• Industry Media and advertising
• Engagement Embedded BI developers, ongoing
• Scale 40+ bespoke dashboards, the source of truth for client performance reporting
• Stack Power BI, BigQuery, Dataform, Funnel.io, Power Apps, Power Automate, n8n

Campaign reporting that was being rebuilt by hand, every month

A media agency's activation teams report performance back to their clients on a cycle. The data for that reporting sits across every channel a campaign runs on: paid social and search platforms such as TikTok, YouTube, Meta, Google Ads and Apple Search, alongside AV covering linear TV, BVOD and audio.

Those two worlds do not report the same way. Digital platforms each export in their own format and labelling convention, while AV arrives on a different cadence and a different measurement basis entirely. On top of that, each client measures success differently, with its own KPI definitions and commercial logic behind the numbers. So assembling a client report was substantially manual work, repeated every cycle, done by marketing managers whose job is campaign performance rather than data preparation.

That hidden tax was measurable: reporting historically consumed around 20% of the agency's total operational capacity. That time went into sourcing, segmenting and formatting data rather than acting on it.

Alongside that, the internal teams needed something different again: a view of pacing, actual spend and delivery against plan, while a campaign is still running and can still be corrected.

The agency needed developer capacity that understood both Power BI and the media data model well enough to build this without a long ramp-up.

What we built, and in what order

DA Lytics works as part of the Havas Media UK BI team, on the team's backlog and in the team's tools. The work splits across four areas, all part of our Power BI and data engineering services.

The data foundation

Campaign data lands through Funnel.io, where we built the initial platform connections, and into BigQuery. Transformation logic lives in Dataform, which gives the warehouse version control, dependency graphs between models, and a place for business logic to live that is not inside a report.

That matters more than it sounds. Once a client's KPI definition sits in a Dataform model rather than in a DAX measure inside one report, it can be reused, tested and corrected in one place.

We maintain the BigQuery datasets as an ongoing part of the engagement, working with both structured platform data and unstructured sources.

Reporting built per client

Each client gets its own Power BI report, carrying that client's KPI definitions and commercial logic rather than a generic template with a logo swapped in. More than 40 bespoke dashboards are live today, each serving as the definitive source of truth for that client's performance.

The cross-channel view brings digital and AV onto one reporting surface, so a campaign running across both is read in a single place rather than reconciled from separate exports on separate cadences.

Some of this work covers UK market clients. Some covers global clients, which within Havas would normally route to the global BI team.

Internal pacing and company-wide solutions

Separately from client-facing reporting, we build the internal tooling: pacing reports comparing actual against planned delivery, so campaigns get corrected mid-flight rather than explained after the fact.

Some solutions are built for a single client. Others are company-wide, serving every team in the agency.

Automation and platform administration

Reporting alone does not remove manual work if the process around it is still manual. So the engagement also covers:

• Power Apps for capturing data that has no system of record, written back into the reporting layer
• Power Automate for approvals, alerts and scheduled distribution
• n8n workflows for automation that sits between tools where standard connectors do not reach
• Power BI workspace administration: workspace and app structure, access control, deployment pipelines and refresh scheduling

What changed

• Five full-time employees' worth of capacity reallocated. Over the last two years, the productivity equivalent of five FTEs has moved from manual data preparation to high-impact planning and activation.
• 40+ bespoke dashboards live. Each serves as the definitive source of truth for its client's performance, carrying that client's own KPI definitions rather than a shared template.
• Digital and AV in one view. Paid social and search alongside linear TV, BVOD and audio, consolidated into a single cross-channel report.
• Global clients served outside the global BI team. Work that would ordinarily route to Havas's global BI function is handled by the UK team we develop for.

Reporting once consumed around 20% of the agency's operational capacity. Automating the low-value, high-effort work of sourcing, segmenting and formatting data returned that capacity to the work clients actually pay for. The reporting is no longer the bottleneck in the client reporting cycle: activation teams pull the numbers rather than assembling them, and the definitions behind those numbers are consistent between one cycle and the next because they live in the warehouse rather than in a person's spreadsheet.

As the agency's business operations lead puts it: "We haven't just saved time; we've increased our ability to deliver radical transparency to our clients."

What the client says

"Throughout its engagement with Havas, DA Lytics has demonstrated a proven track record of delivering challenging analytics solutions on time, even within tight deadlines and evolving requirements. Their ability to translate complex data landscapes into clear, actionable insights has been instrumental in supporting data-driven decision-making across multiple stakeholders. Beyond technical excellence, DA Lytics stands out for its strong soft skills and collaborative approach. The team maintains proactive communication at all stages of delivery, ensuring transparency, early identification of risks, and alignment with business expectations. Their responsiveness, stakeholder management, and ability to adapt quickly have made them a trusted partner."

Benjamin Loosli, BI Manager at Havas Media UK

Where the engagement is heading

The next phase is self-service analytics: a single governed source that teams across the agency can query themselves, including through AI tools, without each department building its own extract.

That work is the same discipline as everything above. An AI assistant returns a reliable answer only when the data underneath it is clean, consistently structured, and described well enough to be interpreted without guessing. The Dataform models and semantic layer built over the last phases are what make that possible.

If your reporting cycle still runs on exports

Multi-platform campaign data, a different KPI definition for every client, and a reporting cycle that eats days of someone's month. If that describes your agency, the shape of the work is usually familiar even when the stack is different. We have written before about what Power BI does for an agency. Tell us where the manual step is and we will tell you what it takes to remove it. Book a call or see our services.