased result under a specific business scenario and cannot be directly applied to other industries or companies. It reflects that the operating return of marketing investment is closely related to the coordination of business opportunity reception, customer nurturing, and store conversion.

Ebrun: You mentioned that the further value of this case is to enable existing brands and store networks to form a continuously optimized operating system. How does this "operating system" specifically operate? In the next stage, Ofero plans to connect national official accounts and multiple store accounts through an AI hub. What problem does this solve?

Sun Licui: The further value of this case is to enable existing brands and store networks to form a continuously optimized operating system: customer questions enter the knowledge base; communication and conversion data support marketing review; verified experience is then used to improve execution. Ofero proposed in its sharing that in the next stage it will connect national official accounts and multiple store accounts through an AI hub, balancing official account distribution and store autonomous operations, and exploring replication across countries. This is a subsequent upgrade direction, with the core being to let headquarters' brand and knowledge capabilities support local stores while retaining the flexibility of local operations.

Ebrun: Regarding computing power and Token costs, how does Xinggu Cloud calculate them? In a single customer intelligent customer service scenario, one session consumes an average of about 500,000 Tokens, and the average daily total consumption is about 200 million Tokens. How should this magnitude be understood?

Sun Licui: We calculate costs by business scenario, not just by Token unit price. We mainly focus on average daily call volume, the number of input and output Tokens per call, and the inference price of the model used, then include supporting resource costs such as retrieval and storage. For implemented scenarios, we focus more on "the cost of completing one effective task" and compare it with saved labor time, processing efficiency, or new revenue. Taking the currently counted single-customer intelligent customer service scenario as an example, one session consumes an average of about 500,000 Tokens, and the customer's average daily total consumption is about 200 million Tokens. The two figures respectively correspond to a single session and the customer's overall average daily usage. We have also been controlling costs through model tiering, prompt optimization, caching, and batch processing.

Ebrun: In advancing large-scale AI implementation, what supporting role do data elements play? Does Xinggu Cloud focus more on deep cultivation of its own industry data or on advancing cross-entity and cross-industry data collaboration? What are the core challenges, and how do you respond?

Sun Licui: Data competitiveness is reflected in whether it can support marketing and sales judgments and continuously improve through execution feedback. Since its founding in 2010, Xinggu Cloud has accumulated overseas marketing practices, and has cleaned, desensitized, and processed industry knowledge and authorized website behavior, social media, and business communication data into five categories of thematic datasets:

First, multidimensional profiles of global manufacturing buyers: supporting demand identification, target customer matching, and marketing outreach.

Second, cross-border multilingual business communication and business opportunity conversion: supporting professional Q&A, inquiry follow-up, and customer nurturing.

Third, multimodal industrial product marketing and content generation: supporting website, graphic, and video content production.

Fourth, international trade compliance and risk control knowledge: providing rule references and risk alerts for automated execution.

Fifth, full-chain overseas strategy and execution effect evaluation: supporting conversion analysis, strategy recommendations, and execution calibration.

We prioritize data governance and scenario validation for customer enterprises themselves, enabling content, inquiries, follow-up, and conversion information to become linked. The core challenge is inconsistency in data standards, quality, and ownership, so we advance collaboration through unified standards, quality management, and tiered authorization; specific customer data follows their respective authorization boundaries, and key decisions are confirmed by humans.

Ebrun: You organized data into five categories of thematic datasets. Among them, "full-chain overseas strategy and execution effect evaluation"—what is its relationship with the previous four categories? Why is it placed last?

Sun Licui: The first four categories are buyer profiles, business communication, marketing content, and compliance risk control, solving "finding the right people, saying the right things, making the right content, and staying within the rules." The fifth category, "full-chain overseas strategy and execution effect evaluation," conducts conversion analysis and strategy calibration based on the first four categories of data, so it is placed last. Its role is to make the previous data not just records, but also able to guide the next round of marketing and sales actions in turn.

Ebrun: How do you view industrial agents rewriting B2B? Is there a long-term development necessity for vertical large models? Where is the AI moat of industrial internet companies?

Sun Licui: The value of industrial agents is to bring a company's existing professional capabilities into daily execution. In marketing and sales scenarios, what truly needs to be solved is whether AI can understand products and procurement needs, advance tasks according to sales rules, and hand over to humans at key nodes. Whether vertical large models are worth developing should be determined by professional tasks, data conditions, and actual returns; we value the combination of models with industry know-how, enterprise knowledge, and business tools.

The long-term moat comes from the data flywheel and organizational evolution: customer questions improve the knowledge base, conversion feedback calibrates content and follow-up strategies, and excellent sales methods enter the shared processes of teams and AI. Relying on nearly 20 years of industry know-how and deep services, Xinggu Cloud continuously transforms frontline practices into products and industry solutions.

Ebrun: Regarding overseas layout, which core markets has Xinggu Cloud currently entered? What key localization adjustments have been made?

Sun Licui: Xinggu Cloud has established a presence in Singapore and regularly participates in local overseas expansion exchange activities, including summits organized by Wu Xiaobo Channel, and has visited customers in markets such as Indonesia, Singapore, and Malaysia. We optimize content expression, customer communication, and collaboration processes based on local language and culture, brand communication, channel structures, and store operating characteristics, promoting the connection between headquarters strategy and local execution.

Ebrun: In building a deep overseas value chain, what model has Xinggu Cloud formed? What value does it create for overseas partners and local industries?

Sun Licui: Our capability iteration revolves around "how headquarters capabilities are transformed into local operating results": connecting brand content, product knowledge, and sales methods to overseas customer acquisition, lead reception, and store follow-up, enabling headquarters and local partners to work around customer continuity. AI continuously executes communication and nurturing tasks, while humans handle complex needs, business decisions, and partner relationships; operating feedback is then used to improve industry solutions.

In October 2026, the two-wheeler industry group went to Jakarta, Indonesia, to exhibit and visit Ofero, learning about the local market and store operating practices. With AI and the internet, Ofero connects online customer acquisition, customer communication, store visit invitations, and offline conversion; according to this visit, a single store is operated by 2 people and can achieve monthly sales of more than 100 units. Such frontline practices help us continuously improve industry solutions in combination with local channels and staffing.

Ebrun: You mentioned that a single Ofero store is operated by 2 people and can achieve monthly sales of more than 100 units. What does this store model mean for Xinggu Cloud's industry solutions? Why can "headquarters capabilities + local small stores" work?

Sun Licui: This store model shows that if headquarters deposits brand, product knowledge, and marketing capabilities into AI and the platform, local stores only need a small number of people to handle offline experience and conversion, and can still generate sales. Xinggu Cloud's capability iteration revolves around "how headquarters capabilities are transformed into local operating results": AI continuously executes communication and nurturing tasks, humans handle complex needs, business decisions, and partner relationships, and operating feedback is then used to improve industry solutions. Such frontline practices help us continuously improve industry solutions in combination with local channels and staffing.

Ebrun: In the process of overseas expansion, what core support do AI and digital capabilities provide for cross-border business?

Sun Licui: The value of digital intelligence is to connect brand assets, customer information, and sales processes, supporting the continuous conversion of marketing investment. DeepWebsite and content and GEO services operate official websites and brand knowledge assets; the MA automated marketing platform connects channels and first-party data; DeepSales advances business opportunities; O2O store AI operations connect offline terminals. The presentation, citation, and recommendation effects of AI search need continuous evaluation.

A unified SaaS foundation, GogloClaw, and SingooWork support AI applications and operational collaboration. Enterprises can connect content outreach, communication, follow-up, and conversion information around the same customer, enabling headquarters to review marketing and sales performance, regional teams and stores to continue service, and effective methods to be reused across teams.

Ebrun: Could you introduce the iteration of Xinggu Cloud's profit model? How do you view the evolution trend of profit models for industrial internet companies? How can Token payment become a standard industry revenue model?

Sun Licui: Xinggu Cloud serves customers through platform capabilities, industry solutions, and professional delivery. Fee items include platform subscription, delivery, implementation configuration, AI usage, and advertising placement, with scope and investment clarified by project. As AI goes deeper into marketing and sales processes, service value will be more reflected in continuous improvement of business opportunity reception, sales efficiency, and customer operations.

Token is a dimension for measuring technology usage; long-term payment relationships also require transparent usage, clear task boundaries, and verifiable results. Customers need to be able to judge: how many effective receptions and follow-ups these investments supported, what the execution quality was, and whether conversion efficiency improved. Business models should connect technology costs with customer-perceptible operating value.

Ebrun: To coordinate AI capability implementation, domestic value chain upgrading, and overseas business expansion, what key adjustments has Xinggu Cloud made in organizational structure or governance mechanisms? What are the biggest internal bottlenecks and external obstacles currently restricting large-scale AI implementation and deep global expansion?

Sun Licui: The internal challenge is to bring scattered knowledge and information into collaborative business processes. For example, can product materials support sales Q&A, can marketing leads be handed to sales and stores with communication records, and when AI executes versus when it hands over to humans. This requires clear data standards, role permissions, and review mechanisms.

The external challenge is that language and culture, procurement needs, channel structures, and product requirements differ across markets. Companies need to maintain consistency in global brands and professional standards while allowing local teams to flexibly adjust content, communication, and follow-up methods, so as to transform existing brand and channel advantages into sustained operating capability.

Ebrun: If you were to give one core piece of advice to peer companies just starting out and hoping to build both AI competitiveness and globalization capabilities, what would it be?

Sun Licui: Let business goals drive AI building. From brand customer acquisition, inquiry reception, sales follow-up, or customer operations, choose the link most in need of improvement, clarify the division of labor between AI and humans, and validate with business results.

Startups can first run through the process from customer acquisition to conversion in one representative market; companies with existing overseas brands and channel systems should prioritize improving coordination among headquarters, sales, and local partners. Accumulate effective knowledge and methods, then promote them in combination with local realities, allowing one scenario improvement to gradually transform into sustainable organizational capability.

As an annual bellwether in the TO B field, the "2026 8th Ebrun Industrial Internet Annual Conference and First Industrial Agent Conference" will be held in Beijing on November 20, 2026. The theme is set as "Modal Evolution Brought by AI and Overseas Expansion," focusing on deep thinking around frontier topics such as industrial agent implementation, AI supply chain value reconstruction, data elements and TOKEN economy, and new industrial overseas ecosystems. The conference will gather academic authorities, central and state-owned enterprise benchmarks, industry leaders, investment institutions, and service ecosystem partners to bring together core industry forces, break down industrial digital-intelligence transformation paths, and share frontier practical experience. At the same time, the annual industrial internet report and benchmark cases will be released, and a decision-maker closed-door meeting and the "Qianfeng Award" awards ceremony will be held.

This article was first published on the Ebrun official website.