https://cynozure.co.uk/wp-content/uploads/2019/04/tungsten-case-study.jpg

Refreshing Data Architecture at Tungsten Network

Tungsten Network helps global businesses manage and control their supply and value chains. Based in London, it has developed a business transaction network, plus a suite of services around supply and value chains. It supports 300,000 suppliers with over £164 billion worth of transactions in the 2017-18 financial year. That support encompasses e-invoicing, financing, analytics, procurement, and invoice flow. This data architecture case study will explain how Cynozure helped Tungsten achieve their data-driven aspirations.

Background, aims, and objectives

When it approached Cynozure, Tungsten Network was an 18-year old business hindered by legacy data infrastructure. This failed to support Tungsten’s ambitions to become more data-driven, efficient, and to unlock the full potential of their valuable customer data

Cynozure was tasked with untangling a complicated data architecture – without any of the original developers, who had since left the company. Tungsten also wanted help in changing the mindset around data use, to transition from a role-based customer viewpoint, to a product-based one that would improve cross-selling.

What we did

1. Audit and document the legacy data systems

Tungsten’s biggest challenge was its legacy data architecture, built in modules over the course of 18 years. No original data architects were still employed by Tungsten, so nobody was on hand who fully understood the data architecture or the processes driving it.

Cynozure began by creating and documenting a map of the existing data architecture, and in parallel assessed the ability of the system to adapt to fulfill Tungsten’s future goals, such as:

  • Improving cross-selling opportunities
  • Growing revenue
  • Establishing new product lines
  • Refining efficiencies in processes and payments

 

Tungsten had previously created a program in-house to try to fix its data architecture. However, there was a lack of specialist internal skills to help them create new conceptual data architecture that would grow with their business. There was also confusion around data terminology – an ‘account’ for example, could mean an individual, a supplier or a ‘product’ depending on the department in question.

 

2. Map out and implement a data architecture for the future

Cynozure developed a strategy exercise that was heavily focussed on creating a data architecture that would fulfill Tungsten’s future goals.

Several workshops were set-up with Tungsten employees to better understand how the business worked. Cynozure used feedback from these sessions to create a data model that covered all the different use cases within the business. From this, they developed an ‘ideal’ data strategy that took into account how the Tungsten business operated, how data could support this, and that wasn’t constrained by the existing architecture. This took around one month to complete.

The ‘ideal’ data model was then translated into the Salesforce software suite. Salesforce was integrated with the existing architecture and Cynozure worked to ensure synergy between the two systems. In this way, Tungsten benefits from an ideal data model formed in Salesforce, but with its legacy architecture intact and not incurring any data loss.

3. Data cleanse and document processes

To help the business fully understand the legacy systems, Cynozure helped Tungsten’s in-house team to learn the newly documented processes via an internal wiki. In this way, the organisation is future-proofed and new employees can easily understand and access the system in a logical and efficient way.

During self-registration, many clients were unable to duplicate details like VAT numbers on the system. This meant that they would input incorrect data (such as extra characters) in order to be able to register. Cynozure worked with Tungsten’s team to clean the data and get it to a point where it could be combined with external data sets. This also prevented the new Salesforce system being affected by bad data.

Additionally, Cynozure helped to evolve Tungsten’s mindset around data. Previously, buyers and suppliers were kept separate and offered different products based on their assigned roles. Instead, a new product-orientated approach was proposed. This was vital in helping Tungsten to cross-sell.

Better data architecture enables new business opportunities

Tungsten now has a fully documented data architecture that will enable it to make and achieve future data plans. They are now looking to deliver other pieces of major infrastructure, such as AWS, amongst other data goals.

Thanks to the data cleansing work, Tungsten was able to uncover new business opportunities and win more business from existing clients by cross-selling; plus the efficiency of processing and payments has been improved.

Finally, documentation and the creation of the internal wiki ensured that when a developer leaves, the knowledge of Tungsten’s data architecture is not lost to the company. Plus it prevents legacy data issues from reoccurring in the future.

Cynozure played a critical role in helping us understand Tungsten’s data architecture, and setting it up for the future, along with changing the wider data culture. There’s been an added benefit in being able to cross-sell and upsell to more of our clients, as well as target parent organisations. There’s greater knowledge of Tungsten’s system, consistent terminology, and everything runs much more smoothly now.

David Williams – CFO, Tungsten Network

You can find more details on the organisations that Cynozure has supported on our Work page, in addition to our white papers.

Data Architecture Case Study