It would be natural to tackle this project with traditional business intelligence (BI) tools — products like PowerBI, Tableau or Chartio — which is what WVI expected we'd recommend.

However, as with all our projects, Suru conducted a customer-oriented research and planning phase which challenged WVI's assumptions and highlighted the needs of the users, not the features of the technology.

Suru recommended an alternative approach using open source technologies, containerisation for robust deployments and a modular architecture. This gave WVI the flexibility to fine-tune how the system works for end users and created a best-in-class experience.

WVI had some initial design ideas for what visualisations they thought would be useful to their customers. Suru critiqued and improved those designs by validating them against real customer data. In many places we developed better ways to visually communicate so the data told the right story and helped end users — mostly agency leadership — understand the complexity on the ground and glean insights from their data.

Suru also worked with WVI to approach these new data visualisation capabilities in an agile way, breaking up their long term vision into smaller pieces which could be independently released. This approach, along with Suru's modular system architecture, allowed WVI to release a few components early and then test demand and usage patterns before progressing to the other areas, guiding their investment.

With numerous on-premise instances of the LMMS product deployed in countries around the world, rapid iteration is a core challenge. Suru's proposed architecture involved cloud-deployed components tightly integrated into the LMMS experience to allow WVI to update to the platform in days rather than months, but to do that the application needed to be centralised.

WVI couldn't simply read data from the in-country instances — doing so would break data privacy laws, not to mention adding security and operational uptime risks — so Suru designed the system to have in-country components that strip PII and other information, pre-aggregate at a low level and then transfer entirely anonymous data to the central data visualisation system.

Suru's architecture also increased the velocity of future development. The system's various logical components are cleanly separated so they can be developed, swapped out and scaled up or down entirely independently from each other.

This flexibility brings real value to organisations such as World Vision who are needing to respond rapidly to the changing landscape of their delivery. Having technology that is inflexible in such a world increases the risk of not being able to deploy useful tools and hampers charities' effectiveness.