Data strategy


I lead a taskforce to define and measure our cross-business KPIs, aggregate that data into an MVP dashboard and increasee the data-reliance at EPUK.

This involved:

  • Defining the UK’s strategic KPIs
  • Building an MVP dashboard to aggregate them
  • Increasing the data-literacy of the wider UK team
  • Certifying as a Qlik data analyst and training myself on how to build efficient data models

This project was no mean feat. In the words of the consultant data strategist at the end of the discovery phase, “this buisness is perhaps the most complex I’ve ever come across”.

Skip to: 💡 Discovery process 🎯Strategy


This project is perhaps best thought about in two distinct phases of work; the initial discovery and proof-of-concept, then the exectuion of the longer-term strategy.

Data discovery and prood of concept

Working with data has always been a core part of my method; watching the numbers go up on my 2023 running results graphs was a key motivation for achieving the goals. In my tenure at EP (and as part of We Got POP, prior to acquisition) I have been responsible within my team for collating and reporting our product adoption and customer engagement metrics - but this project required a much more considered approach.

The UK team was made up of four businesses that had all been recently acquired and asked to integrate - each with their own ideas about what was important. As such, the reporting was a bit of a mess. Some teams had fairly mature practices (eg. finance, producing a weekly report) and others had nothing at all.

We needed help to prioritise the right metrics going forward. The CTO gave my boss and I a budget and I was asked to pull together a team of contractors to assist with both the strategy and execution.

The initial brief:

  1. Work with leadership stakeholders from across the business to define and prioritise the strategic KPIs across all teams
  2. Conduct a data discovery exercise to determine which data we have access to (and where we might need to change our processes)
  3. Construct the data model(s) and build an MVP dashboard that
  4. Define a roadmap for all of the data points identified

We hired two consultants, a strategist and a data engineer, then got to work:

  1. We engaged leaders from across the business to discuss and prioritise the most important metrics for their individual function (as well as information they needed from other teams)
  2. We created a metrics tree, mapping how all of the subsequent KPIs related to the north-star metric
  3. We built two data models (one for each business line, Casting Portal and Production Portal) that aggregated the KPIs into a uniform reporting model
  4. We constructed a static prototype dashboard using Excel and Qlik
  5. We ran a roadshow, engaging the leadership and wider UK team to demonstrate what we had achieved

Within 10 weeks, we had completed the brief and the business was clearly hungry for data - but our prototype took considerable manual effort to update and lacked functionality that would enable the team to self-serve their most common follow-up questions.


Upon completion of the initial phase, the Managing Director of the UK arm tasked me with the job of executing the next stage of the project. This involved:

  1. Getting the data into peoples hands
  2. Changing the wider culture across the business, increasing the reliance on data

Building the MVP report

  • Establishing processes within teams to get the data we needed
  • ETL into Domo to aggregate the data and normalise the reporting dates
  • Prototyping the dashboard (a summary of KPIs and )

Changing the culture

  • Setting up monthly check-ins with leadership to ensure they have access to the answers they need
  • Baking the use of the dashboards into key processes with teams
  • Publishing regular deep-dives to engage wider teams with