New cloud data platform increases traffic


Missguided (NEW)_hero illustration_large
The client

About Missguided

Founded in 2009, Missguided is a UK-based online, multi-channel fashion store for women. Their mission is to empower females globally to be confident in themselves and be who they want to be. Informed by their customers along with global influences like social media, street style, and popular culture, they have created a destination that delivers and encompasses everything it means to be a girl on the go in the world today.

Missguided had long term plans for expansion with a business strategy focussed around growth - increasing revenues and profit margins alongside the conversion rate of visitors to its website. The goal was to remain extremely competitive in the low margin world of online retail.


Key Outcomes

We worked in partnership with the team at Missguided to make the most of a new cloud data platform, which greatly increased their pace of feature releases resulting in increased search traffic and site rankings.

  • Data 9TB Over 3 years of data migrated to the cloud
  • Scale 26m Objects contained within the new AWS data lake
  • Pace 300% Increase in feature releases every week
The Challenge

Customers weren’t converting

Having recently migrated to AWS in order to solve scalability issues and deal more effectively with traffic spikes, Missguided identified that recent investment in advertising campaigns was not yielding the expected increase in customer conversions.

At the time Missguided was using an on-premises SQL server as a data warehouse, combining data from SAP, Magento, Google Analytics, and Facebook ads. Unfortunately, the system was constantly running out of memory and required continuous manual upkeep.



Building a new data platform in the cloud, and new capabilities within Missguided’s team

Missguided took a strategic view that to keep up with its competitors it needed to better engage with its website visitors and introduce a more personalised experience.

It also needed to update its website and app more often with offers and features. Realising it needed technical guidance to address these issues, Missguided turned to AND Digital for help to build a new data platform in the cloud. Our focus on building long-term capability within their teams so that they could effectively manage their own infrastructure made us a compelling choice for Missguided.

The Delivery

Migrating multiple data sources to AWS

Working in partnership with Missguided we combined multiple data sources into an AWS data lake, we then cleaned and visualised the data, while building the internal capabilities of Missguided's teams.

Cost-effective connectivity between data sources

We plotted out all of Missguided’s data sources to figure out the best and most cost-effective ways to connect them. It was established that there was a lot of value in joining the behavioural data of customers from Google Analytics with transactional data from Magento and other sources to get a clearer picture of the full customer lifetime value.

Missguided had more than 84TB of data that needed to be combined onto an Amazon Simple Storage Service (Amazon S3) data lake before it could be properly analysed. The data was stored on multiple datasets of various sizes and in different formats. To be cost efficient, we used several different AWS services to transfer the data sources into the data platform.

Transferring, transforming and cleaning the data

The largest data sources were transferred using Amazon EMR, a data platform for running large-scale distributed data processing jobs in the cloud. We used Python scripts to pick up data from one source, transform it, and then drop it into the main data lake. More than 3 years of data—over 9TB—was transferred this way for minimal cost.

For smaller datasets the team used a combination of AWS Step Functions, a low-code, visual workflow service, with AWS Lambda, a scalable, serverless compute service for code-running. For everything else, we used AWS Glue, a serverless data integration service that combines data for analytics, machine learning, and application development.

Querying and visualising the data

Once all the data had been properly stored, cleaned, and prepared, Amazon Athena, a serverless interactive query service, was used by the Missguided data scientists to query the raw datasets. The data was formatted in Parquet files, which was chosen over CSV/JSON as it results in faster queries, smaller file sizes, and further reduces spending on Amazon S3. 

AWS Lake Formation provided granular access to the data and properly secured any personally identifiable information. Amazon QuickSight, a serverless BI service, was then used to send engaging visualisations of relevant data to its staff.

Building the capabilities of Missguided’s teams

Throughout the build process we worked in partnership with Missguided’s internal teams to ensure that they were equipped with the skills and expertise that they would need to both effectively manage the infrastructure of their new data platform, and to continue to innovate - driving new value for the business through increased feature releases.

The Result

Driving business growth

Greater productivity and increased performance

We worked closely with the team at Missguided to help them transform their ways of working and their use of data, in a cost-effective way. This has resulted in improved website performance and customer experience.

Visually compelling data delivered cost-effectively

Business analysts from Missguided now have access to visually compelling data from multiple datasets. The 84TB data lake contains more than 26 million objects and allows Missguided to create informed business reports from all its data sources. In addition, as the project used serverless technologies where possible, and only paid for compute resources when required, costs were kept to a minimum.

Coding practices transformed

We provided six technical staff to work with Missguided and, within a few months, had helped reskill the team. Missguided’s coding practices have since been transformed. Its developers now take more responsibility for their code and have been sufficiently upskilled to manage the company’s updated infrastructure.

Increased releases, site traffic, and search rankings

The project has resulted in the Missguided development team significantly improving its productivity. Previously they were managing between three and five new releases each week; they now average 14, an increase of around 300 percent. These updates have helped improve the number of customers who visit the company’s websites and resulted in a better search engine ranking.

Case studies

See our impact elsewhere

More successful digital transformation projects...

Want to learn more about the work we do with clients? Take a look below.

Talk to us.

Get in touch to discuss your digital ambitions.

Contact us