How Visionwork enabled product managers to make data-driven decisions 4x faster.

The data access bottleneck

Visionwork's product team had embraced data-driven decision making in theory, but logistics made it nearly impossible in practice. With 45 product managers and only three data analysts, the math did not work. Each analytics request took an analyst 2-4 hours to complete—write SQL queries, pull the data, create visualizations, and explain the findings. With new requests coming in faster than analysts could complete them, the backlog grew to 60+ tickets.

Product managers adapted by avoiding data requests. They would launch features based on user interviews and intuition rather than wait two weeks for metrics. They would run A/B tests but make decisions before the analysis was complete because they could not hold up roadmaps. The VP of Product knew this was leading to poor prioritization, but hiring five more analysts was not realistic and would not really solve the underlying access problem.

Self-service product analytics

The solution emerged from a conversation with a product manager who had gotten frustrated and tried to build her own dashboard using Lumis. She had connected it to their event tracking data and created exactly the metrics she needed to evaluate a feature launch—completion rates, time to value, user segments, and retention impact. The whole thing took her twenty minutes. When she showed the VP of Product, his immediate reaction was why is not everyone doing this.

The rollout was straightforward. The data team built templates for the most common product metrics—feature adoption, user engagement, funnel conversion, cohort retention, and experiment results. They connected Lumis to their event tracking system and user database, then gave every product manager access. Training took one hour, focusing on which template to use for different questions rather than how to write SQL or build charts.

From blocked to empowered

The impact was immediate. Product managers who had been waiting weeks for simple dashboards were creating them in ten minutes. The analytics request backlog dropped 95% within a month as PMs handled standard metrics themselves and only engaged the data team for complex statistical questions. Feature development accelerated noticeably—teams were running 180% more experiments because they could analyze results immediately instead of waiting in the queue.

Product velocity increased 47% in the quarter following implementation. The VP of Product attributed this directly to faster decision-making enabled by self-service analytics. Teams were shipping features with confidence, learning from user behavior in real-time, and iterating based on data instead of assumptions. The data team, freed from dashboard factory work, focused on building better data infrastructure and conducting complex analyses that actually required their statistical expertise.