Haidilao’s 6-step recipe for successful digital transformation with OceanBase
Photo credit: Haidilao US
Rich, flavorful broth. Fresh ingredients. Staff turning a lump of dough into springy noodles in a skilful dance. Excited diners, eagerly waiting to slurp up the delicious goodness.
This is the top-notch dining experience one gets at Haidilao, one of the world’s most popular Chinese hotpot restaurant chains. And for Haidilao members, this is dialed up through its membership program.
“Patrons can access their accounts on in-restaurant iPads upon being seated to see what exclusive vouchers and freebies are available to them,” shares Yao Xiangyu, chief information officer of Super Hi International, the company behind Haidilao’s operations outside of Greater China.
However, while this membership system has been a key element of Haidilao’s operations ever since it expanded outside the region in 2012, it had to undergo a major transformation to become what it is today, overhauling its backend with the support of distributed relational database provider OceanBase.
Too hot to handle
In 2018, Haidilao moved to a digital system for its memberships. However, as the company expanded into more markets and its member numbers increased, it found that its existing database system couldn’t handle the load.
“We experienced system instability and slow response time, which affected daily operations especially during peak hours,” says Yao.
Yao Xiangyu, chief information officer of Super Hi International / Photo credit: Super Hi International
Additionally, the way the system was set up proved to be taxing on the firm’s IT team, as its maintenance was costly, inefficient, and very challenging.
“The original system design assumptions and database adoptions became a bottleneck,” Yao shares. “We needed a more robust, scalable, and cost-efficient solution.”
In a bid to address this, Yao and his team decided to go back to the basics and revamp the entire database solution.
And much like how an empty hotpot represents an opportunity to craft a meal to one’s liking, finding a new solution meant that Haidilao now had the opportunity to put together a personalized database, working with OceanBase to do so.
For those who want to recreate Haidilao’s journey, here are the steps.
A recipe for digital transformation
Ingredients
- 1 CIO determined to make a change
- Unsatisfied engineers looking for something better
- Database experts
- A database project team
- An OceanBase database solution
Instructions
1: Determine the key features you want
Haidilao wanted to completely overhaul its database, so this was an opportunity for the company to build something that would meet its long-term needs.
“We had a long list of criteria: We needed something that could handle massive amounts of data without archiving,” explains Yao. “We also needed something that supported global indexing, as well as high data compression and automatic fragmentation management.”
Haidilao also wanted to avoid being locked in with a single cloud computing platform, as it needed to use this database across different cloud providers in different markets.
2: Identify the right database to use
With these features in mind, Yao and his team began searching for the right database partner to support Haidilao’s needs. After much research, they decided on OceanBase.
Aside from meeting the team’s exacting criteria, OceanBase also has the backing of Ant Group, as well as a proven track record of powering businesses such as Alipay, GCash, and Trip.com. It’s also supported major ecommerce events such as Alibaba’s 11.11 Global Shopping Festival for over a decade.
“OceanBase is definitely a prominent database system in this industry nowadays – they showed us case studies from other customers, and we felt that it has really good enterprise experience,” Yao shares. “It’s extensively proven to be able to handle massive amounts of data and ensure stability and operation continuity during peak hours, which is what we needed.”
3: Speak to the database experts to figure out what you need to do
Once Haidilao decided on OceanBase, Yao and his team sat down with some of OceanBase’s experts to determine how to go about this process.
“We had some interviews about the Haidilao membership business, about the rules and the way it worked, and they also showed us some inherent features of OceanBase,” says Yao. “From there, we formed an optimization and incorporation plan for the database together, conducting mutual reviews to ensure that we had the same understanding about the business and what needed to be done.”
4: Develop a proof of concept to validate your approach
From there, the next step was to test the process before engaging in a full-scale optimization of Haidilao’s membership database. Together with OceanBase, Haidilao developed a proof of concept (POC) for the optimization process, focusing on critical transactions in order to verify that the database was able to support the proposed schema, queries, and the expected volume of data and throughput that Haidilao needed to manage.
“We used the OceanBase Migration Service (OMS), a one-stop solution to support database optimization, as part of this,” says Yao.
5: Put your optimization plan into action, step by step
With the successful POC, it was clear that OceanBase was able to meet Haidilao’s needs. From there, the firm began the process of optimizing and incorporating all its membership data into this new database, once again tapping on OMS.
6: Success!
With the incorporation complete, Haidilao’s membership system is now operating on a more robust database that can support the demands of its customers effectively.
Good soup
With the optimization done, Haidilao was able to eliminate lag during peak dining periods, reducing the amount of time the system needs to recover if it goes down to less than 30 seconds. This enabled it to handle significant peak customer loads, such as during seasonal holidays, without incident.
The use of OceanBase’s database solution also helped ease the burden on Haidilao’s DevOps team, cutting down on development costs and the effort required to manage the database. Haidilao’s database costs dropped by 50% as a result.、
Photo credit: Haidilao Singapore
This new database has also made a difference to Haidilao’s marketing efforts, increasing computing power for real-time analytics by as much as 45%, enabling the firm to leverage operating insights more effectively.
“With improvements to this database and the back end, our customer relationship management platform has also improved,” Yao shares. “Marketing campaign information and promotions can be sent to Haidilao customers more efficiently, which has led to a significant increase in new member registrations.”
Haidilao currently has around 4 million members across 112 restaurants in 12 countries outside of Greater China. Yao is confident that this new database system, with multi-cloud architecture to support different cloud computing providers in different markets, will support this growth.
“We just opened our first store in Dubai in March this year and will be expanding into the Philippines very soon,” he shares. “We’re sure this membership system will be able to hold the load that we need in the future, supporting our international business development and growth.”
OceanBase Database is a distributed relational database developed by Ant Group. Supporting Alibaba’s 11.11 Shopping Festival for over a decade, it adopts an independently developed integrated architecture, which encompasses both the scalability of a distributed architecture and the performance advantage of a centralized architecture. To date, it has helped over 400 customers across industries upgrade their core systems.
