Boots is an iconic British retailer. Founded in 1840, it has more than 2,200 stores, from large city centre health and beauty stores to small local community pharmacies. It is one of the most recognizable brands on the UK high street.
However, retail is changing quickly, so Boots is also becoming an increasingly digital business that has to pivot to new digital channels. This is a tricky task: Boots’ customers span all age groups; the beauty sector is notoriously fickle; and the pharmacy business demands strict data protection.
“Our goal is to make sure every Boots customer can shop with us in the way they want to shop,” says Steve Westgarth, head of engineering, Boots UK. “That means we need to bring our digital experience much closer to the customer.”
To complicate matters further, the same customer might engage differently while shopping for mascara or a repeat prescription.
Additionally, as a chemist, the past couple of years have put Boots on the frontline of the UK’s COVID response, from testing to giving advice and administering vaccines.
“This is a business that impacts people’s lives in many different ways,” says Westgarth. “In order to know our customer, data is king.”
The challenge for Westgarth’s team is to bring together multiple sources of data, to aggregate quickly, and to present it in a way that makes actionable insight possible.
“We’re looking to make technology choices that best serve the business, our customers and our engineers,” he explains. “We’re creating a learning culture based on continuous improvement.”
MongoDB Atlas on Microsoft Azure is central to Boots’ digital ambitions. It supports transactional, search, analytics and mobile use cases while using a common query interface.
“MongoDB was an obvious choice,” says Westgarth. “Our engineers are familiar with the MongoDB API; it’s become an industry standard in terms of manipulating and managing data. The fact that our engineers enjoy working with MongoDB is a big thing for us. We want to make the engineering experience enjoyable.”
Database technology, he continues, is an area where standardization is obvious. Supporting multiple databases is not where Westgarth wants to be focusing resources: “Where possible, we don’t want to fragment the technology stack. We can do everything we need to do from within MongoDB, which means we don’t have to bring in third-party applications.”
Standardizing on MongoDB is part of a wider project to simplify Boots’ infrastructure stack. As expected with a long-established business spanning multiple market segments, Boots has built up an assortment of infrastructure components and applications. This is unnecessarily complex, he says.
“We don’t want our engineers wasting time learning the skills to work on unsupportable technologies. Their energy should be focused on delivering value. We need to find simpler ways of working.”
Understanding the real-world consequences of database performance
Selecting MongoDB, Westgarth continues, was not a response to crippling pain points. It was a strategic decision to connect Boots to the industry-leading database platform.
“In effect, the decision creates a centre of excellence around MongoDB. It enables our engineers to focus on one industry standard. This removes duplication and distraction.”
The impact is an environment where engineers enjoy work, can produce great work and, in turn, attracts the best new talent.
“By really engaging with the technology platform we’ll be better able to leverage the roadmap of new features,” he says.
By decluttering the database landscape, Boots engineers have more time to, in Westgarth’s words, “understand how our applications impact the real world”. For example, what a processing delay might feel like for a Boots customer at the online point of purchase, or the impact of a database query failing as someone waits for a test result.
“Engineers need to be aware that the performance of database technology can have real-world consequences, particularly for a business that works in the healthcare and pharmacy sectors,” Westgarth adds. “There may be revenue loss, reputational damage or unnecessary stress caused by a delayed test result.
“The database is critical to our customer experience, using MongoDB means we have robust, reliable technology supporting us.”
The most critical of applications running on MongoDB is Boots’ order management system (OMS). This routes orders from boots.com to the company’s many warehouses. The OMS would be business-critical in any normal year; through COVID, with online sales rocketing, its effectiveness was existential.
“We experienced Black Friday levels of orders in March. This was unprecedented. Usually, we’d work for months in advance to prepare for Black Friday. Our warehouses risked being overwhelmed,” Westgarth explains. “Our OMS is the glue that holds the business together.”
Boots built an additional service layer, allowing customers to track an order or request a change of delivery address or branch collection. This all took place against a raging pandemic, health uncertainty and sky-high customer expectations around online retail. MongoDB was instrumental in managing this complexity, and became part of a new, microservices application development approach.
“MongoDB means the scrum teams can iterate quickly, especially when it came to building an OMS to move orders. Adapting the OMS meant we could move orders between warehouses, or take stock from quieter, out-of-town stores,” Westgarth says.
“The OMS application has had to scale quickly and it was vital that MongoDB scaled alongside. We could test performance, and plan for new scenarios.”
He says the way the business has responded to COVID, and the rapid progress made across digital touchpoints, bodes well for the future.
“The way we’ve had to work during COVID – quickly, with agile development is how we’ll have to work as we accelerate our digital transformation. I expect us to be making more use of MongoDB in the next 12 months.”
Steve Westgarth, Head of Engineering, Boots UK