In the era of big models, the three key marketing strategies

With the advent of the big model era, the marketing field is undergoing a profound transformation. Companies need to embrace AI and use data analysis and intelligent tools to improve marketing efficiency and effectiveness. In this article, the author explores how companies can find growth points through precision marketing and data management in the current market environment.

Many brands have expressed similar views to Yilan Business: Marketing has changed in the past two years.

"Last year, the traffic budgets of two of our e-commerce platforms were cut in half. If the ROI remains at the current level, we will consider focusing only on the best-performing e-commerce channels," a beauty brand owner told Yilan Business.

Since last year, it has become the norm for companies to cut their budgets for seeding and traffic. As the mobile Internet has reached a mature stage with users and traffic reaching their peak, companies are becoming more cautious and conservative in allocating budgets for channels. The trend is reflected in the fact that marketing budgets are shifting from seeding to conversion, and companies are paying more attention to "certain business opportunities."

Why do so many companies fail to do marketing well? On the one hand, the retail industry has more channels. For example, the mattress brand Xilinmen has 9 online channels and more than 5,600 offline stores. How can it reach the audience in a refined way? On the other hand, with the return of the era of rational consumption, consumers are more cautious about how to use their "purses". As the incremental market turns into a stock market, how can companies convince consumers to buy their products?

In the final analysis, the starting point for doing a good business is precision marketing, and the starting point of precision marketing is data management. As Peng Xinyu, Vice President of Alibaba Group and CEO of Lingyang, said: If you can't control the data, you can't control the market.

The data control here is not just report data or GMV data, but means that enterprises need to have a clear insight into global user data. In 2024, enterprises must not only control data, but also control AI. They must not only use data for their own benefit and provide a basis for decision-making, but also use data "wisely" to find certain business opportunities in an uncertain era.

1. If you don’t break the data dilemma, you can’t do marketing well

A company trapped in data cannot escape these industry pain points:

• There are many service providers and systems, and people in different systems mostly collaborate through project management, not business and data collaboration. Because there is a lack of links between them, it leads to inefficiency and fragmentation among the group's businesses.
• It is difficult to achieve conversion on one hand. Data is very important but its richness is limited. The concept of data intelligence is beautiful but it is difficult to match business value and find a landing point. How to use global data and AI intelligence in compliance with regulations to help enterprises find new growth points in fierce competition has become a new topic.

To solve these pain points, companies are in urgent need of a full-link marketing tool that not only places all marketing scenarios in the same space, but also has AI capabilities for connecting and processing data.

It is worth noting that IDC released a report today analyzing China's current omni-channel marketing platforms. In the report, IDC rated Lingyang Quick Audience as a leader and affirmed its ability to help enterprises achieve omni-channel marketing through omni-domain data and AI intelligence capabilities.

Lingyang Quick Audience (hereinafter referred to as QA), with its automated marketing capabilities, has become a powerful weapon for a world-renowned clothing brand to break through the data dilemma.

The pain point of operators is that the dormant population on e-commerce platforms such as Tmall is about to be lost, and there is no effective way to activate the dormant population. With the help of QA, operators selected Tmall dormant populations who reside or work within 5 kilometers of 42 stores in Beijing, Shanghai, and Guangzhou, and combined with the intelligent screen light timing capability to push Lingyang Chaoxin at the time when the device predicts activeness, to achieve offline store recall of dormant populations. Compared with the control group, the probability of store visits in the experimental group increased by 18%, the offline conversion rate increased by 35%, and the omni-channel conversion rate increased by 40%. The growth effect is significant.

Wang Chao, director of the marketing digital product operation department of the Galan Group Big Data Center, believes that a brand is actually the sum of consumer relationships. With the data center, enterprises can clearly know where the large amount of market data generated by the business end can flow back for further analysis and result export, and then apply it to multiple business scenarios including marketing, forming a positive feedback loop and truly revitalizing data application.

(Source: 102 Growth Examples: Leaders in Digital Growth)

After solving the data dilemma, the next step in marketing is to gain insight into “three-dimensional people”.

2. If you don’t capture “three-dimensional people”, you can’t do marketing well

In the past, it was difficult for retailers to obtain real and effective C-end user data because the labels were flat.

Why did flat labels work in the past? Because it was an incremental market in the past, and even flat labels could help companies grow. But in today's stock market, the more obvious trend is that companies cannot do marketing well if they can't capture three-dimensional people.

How to understand a three-dimensional person? For example, a person can be pregnant and love sports. A person may also like outdoor activities when buying a Frisbee for a pet. A person cannot be defined by a single label such as gender, city, education level, etc. Everyone is complex and comprehensive.

However, a large number of merchants still face this situation: although the membership base is large, they lack deep insights into member characteristic data, resulting in the existing CRM (customer management system) system being unable to carry out accurate marketing push based on data and can only carry out general push.

The high-end women's clothing brand Lancome is a typical example of achieving growth by "capturing" three-dimensional groups of people through data + intelligent insights.

In order to activate consumers with different consumption characteristics, Lancome Group adopts a multi-brand matrix strategy and fully deploys online and offline business channels. Its women's clothing business segment includes Lancome, Rhine and other brands.

While fellow apparel peers are still struggling with how to handle the complex relationships among multiple platforms, multiple channels and multiple brands, Landis has already begun working with Lingyang to reach out to "three-dimensional people".

Specifically, QA divides the entire contact process into four steps:

1. Labels such as predicted age, predicted gender, and predicted place of residence are used as the basis for stratifying people and goods.
2. Quickly match phone, OpenID, DeviceID and other information to gain consumer insights;
3. Further enrich the application of scenario marketing and customer segmentation such as price range, content, style, channel, etc.;
4. Customize a label system with brand characteristics for long-term use and accumulate unique consumer assets.

"Even with the gift strategy, Lancome knows to provide 'sports equipment' rather than 'maternal and child products' for single women who love outdoor sports, and uses gift preferences to increase users' impulse to place orders. In essence, this shows that they understand consumers well enough," Ling Yang explained.

With the help of Lingyang, Langzi has formed a main public domain platform consisting of Tmall, JD.com, Douyin, and Video Account, and continues to carry out content, services, marketing, and sales in the private domain, achieving a member GMV contribution of 40-60%.

Lancome's tag setting logic is also applicable to all walks of life. For example, in the catering industry. The catering industry often has a demand for people-place tags. QA can help it match the user's permanent residence with the corresponding store location, and match people and goods based on historical purchase records. In addition to pushing to the core group, it may also conduct targeted observations on the group that generates product linkage.

After defining the target audience, QA also pays attention to every detail in terms of precise reach. Yilan Business learned that there is a specific component called "Smart Timing" in QA's automated marketing scenario. When sending, the system will automatically match the time periods when different IDs are active, helping companies achieve more efficient marketing coverage.

After reaching users accurately, companies will naturally focus more on private domain operations. After all, private domain members are often the most promising consumers of a company, and fine-tuning operations for these users is the key to business growth. In this process, AI tools have begun to replace traditional operations and become the main force.

3. If you don’t know how to use AI, you can’t do marketing well

When AI was not yet popular, automatically generated marketing content was always crudely produced, and ranking members’ purchasing intentions was a fantasy. However, as AI began to be widely used in corporate marketing, these problems were easily solved.

A group of companies that were the first to embrace AI have long understood the necessity of using AI tools throughout the entire marketing scenario: they can achieve refined user operations, improve operational efficiency, and accumulate private domain data.

For example, the women's clothing brand Evely has locked in four membership marketing scenarios through the targeted algorithm capabilities of Lingyang Marketing AI, and has effectively managed costs by presetting ROI, adding certainty to operations. In the words of Crystal, Evely's e-commerce operations director, "Lingyang's 'data + targeted algorithm' has made every penny we spend on marketing count as 1 yuan, or even 10 yuan."

So, what do companies, including Evely, need to do to achieve such growth? We have summarized three directions:

• Increase the repurchase rate in the existing market.
• Carry out refined marketing outreach to private domain members.
• Prepare personalized yet standardized display templates for daily marketing and brand marketing activities.

It must be acknowledged that since the current consumer market is a stock market, marketing conversion is naturally more difficult. Therefore, the repurchase rate is increasingly important for the re-growth of enterprises.

Imagine if you are an operations staff and your boss wants you to start preparing the marketing strategy for the Double 11 promotion now. How can you find a way to activate repurchases from old members?

At this time, you would hope to be able to rank the old members who are most likely to repurchase so that you can reach them one by one and improve efficiency.

QA's built-in algorithm model can achieve your expectations.

This is how it is done: QA's repurchase prediction model can train the brand's existing population data. By throwing the model into millions of consumers, it will automatically sort the repurchase tendencies, and then based on the marketing budget and the boundaries of the target population, it will push marketing to people with high repurchase intentions.

Targeting users with high repurchase rates and reaching them accurately are all solved for you at once. It also saves you extra time in designing marketing content. After all, QA's AI raw pictures have rich content ammunition that can be used out of the box.

At the core membership management level, QA can also help companies to be more granular, such as gaining insights into the needs of different private domain members and recommending products with different functions.

Obviously, QA has done its best in automated marketing. No wonder IDC commented: "It provides a full-featured marketing solution."

IV. Conclusion

In the era of big models, marketing strategies must return to the essence of marketing, which is to find the right people, say the right words at the right time.

In the eyes of Yilan Business, in the marketing activities in the era of big models, each business segment is no longer separated, but coordinated and digitalized. This digitalization includes every step of marketing. Departments such as merchandise, content planning, integrated marketing and e-commerce channel sales together form a "large marketing team."

In order to achieve precision marketing, this "large marketing team" needs to have the ability to master data, the ability to define user groups, and the ability to accumulate private domains, all of which are inseparable from digital marketing tools.

It must be acknowledged that with the continuous emergence of "enterprises that take the lead" such as Galan Group and Lanzi, in the future marketing era, we will witness a profound transformation driven by AI and data simplification. In this transformation, only enterprises that prioritize "mastering" data + AI will come out on top.

Li Yan | Author Mu Yu | Editor This article is written by the operation author [Yi Lan Business], WeChat public account: [Yi Lan Business], original/authorized to be published in the operation party, and any reproduction without permission is prohibited.

The title image is from Unsplash, based on the CC0 protocol.