Amazon Web Services Li Jian: Chinese companies face five major challenges when going overseas, and generative AI will become a new engine for Chinese companies going global | Alpha Summit

On December 21st, Li Jian, General Manager of Industry Cluster for Amazon Web Services in China, was a guest at the "Alpha Summit" co-hosted by Wall Street News and China Europe International Business School, where he deeply analyzed how generative AI can assist Chinese enterprises in going global to face new challenges.

He pointed out that Chinese enterprises going global have entered a new stage, requiring companies to accelerate their overseas market layout while actively integrating into the overseas ecosystem to achieve "win-win" development. Li Jian believes that generative AI will become a new engine for Chinese enterprises going global, helping them tackle four major challenges: product adaptation, efficient expansion, integration into the overseas ecosystem, and safety compliance.

Li Jian used excellent overseas companies such as Anker Innovations, Haier Smart Home, Muto Technology, and WPS as examples to demonstrate how generative AI technology can help enterprises achieve significant results in product innovation, efficiency improvement, and market expansion.

He also emphasized that companies need to change their mindset, integrate generative AI into corporate culture, cultivate employees' AI application capabilities, and establish effective risk assessment systems to gain a competitive edge in the AI era.

Wall Street News summarized the highlights of this speech:

• Chinese enterprises going global have entered a new stage, with intensified competition, where speed and efficiency become key, while companies also need to actively integrate into the overseas ecosystem to achieve "win-win" development.
• North America remains the preferred market for Chinese enterprises going global, but the importance of the European market is rising, and emerging markets such as the Middle East, Southeast Asia, Africa, and South America also hold great potential, providing diversified choices for enterprises.
• Generative AI is becoming a new tool for Chinese enterprises going global, helping them address challenges in multilingual interaction, market analysis, product innovation, customer service, and more.
• In the AI era, enterprises need to change their human resources strategy, reallocating the human resources saved from "streamlining" repetitive jobs to more valuable work, such as business development and product innovation, while using AI to enhance overall work efficiency.
• Implementing a generative AI strategy requires multifaceted efforts, including gaining support from top management, forming diverse professional teams, integrating AI into corporate culture, conducting repeated testing to find the best balance, starting experiments and learning from low-risk use cases, implementing protective measures, and building responsible generative AI projects.

The following is the full transcript of the speech:

Hello everyone, I am Li Jian. Recently, I have attended many sharing sessions and found two particularly prominent themes: one is going global, and the other is generative AI. As a practitioner in Amazon Cloud Computing, these two topics are highly relevant to our business.

Now, more and more people are beginning to pay attention to how generative AI can assist in the development of cross-border business, which is precisely my area of expertise. At Amazon Web Services, I mainly oversee the gaming and retail e-commerce industries, which are closely related to the current themes of AI and going global.

Today, I would like to share with you based on the new challenges and issues encountered by clients in various industries during their global expansion, as well as the new opportunities and innovative possibilities that generative AI brings to going global In 2024, competition in overseas markets is more intense than in previous years. Whether in new energy, electric vehicles, gaming, fast fashion, or the food and beverage industry, competition is escalating across various sectors. As a result, companies are placing greater emphasis on the speed of going overseas, hoping to achieve greater returns in international markets. Going overseas has become an important driving force for China's economic development, and we observe that 80% of Chinese clients going abroad choose to conduct business on Amazon Cloud.

From the early days of TikTok, e-commerce platforms like Shein and Temu, the expansion has now reached more fields. For example, in the healthcare sector, with the narrowing profits of innovative drugs domestically, products from HengRui, BeiGene, and Junshi have begun to go overseas after being approved as "tumor-suppressing drugs."

In the gaming and entertainment sector, this year's highly anticipated "Black Myth: Wukong" has surpassed 23 million global sales on Steam and won the Best Action Game award at the 2024 TGA. Additionally, my client's games, "Endless Winter" and "Whiteout Survival," have consistently topped the overseas mobile game revenue charts. The short video application ReelShort has also ranked first in the iOS entertainment charts in the United States. These successful overseas cases are closely related to Chinese listed companies, such as Century Huatong, whose main business highlight comes from "Endless Winter"; and the significant rise in stock price for Zhongwen Online this year is primarily due to the expansion of its overseas business, namely ReelShort.

In addition to traditional gaming and media apps going overseas, industries such as smart manufacturing, electric vehicles, cosmetics, and video live streaming are also beginning to enter international markets, with the scope of going overseas continuously expanding.

As a current student in the China-Europe EMBA program, my research topic focuses on how to leverage AI to assist companies in going overseas, and I have conducted in-depth research and gained personal insights in this field. So, where are the current opportunities for going overseas?

From a market perspective, North America remains the most important destination for going overseas. This is mainly due to the strong consumer power in the United States, where 70% of its GDP is driven by consumption. Therefore, whether in fast fashion, smart manufacturing, smart devices, or the gaming industry, North America is viewed as a key market.

At the same time, the importance of the European market is increasing day by day. Europe has become a market that both China and the United States are actively exploring. Many Chinese electric vehicle companies choose Europe as their first stop for going overseas, and other industries are also making their way into Europe. Of course, compliance with GDPR in the European market is an important challenge that cannot be overlooked.

There is an interesting saying in the industry: innovation in the United States is reflected in technology, innovation in China is reflected in market research, and innovation in Europe is reflected in compliance and regulation. Looking back at history, this saying holds some truth: typically, the U.S. leads in technological innovation, China innovates and promotes in market application, and finally, Europe guides the standardized development of the industry through regulations. This reflects the significant differences among the three major markets.

Another market worth noting is the Middle East. The Middle East not only has strong economic power but is also undergoing digital transformation, shifting its economy from traditional oil and petrochemical industries. The governments of Saudi Arabia and the UAE are vigorously promoting this transition. For example, Ctrip recently held a partner conference in the Middle East, and the UAE is actively attracting gaming companies from Shanghai to establish studios locally. The digital transformation of the Middle East market provides important opportunities for Chinese companies going overseas Another market to consider is Southeast Asia. The Southeast Asian market has two significant characteristics: first, a large population base, especially India and Indonesia (the dual India market), which has always been an important battleground for cross-border e-commerce; second, a young population structure with many young people. Coupled with consumption habits similar to those in China, it has become an important destination for Chinese companies going overseas.

In addition, there are two emerging markets worth noting:

One is the African market: not only is there the successful case of Transsion mobile phones, but the internet development speed in Africa is astonishing, and its population structure is even younger than that of Southeast Asia. Moreover, Africa has directly skipped the PC era and entered the mobile internet era, presenting enormous development potential.

The other is the South American market: it is increasingly favored by e-commerce and gaming companies and has become an important strategic layout area.

From the perspective of Amazon Web Services, the layout of cloud service infrastructure often indicates economic development prospects. As Ma Huateng said, the usage of cloud computing can reflect the level of economic development. From a global perspective, this is indeed the case—regions with good economic development trends and strong demand for digital transformation often attract investments from cloud computing giants.

Based on Amazon Cloud's experience serving numerous overseas clients in China, we observe that the current wave of going overseas faces several major challenges:

1. Product adaptation: How to develop better products to meet the needs of target markets

2. Efficient expansion: In the current era, how to leverage leading technology to facilitate rapid global business expansion

3. Integration into overseas ecosystems: A single blockbuster product is difficult to dominate the market; better integration into the overseas business ecosystem is needed

4. Overseas compliance: How to integrate into the legal system of the host country and how to avoid compliance risks related to cross-border data transmission

Among the clients we serve, many companies have achieved significant results in market insights and product innovation. Taking Anker Innovations as an example, the company has developed from Hunan Haiyi in 2016 to now, expanding its product line from initial accessories like chargers and power banks to smart cameras and energy storage products. In overseas markets, Anker has established a high-end brand image, and its products no longer rely on low-price strategies. They fully utilize cloud computing and AI technology to analyze customer data and market data, which is much more in-depth than simply relying on data from Amazon's cross-border e-commerce platform.

Haier Smart Home is another example. They need to evaluate 80,000 home appliance design proposals each year, and the scale of screening and data analysis is extremely challenging. Through generative AI technology, they have shortened the analysis cycle from months to days, greatly improving product design and rendering efficiency. These demonstrate how to achieve product innovation and market insights through cloud computing and AI technology.

In terms of rapid development, the gaming industry provides a great case. Taking the mobile game "Mobile Legends: Bang Bang" (MLBB) from Moonton as an example, they faced challenges in localizing for less commonly spoken languages when expanding into the Southeast Asian market. Traditional machine learning translation solutions mainly target mainstream languages such as Chinese, Japanese, English, and German, with weak support for less commonly spoken languages and limited by insufficient training data. However, through generative AI technology, after inputting industry-specific data and less commonly spoken language data, they developed better solutions For example, some abbreviations commonly used by gamers, such as "let's rush together" or "idle," may have unique expressions in different languages, which traditional translation software struggles to convert accurately. Generative AI can effectively solve this problem, significantly lowering the market entry barrier.

Another client is WPS, whose overseas development is impressive. During my recent visit to Zhuhai Kingsoft Office, I was amazed by WPS's achievements. They have directly transitioned from traditional office software to the mobile era, accumulating over 100 million users overseas. Their applications in generative AI are particularly impressive.

For instance, when writing monthly reports, WPS's AI assistant can intelligently recommend phrases based on context and even automatically calculate data such as month-on-month growth. This deeply integrated office assistant greatly enhances work efficiency. It is foreseeable that in the future, every office-related scenario will be equipped with AI assistants, which will be one of the most significant application areas for improving work efficiency.

Another example is smart hardware and automobiles. For instance, when Chinese smart home products enter overseas households, they need to interconnect with existing smart home ecosystems. This requires various certifications, security protocols, and standardized interfaces. If companies were to complete these processes independently, it would be very challenging, but cloud service platforms can significantly simplify this process. For example, SAIC can integrate Tencent Music or QQ Music in domestic cars, but overseas, it needs to collaborate with Prime Music, and voice assistants also need to interface with Alexa, all of which involve ecosystem cooperation.

In recent years, an increasing number of B2B SaaS products are also choosing to go overseas, including ERP systems and database products (such as PingCAP's TiDB and Ant Group's OceanDB). When these B2B software products go overseas, they need to embrace the overseas ecosystem more. Unlike B2C products that promote customer acquisition through App Store or Google Play, the overseas expansion of B2B products requires a completely different methodology and a deeper integration into the local business ecosystem.

Currently, compliance issues have become the primary consideration for Chinese companies going overseas. Compared to 8-10 years ago, companies have shifted from a mindset of "violate first, rectify later" to "comply first, operate later." Especially considering the potential political changes that may arise with Trump's return to power in January 2024, compliance issues will become even more critical.

In terms of data compliance and security, particularly in fields involving personal privacy such as healthcare, the requirements are extremely strict. Amazon regards security as its primary principle, "Security is Job Zero." Compared to the Chinese market, overseas markets have stricter requirements for personal privacy protection. In China, users may have become accustomed to precise push notifications based on personal behavior, but in overseas markets, such practices may lead to user complaints or even product removals. Therefore, for companies going overseas, choosing a mature cloud service platform like Amazon Web Services can better address these security and compliance challenges The Global New Technology Application Cycle Accelerates, and Overseas Enterprises Urgently Need Countermeasures

The rapid iteration of technology has intensified competition among enterprises, which is particularly evident among overseas companies. Whether it is the Shanghai Stock Exchange 50, the Shanghai Stock Exchange 300, or the turnover rate of the components of the S&P 500 in the United States, it may far exceed that of the past decade. The significant rise in Nvidia's stock and the rapid growth of AI-related products confirm this. In this era, for overseas enterprises, the key lies in how to embrace technology, especially generative AI, transforming it from a conceptual level into tangible business value. Since the emergence of large language models two years ago, various models have continuously emerged, marking not only the arrival of the AI era but also bringing new opportunities and challenges for enterprise development.

Although the general public mainly knows about ChatGPT, there are actually more than ten large language models available now, each with unique advantages in their respective fields. Over the past two years, my team and I, along with the Amazon Web Services team, have summarized a development path from business conception to business value: Learn, Build, Determine, Lead Action, and Set Sail Again through in-depth communication with numerous Chinese overseas clients. Let me elaborate on each step.

In the past two years, enterprises have spent about 70% of their time trying and exploring, and 30% of their time harvesting the value brought by these attempts, and they are willing to try more. We have summarized six main application scenarios, several of which are particularly important for overseas enterprises:

1. Multilingual Interaction: This is a major barrier for many enterprises, especially small and medium-sized enterprises, entering overseas markets. Unlike giants like Tencent and NetEase, which have a large number of algorithm talents, customer service teams, and multilingual talents, small and medium-sized enterprises often have to focus only on the English market due to language limitations. The emergence of large models fundamentally changes this situation.

2. Secondly, when we enter a new market, we often face many unknowns. Generative AI can mine and analyze massive amounts of market data. Based on historical data and models, it can simulate and evaluate the potential impacts of different decision-making options in the future, providing accurate decision-making basis for enterprises to formulate market strategies.

3. Insights and Decision-Making: When we enter a new market, we often face many unknowns. Generative AI can mine and analyze massive amounts of market data. Based on historical data and models, it can simulate and evaluate the potential impacts of different decision-making options in the future, providing accurate decision-making basis for enterprises to formulate market strategies.

4. Customer Service and Marketing Analysis: From the experience of using various apps daily, the importance of customer service is self-evident. More and more enterprises now regard customer service data as a core analysis object. Whether it is customer comments, complaint rates, reasons for product defects, or suggestions made by customers, these can all be quickly analyzed through large models, helping enterprises to identify problems and opportunities in a timely manner.

5. Building Internal Knowledge Bases: For building internal knowledge bases in enterprises, such as at Amazon, we particularly emphasize document writing, and all plans need to be documented. Writing in English used to be a very painful process, but now, with the help of large models, this process has become much easier I believe that in the future, China may also allow the use of AI in certain scenarios. This means that the next generation, and even our generation, needs to master this skill. Our energy should be invested in more valuable areas, rather than rote memorization. If we are to embrace AI now, the first step is to learn—observe the successful experiences of others and then put them into practice.

All industries are actively trying AI applications. Last week, I attended Amazon's re:Invent conference in the United States, where Amazon CEO Andy Jassy announced the company's own Nova large model. It is worth mentioning that Amazon has already applied large models in various scenarios in the e-commerce field, such as customer review analysis and shopping assistants. For example, if you want to understand the difference between running on a treadmill and on a greenway, you just need to ask Amazon, and the system will explain the differences and recommend suitable products. This is different from traditional Baidu searches, where users need to judge the accuracy of the information themselves and then separately choose brands to place orders, while the AI shopping scenario makes this process much smoother.

At the level of generative AI technology, we have experienced multiple levels of development over the past few years:

• The bottom layer is infrastructure, which is also the reason for NVIDIA's stock price surge, as all AI applications require such infrastructure. However, most users and customers do not need to delve into this layer.
• The middle layer consists of tools built on various large models, which is the level most likely to be used by enterprises, leveraging the capabilities of different large models to meet overseas or other business needs.
• The top layer consists of applications, such as the products we personally use like Doubao and Kimi.

In terms of construction, we have various choices including open-source models, closed-source models, ChatGPT, Google, etc. In the field of large models, there are almost daily breakthroughs and changes in rankings, and the market landscape is changing rapidly. It's similar to last year's "hundred model battle" in China, but in the end, there may be no more than ten that survive, and in the future, there may only be three or four left, which is akin to the development trajectory of the electric vehicle industry.

For ordinary enterprises, when choosing models or building applications based on generative AI, it is necessary to strike a balance between three key factors: accuracy, performance, and cost.

Let me explain with specific examples:

1. Accuracy: For example, in translation scenarios, the difference between 99% and 95% accuracy will have varying levels of user acceptance.

2. Performance: Different scenarios have vastly different requirements for response speed. For instance, in a shopping scenario when conversing with customer service, a one-second response time is acceptable. However, in a gaming scenario, especially during battles or team play, if the command "charge" is delayed by one second before being understood, the player may already have been defeated.

3. Cost: This is especially important in B2B businesses. In practical applications, if it costs $10 to achieve 98% accuracy, or $1 to achieve 97% accuracy, most enterprises will choose the latter.

This also explains why not all B2B companies must use ChatGPT—it may be the most expensive option, while there are many other models on the market that may perform equally well or better but at a lower cost For enterprises, the most important thing is to recognize that there is no need to invest a lot of energy in developing exclusive large models, as this is not suitable for all companies. The key task is to make full use of existing data while changing the perception of data in various business departments, as well as the corresponding management mechanisms and security monitoring mechanisms.

In the era of large models, data openness may be more valuable than strict restrictions, which requires a shift in mindset. This does not ignore security but aims to provide more data to improve the accuracy of models and reduce the phenomenon of "hallucination." The key is to establish effective security control mechanisms at the data usage end to ensure that information is only displayed to appropriate personnel. Therefore, when determining principles, data-related processing mechanisms are particularly important.

The speed of corporate transformation is accelerating, and both the composition of personnel recruitment and corporate culture construction differ significantly from the pre-AI era. We need to deeply integrate generative AI into corporate culture.

For ordinary enterprises, employee empowerment requires: 1. Cultivating employees' ability to master prompts; 2. Cultivating critical thinking to identify the limitations of the pre-AI era; 3. Enhancing each employee's awareness of generative AI.

For example, Amazon Web Services in China requires all employees to pass generative AI certification, including all positions such as operations, marketing, front desk, assistants, and clerical staff.

In the current era, I recommend that enterprises adopt various methods to promote employee exposure to and use of AI: first, introduce high-end talent, which, although costly, can quickly enhance capabilities; second, systematically conduct internal training; third, introduce new application scenarios for employees to practice. Regardless of the method used, the ultimate goal is to deeply integrate generative AI into the corporate DNA. There are many specific methods to enhance skills, such as developing systematic training programs that allow employees to experience the convenience and efficiency brought by AI in their actual work.

Some enterprises have adopted good practices, such as holding hackathons within the company. Employees from different departments are encouraged to propose creative ideas, and then use generative AI technology to realize these ideas, followed by a competition for evaluation. This competitive learning approach not only stimulates employees' creativity but also allows them to deeply appreciate the practical application value of AI technology.

It is important to emphasize a key concept: responsibly building generative AI applications. I often conduct an experiment: I ask the same question to two different domestic AI chatbots and often receive different answers. Only through personal searches and verification can I determine which answer is accurate.

This "hallucination" phenomenon is still prevalent in mainstream AI apps. For example, last month (November), when I inquired about the "2024 TGA Game Awards winners list," one app fabricated and listed the winners, while another app accurately responded that the awards ceremony would take place on December 12 but could provide the announced nominees.

For enterprises, especially B2B companies or those involved in actual business and financial transactions, such errors are absolutely intolerable. This is because it may directly affect business decisions and customer trust, leading to actual economic losses and damage to reputation In terms of business risk assessment, we have established a three-tier risk assessment system to evaluate the risk levels of different scenarios such as marketing promotion, financial reporting, operational contract drafting, and predictive analysis. When enterprises begin to apply AI, they should prioritize scenarios with the lowest risks and gradually expand to more complex application scenarios.

Of course, this risk assessment system needs to be adjusted according to the characteristics of the enterprise. For example, B2B enterprises may pay special attention to the risks of contract accuracy, as contract errors can directly affect business relationships; B2C enterprises may place more emphasis on the security of personal privacy data, which relates to user trust and compliance requirements; for supply chain enterprises, the accuracy of supply chain maintenance and forecasting may be the most critical, as this directly impacts inventory management and operational efficiency.

Each enterprise needs to establish risk assessment standards that align with its own business background and industry characteristics, starting from the most controllable risk scenarios.

Based on our past experience supporting numerous enterprises, especially those going overseas, successfully implementing generative AI projects requires following six key steps.

The first is to gain recognition and support from the executive level, which is extremely important. This involves several reasons, one of which is data, especially cross-departmental data. Taking a gaming company as an example, relevant data is usually scattered across multiple departments: the user growth department holds user behavior data, the game operation department where producers work holds game content and operation data, and the R&D department holds data related to technical implementation. These departments typically operate relatively independently, and if any one department's data is missing, the effectiveness of the AI application will be greatly reduced. Without top-level promotion, it is difficult to break down these inter-departmental data barriers.

The second key step is to form a diverse interdisciplinary team. This team needs to understand both the business and AI models. When forming such a team, enterprises usually face two choices: one is to build a team from scratch by recruiting AI experts and algorithm engineers, but this approach is very costly, especially as the salary levels of AI talent in the market continue to rise; the other is to leverage technical support and architect services provided by professional companies like Amazon Web Services. We have observed that most clients tend to choose the latter, as generative AI projects often start in the idea validation phase, and the returns are uncertain. By utilizing the existing technical capabilities of the platform, they can quickly validate the feasibility of ideas while controlling costs. This approach allows enterprises to flexibly carry out AI innovation without incurring excessively high labor costs.

The third key step is to implement some protective measures, such as those related to data security and risk assessment mentioned earlier.

The fourth step is to deeply integrate generative AI into the corporate culture, because only with comprehensive and full employee embrace of AI can qualitative changes occur within the enterprise.

The fifth step is repeated testing, which is particularly important. This comes from our profound experience working with clients; even when collaborating on generative AI projects with large clients like Tencent, miHoYo, Trip.com, and Dewu, the testing phase usually requires 4-6 months. This long testing cycle is primarily to find the best balance among performance, cost-effectiveness, and accuracy For example, in terms of accuracy, it is necessary to repeatedly test performance in different scenarios; in terms of performance, it is essential to ensure that response speed meets business needs; in terms of cost, the economic viability of long-term operations needs to be assessed. Another reason for extending the testing period is that during these 3-4 months of testing, new models and more favorable cost structures often emerge in the market, and the team always hopes to keep pace with technological developments and continuously explore new possibilities.

But I want to emphasize one point: model iteration is endless, and I advise companies not to pursue perfection excessively. As long as it is confirmed that the cost-effectiveness of using generative AI has significantly improved compared to the original solution, it can be put into production. If one always pursues a better new model, it may instead miss market opportunities and affect business development.

The sixth step is to maintain a certain level of vigilance. Since models may produce "hallucinations," we need to establish corresponding limiting mechanisms. There are now many technical means to achieve this, such as blocking hallucinations, filtering sensitive words, and setting prohibited responses. These tools and software can help companies mitigate risks and ensure the reliability of AI outputs.

Now, whether or not it involves overseas business, every CEO is pondering the same question: What is the generative AI strategy of the enterprise?

If everyone present is thinking about this question, or has started to rethink this question through today's discussion, that is a great start. This question concerns not only enterprises but also extends to every family. Many people ask me, "Should children still learn programming?" "Will my job be replaced?" These are anxieties in the context of changing times.

Indeed, we see that many companies have "streamlined" a lot of personnel after applying AI. Because AI is best at handling repetitive tasks, the saved human resources should be redirected to more valuable work such as business expansion and product innovation, and consider how to let AI bring more value to business expansion and products? This is not just simple personnel optimization, but rather thinking about how to leverage AI for the overall transformation and upgrading of the enterprise.

Returning to today's theme of going overseas, to be honest, under the current context of Sino-U.S. relations, Chinese enterprises do face many challenges in AI development, whether in terms of computing power acquisition or the use of large models, which are subject to certain restrictions. However, our overseas products must compete directly with European and American companies. Without a sound AI strategy, it is very likely that we will already be behind at the starting line.

This raises a key question: As an overseas enterprise, if we are at a disadvantage at the starting point of AI application, how can we change this situation? This requires each enterprise to think deeply, not only considering technical feasibility but also comprehensively weighing commercial potential.

A new wave of going overseas is undergoing fundamental changes. From the right side, our original path of going overseas was from 0 to 1, going through data analysis, market research, and innovation, and then through ecological cooperation. Now, we can leverage generative AI, cloud computing, data-driven approaches, security guarantees, and ecosystems to support going overseas. This is completely different from the "gamble" or "treasure hunt" strategies of Chinese enterprises three or four years ago—back then, it might have been betting on a game becoming a hit or expecting a certain product to sell explosively on the Amazon platform. This luck-based treasure hunt model for going overseas is no longer suitable for the current market environment Now, we are shifting towards an "integrated win-win" overseas expansion model. Companies can position themselves as global companies, rather than just overseas enterprises. It is no longer a simple overseas strategy, but a comprehensive consideration of global operational strategies.

In this transformation process, cloud services can provide infrastructure support for enterprises at various global nodes, ensuring stable business operations, while generative AI can bring new innovative opportunities.

I hope everyone can fully leverage the advantages of generative AI to create products that truly meet global market demands, enhance employee work efficiency, optimize overall corporate operational efficiency, strengthen market analysis and insights, and improve the quality of customer service.

All these aspects can be substantially improved through generative AI. For overseas enterprises, in the face of various current challenges, whether it is data security, market access, or customer service, generative AI can be used to address them. I hope that with the assistance of generative AI, your next product can become a true market hit.

Thank you all!