From fragmentation to a focused modernization program
Exec summary:Â Â
We worked on a simple change for one European marketplace team: focus + ownership.Â
Truly, cut the noise and set up three work areas with named leads:Â
- Tooling & Training (clean up SQLs, adopt dbt, improve data discovery with AI).Â
- Data Product Layer (pick a few products, define standards, pair analysts with engineers to add meaning for data).Â
- Metric Layer (create shared definitions, certification and simple governance for the most used metrics).Â
Weekly rhythm. Small and short wins only. 1:1 coaching for each lead. Coalition between data team and business analysts. No “big bang”— just visible progress the leaders can see.Â
In eight weeks they had a plan leaders could support, pilots in motion, and less rework.
Â
Lesson: reducing load and creating ownership beats adding tools every time.Â
Audience takeaway:Â Turn around from messy, reactive analytics into an execution rhythm that leaders can trust, celebrate progress and plan clean steps forward.Â
 Context: Â
Company’s analytics landscape had grown fast but uneven. Teams were doing heroic work, yet standards, ownership, and metric consistency lagged. Constantly bombarded by new requests, escalations and never ending fire fighting to keep the lights on, day over day.
 Challenges (before we started):
- Tooling gaps and hard‑to‑maintain SQLs; slow queries and low data discoverability.
- Inconsistent documentation and practices; analysts carrying end‑to‑end work without support from engineers.
- Fragmented ownership and overlapping datasets; conflicting definitions for key metrics; no shared understanding what data shall be prioritized on the platform.
- A culture of “acceptable dysfunction” that increased technical debt daily.
 What we did (advisory, training & coaching)
- Facilitated a series of internal workshops to establish shared language on data products, roles, and standards.Â
- Ran an advisory sprint to frame strategy into three themes: fix delivery friction, define & reuse key metrics, and stage the platform roadmap.Â
- Co‑created three workstreams with named leads and first actions:Â
- Tooling & Training — dbt Core adoption, metric‑layer approach, and discovery tooling.Â
- Data Product Layer — identify flagship products, pair analysts and engineers, define pragmatic standards and easy certification.Â
- Metric Layer — reuse‑ready strategic metrics, scoping most pragmatic ones.Â
- Set up coaching with stream leads and agreed a four‑week follow‑up to lock scope and top actions. Monthly check-in with data leader to monitor overall progress.
 Outcomes so far:Â
- Clear owner‑led workstreams and a multi‑week action plan to show visible change.Â
- Regular cadence established inside the core team; first PoCs on dbt Core and semantic layer. Higher efficiency within the core team, with better resource utilization.
- Agreement to avoid tool PoCs without clear stakeholder buy‑in, supported by top leadership.Â
 What made the difference:Â
-
We reframed success from “tools first” to visible business‑facing improvements driven in coalition between data & business, with clear workstreams and bite-sized deliverables.
