An energy retailer providing electricity and gas to private and professional customers rebuilt its analytics stack from data collection through to data activation. The project combined a full stack audit, a migration to server-side tracking, and a new digital data product for online behavior data, feeding faster insights, better marketing spend decisions, and lower churn.
If your organization is competing on customer acquisition and retention at the same time, and your analytics stack was never built to support both, this case is written for you.
This is how the engagement was scoped, what was built, and what changed once it was live.
The starting point: fierce competition, an analytics stack not built to keep up
The energy market is a tough place to compete. Pricing is transparent, switching is easy, and every point of customer friction is a reason to leave. This retailer needed to sustain its sales targets while also improving digital journeys for existing customers, two goals that put very different demands on the same analytics stack.
The company needed support to leverage its existing analytics stack properly and improve its digital journeys, for both customer acquisition and retention use cases at once. That meant the stack itself, not just the dashboards on top of it, needed a hard look.
Analytics and growth objectives for the project
Two objectives shaped the engagement:
Sustain sales targets by improving the performance of digital channels for customer acquisition
Increase customer satisfaction by improving journeys on digital platforms, and reduce churn
Acquisition and retention were not treated as separate projects. Both needed to run on the same trustworthy data, which is why the work started at the stack level, not the campaign level.
Rebuilding the analytics stack, from data collection to digital data product
Keyrus consultants worked directly with internal teams and external partners to deliver a set of connected workstreams.
Auditing the analytics stack. The audit covered both data collection and data activation, to find out where the stack was leaking accuracy or value before proposing any fix.
Defining the analytics roadmap. Keyrus defined the roadmap and led execution, coordinating internal teams and external partners so the roadmap actually got delivered rather than sitting in a deck.
Reviewing and optimising data collection. This meant challenging the existing data layer structure and leading the migration to server-side tracking on the Google stack, a move that improves both data accuracy and resilience as browser tracking limitations increase.
Defining and implementing a digital data product. The team reviewed and defined a new data model for online behavior data, then implemented the pipelines and transformations to gather it alongside media and web scraping data, combining BigQuery and Fivetran. The result was a data product ready to be consumed by the broader organization, not just the analytics team.
Defining a growth framework. Keyrus led continuous platform optimisation through new dashboard creation, journey analysis, and A/B testing using Optimizely, turning the rebuilt stack into an ongoing growth engine rather than a one-time fix.
Results: efficiency, spending optimisation and lower churn
Efficiency gains. Faster insights to react to market conditions, instead of waiting on manual reporting cycles.
Optimised marketing spending. A better view on channel mix and contribution, so budget follows what is actually driving acquisition.
Data-driven growth. Platform changes based on proven data rather than assumptions about what customers want.
Reduced churn. By leveraging behavioral signals in the customer portal to catch and address friction before customers leave.
What this project shows
A dashboard is only as good as the data underneath it. The growth framework, the A/B testing, the marketing spend decisions, none of it holds up if the analytics stack it runs on has gaps at the collection layer. That is why this project started with an audit and a data model, not a new dashboard.
This is the kind of project Keyrus builds: analytics stacks and digital data products that hold up under real competitive pressure, not a reporting layer bolted onto data nobody has stress-tested.
AI does not transform businesses. Architected intelligence does.
An analytics stack audit reviews both data collection and data activation, identifying where tracking is inaccurate, incomplete, or not reaching the teams who need it, before any new tooling or dashboard work begins.
Server-side tracking moves data collection from the browser to a server environment, improving accuracy and resilience as browser-based tracking becomes more restricted by privacy rules and ad blockers.
A digital data product is a structured, reusable data model, here built for online behavior data, combining sources like media and web data through tools such as BigQuery and Fivetran, and made ready for consumption across the broader organization rather than a single team.