Application Exploration of TiDB in Chain Fast Food Enterprise丨Massive Transactions and Real-time Analysis

For many modern people, a good-tasting, energy-packed, reliable, and quick-and-easy burger, fries, and fried chicken is a good choice for a busy work day. What's more, the salt, sugar, fat and carbohydrates rich in fast food are also more likely to make people feel satisfied. But when you enjoy the happiness brought by fast food, have you ever thought about how fast food companies can buy coupons online, pick up goods offline, place orders at home, deliver to your door, and use mobile phones while operating hundreds of stores. A series of high-level digital operations such as ordering and store pickup?

Since the epidemic, the overall loss of the catering industry can be said to be heavy, but the first to recover from the blow are the major fast food giants with the largest number of stores and the most complex operations, or more specifically, those with more investment in DTC. and accumulated fast food businesses. Taking Double Eleven last year as an example, a number of top fast food companies broke through the whole-day sales of last year's Double Eleven in less than 12 hours, and the business growth was gratifying.

So what is the sacredness of this DTC that can make many fast food companies flock to it? DTC (Direct to Customer, that is, "direct-to-consumer") is a business model that focuses on online and takes direct-to-consumer as its core goal. Combined with the domestic consumption environment, DTC emphasizes the use of APPs, mini programs or other online channels to provide all consumers with a 360-degree all-round user experience. Compared with the traditional business model, DTC's advantages are mainly reflected in being closer to consumers, paying attention to consumer behavior, and paying attention to consumers' lifestyles. At the same time, by crossing various intermediate links in the business process, enterprises can also more easily and directly exert influence on consumers and evaluate the effect, thereby making the business operation more efficient, lower in cost and more direct in effect.

Multi-channel transactions for a large number of users Of course, like any advanced business model, the DTC strategy also has a high threshold for implementation. The core advantage of DTC lies in its directness and efficiency, which is in line with the consumption patterns of ordinary consumers in the Internet age. However, to achieve these effects, enterprises must not only collect user behavior information aggregated from massive channels, but also do their best while completing various business operations. It is possible to quickly analyze the full amount of data in real time, and form a marketing or service plan for thousands of people and on-demand based on the analysis results.

For fast food companies with tens of millions of users and tens of millions of daily orders, the resulting data pressure can be imagined.

Take a large-scale fast food chain brand that operates in more than 100 countries around the world, has tens of thousands of stores, hundreds of millions of registered users, and handles tens of millions of orders every day. service, and place it at the entrance of the applet, so that customers can get a thousand and one thousand service menu. At the same time, enterprises also need to use big data to analyze the usage and stay of different modules in small programs or APPs to achieve continuous optimization of applications. Through data analysis of consumer behavior, companies can also more accurately understand a series of information such as what consumers bought, how much they spent, how often they came, when they received coupons, and which restaurants they often spent. , so that companies can use big data and algorithms to infer which consumers can be attracted by various marketing activities in which locations.

To achieve such business results, the entire set of data infrastructure that once centered on the public cloud was not ideal in terms of innovation, flexibility, reliability, and initiative for change. Therefore, the company abandoned the original public cloud architecture and rebuilt a private cloud architecture. After a period of operation, this architecture not only achieves a lower TCO than the public cloud, but also solves problems such as the construction of massive applications, rapid iteration, and grayscale release through a powerful software stack. More importantly, the company has solved the huge demand for online transactions and real-time data analysis corresponding to hundreds of millions of users and tens of millions of orders per day under the DTC strategy through the new architecture of the database, and can respond with a lighter body. The amount of data is growing rapidly.

In the logical architecture of the data platform, the OLTP function of the TiDB database serves scenarios such as enterprise orders, payment and supply chain, while the HTAP corresponds to the key real-time reporting and data analysis requirements in the DTC strategy. On the other hand, under the background of the continuous accumulation of user behavior and the continuous growth of the number of users and business volume, the amount of data to be stored in the database will inevitably increase rapidly, and the database must meet the requirements of the first two functions. The rapid growth of the company provides a simple, effective and low-cost response. To meet these requirements at the same time, it is a great challenge for the database.

In the past experience, whether it is an open source database represented by various types of MySQL, traditional commercial databases, or cloud databases launched by major CSPs, it is difficult to escape the sub-database when the amount of data is huge and growing rapidly. The fate of the sub-table. Not to mention the huge workload and management problems corresponding to the operation of sub-database and sub-table itself, the query and analysis difficulties caused by cross-table and cross-region alone will greatly reduce the actual effect of DTC.

Therefore, choosing a new underlying database that can operate in a distributed and cloud-native manner, and getting rid of the performance, operation and management drawbacks caused by sub-database and sub-tables, has become a key step for fast food companies to implement their digital transformation strategy. In the process of DTC strategy practice of this international fast food chain giant, TiDB finally took on the important task with its various advantages.

TiDB is an enterprise-level open source distributed database, which has a natural high degree of compatibility with cloud architecture. It can meet the growing data volume and performance requirements of enterprises through the increase of cluster nodes, and avoid the huge workload caused by sub-database and sub-table. , operational risk and high follow-up operation and maintenance costs. At the same time, TiDB's open source features and active technical community also enable enterprises to achieve fast, low-cost functions and business innovations according to changes in needs and applications.

On the other hand, as a database that supports the HTAP function, TiDB can provide high-performance real-time analysis capabilities on the basis of meeting the needs of online online transactions, helping enterprises to use a set of database architecture to meet the needs of real-time reports and large-scale data under the DTC strategy. The demanding needs of analysis. Through the innovative row-column storage isolation mechanism, TiDB can perform real-time report aggregation and rapid analysis without affecting OLTP business performance, allowing enterprises to focus more on data value discovery and business innovation under the DTC strategy.

After the fast-food company completed the deployment of the private cloud infrastructure, the PingCAP professional service team, in close collaboration with the company's own technical team and ISV, completed the deployment, data migration, and deployment of two TiDB clusters in the cloud environment in only 3 months. Grayscale works online. This deployment speed can still be maintained in the face of massive business systems and huge data volumes, which is obviously beyond the reach of traditional databases.

With the support of the new database, the fast food company has realized second-level data retrieval and analysis of hundreds of millions of users, and can quickly evaluate the effect of marketing activities, thereby realizing timely optimization of marketing strategies. By monitoring take-out orders under different weather, different promotions, different products, different time periods, different dates, etc., the fast food company can better predict the category and sales changes of the store's take-out orders, which is convenient for the enterprise to control raw materials, personnel, Data-driven refined management and regulation of equipment, etc., to improve store operation efficiency and effectiveness.

At the operation and maintenance level, TiDB, which does not require sub-databases and sub-tables and can meet the multi-dimensional needs of users, not only runs stably, but also greatly reduces the workload of database operation and maintenance management. With the support of PingCAP's professional service team, the fast food company has also successfully survived the peak of business brought by the Double Eleven and other major promotions.

TiDB’s ability to provide elastic expansion during peak dining hours. Through the DTC strategy practice supported by the new database, the company can fully understand where customers are, who they are, what they like, and why they come, for new product development, marketing, restaurant Operations and even supply chain management provide more precise insights and enable companies to make more real-time business decisions.

In the digital age, data is the core asset of an enterprise. As an intangible virtual asset, every exploration and utilization of data value by an enterprise is inseparable from the support of the underlying database system. That's why most companies' digital transformation strategies use data as the entry point and the core reason.

Under the macro background of data explosion, rapid growth of business volume, and accelerated innovation, whether the database can bring a new experience of functions, performance, and operation and maintenance management to enterprises through features such as distributed, horizontal expansion, and HTAP is an important issue for enterprises. The key to successfully completing the transformation and realizing business innovation.

The "Global Database Management System Software Market Analysis Report, 2021" released by IDC also pointed out that the trend of "Direct Customer Engagement" represented by the DTC strategy in the fast-moving consumer sector is driving digital applications to high-performance, data-consistent, elastic-load databases. system requirements. In this case, PingCAP's TiDB can well meet the DTC strategic transformation needs of fast food companies with large data volumes, many businesses, fast updates, high transaction volumes, and heavy analysis tasks, laying a solid foundation for in-depth business innovation.

This is exactly the bottom-level solution that many catering, retail, and fast-moving consumer goods companies that are trying the DTC strategy, and even thousands of industries that need to face end users, seek to achieve digital transformation and build a data platform.

Further reading:
Click to see more cases and practices of DTC transformation in the new economy