What’s the best path: centralized data warehouse or separate data marts?

[DanielIT]

I'm so tired of bringing the same boring lunch to work every day. What are your favorite simple lunch ideas that are actually exciting to eat?

I need things that pack well, don't require reheating (or can be eaten cold), and are easy to make ahead. I'm especially interested in budget-friendly meals since I'm trying to save money by not eating out. What are your go-to simple lunch ideas that make you look forward to lunch break?

[GregoryL]

I’m trying to decide if we should use a single, centralized data warehouse or keep our departmental data marts separate as we scale. The main issue is that our marketing and sales teams are now asking for combined reports, and our current setup makes that a real pain without a lot of manual stitching.

[Samuel_A]

We moved to a centralized data warehouse last year after the same pain points; the cross team reports became much easier, and executive dashboards load faster. It took us about six months to stand up the core model, with a small data governance charter and a nightly ETL. The first few weeks were painful, but by month three usage grew and we retired manual stitching.

[NatalieL]

We tried keeping marts separate and adding a semantic layer to stitch in BI, but we still spent a lot of time negotiating definitions. Governance felt heavier than the tech. We eventually started a quarterly data glossary and a single source-of-truth table for key metrics.

[Olivia.P]

Maybe the real issue isn't where the data lives but whether marketing and sales actually agree on KPIs and time horizons. If everyone wants a single report but defines 'open rate' or 'close rate' differently, you're chasing ghosts.

[StephenP]

We spun up a pilot central store and migrated two domains first. ETL runs nightly, then moved to near real-time for a few dashboards. It was messy; some teams bailed early on because data ownership changed, but those who stuck saw fewer manual joins.

[BrandonLJ]

I tried to push this but I'm not sure we solved the problem; we ended up with a two-tier model where some teams still pull from marts for agility but others rely on the central store. The report stitching got easier but not perfect; cost also crept up.

[Matthew15]

Do you have a clear plan for data governance and lineage, or are you rushing to combine things before that is in place?