Lightdash

The seamless integration with dbt is remarkable. Previously, our BI was limited to Google Sheets, but Lightdash enabled visualizing over 150 models within six months. Being developer-first with tight dbt integration made setup easy with a small team. Customer support is top-notch and responsive. There have been growing pains as Lightdash transitions from alpha to production-ready, but they're quickly catching up. The mobile solution could be stronger. They effectively address both internal and external customer issues.

Lightdash closely mimics Looker in user experience, making adoption easy due to familiarity. I particularly like defining metrics within the dbt project, simplifying the start. Combined with CLI tools, it's developer-friendly. The open-source version exists, but even cloud usage is far cheaper than Looker. As a newer tool, some visualizations lag behind Looker (e.g., freezing columns), but the team ships quickly and is customer-centric, so they'll likely catch up soon. It suits data visualization and self-serve analytics for non-technical users.

Lightdash is our primary BI tool, easy to start with and equipped with all essential features. Being part of the semantic layer community, it goes beyond pretty charts, allowing non-technical users to explore data with predefined logic from the analytics team. In the past, granting business users access often led to incorrect conclusions due to unmanaged calculations, but Lightdash solves this by enabling pre-defined metrics and joins. Its full integration with dbt bridges the gap between database logic and BI, making it easier to understand the impact of upstream changes. The ability for users to explore, filter, and group within the guardrails set by the analytics team is excellent. The team is responsive to bugs and feature requests. While Lightdash is still young and missing some features, none are dealbreakers, but we often seek workarounds. These are on the roadmap, and I'm eagerly awaiting their release. Databases are tables of rows and columns; when connected to BI, you need metrics and dashboards. Previously, creating well-curated KPI definitions was nearly impossible, leading to limited or incorrect user access. Lightdash solves this by letting analysts define metrics, joins, and dimensions that business users leverage to create their charts.

The rapid iteration and team's helpfulness in prioritizing useful features stand out. Deploying via CI/CD is a huge time-saver compared to previous solutions, boosting engaged BI users with fast content deployment. The CLI and API offer a great developer experience, enabling integration into various analytics workflows and central version control. The low learning curve for business users offloaded significant work from the data team. There's little to dislike; future features could include better search and discoverability to guide users to relevant content. Charts as code, removing the need to manually create charts for new metrics, would be excellent. The tool effectively delivers metrics to the business, ensuring easy access and stability through a metrics layer.

Years ago, we chose ThoughtSpot over Looker, but I wish Lightdash had been as mature then. We've since switched to Lightdash a few months ago. Key advantages: it's open-source or affordably priced in the cloud, with transparent pricing based on usage or licenses. If Lightdash continues its pace, it may eventually be known as 'Looker is a closed-source Lightdash.' The dbt project serves as the foundation, keeping all data logic in one repo with easy CI/CD. Its modeling functionality is advanced but not yet as mature as LookML, though improving. It's intuitive for non-technical users with minimal onboarding, enabling self-service. BI maintains models and top-level reports, while stakeholders and data champions handle the rest. A bonus: no limits on projects or user sharing, which we use extensively to segregate reporting areas and share with clients—a premium feature in traditional BI platforms. This allowed us to retire DataStudio reports. The team is transparent with requirements, supportive, and responsive. All communication is Slack-driven with a great community, no ticket systems. Some advanced modeling challenges remain, but they're aware and improving. Missing multi-tab reports, and while I'm used to quick responses, even their few-hour replies to my many questions are fine. Lightdash handles self-service well, letting BI focus on strategic work rather than ad-hoc tasks.

Lightdash has been a revolutionary tool, eliminating shadow semantic layers and bridging the gap between business-driven insights and those from the data team. The clear separation of dashboards and charts, all validated through dbt, ensures every dashboard element answers a business question. Stakeholders can mix and match governed, semantically correct charts for their own use. Implementation was straightforward, and we were operational in under two days. We've replaced Tableau, PowerBI, and Looker with Lightdash, and business stakeholder usage has increased. I struggled to find any flaws; there's minimal missing functionality. Potential improvements could include broader visualization types like geo-spatial and clustering maps, which are common outside pure BI. The tool excels at answering business questions, governing thematic data access, enabling better data storytelling, and offering novel alerting and dashboard testing methods.

The tool is highly intuitive after some initial training, enabling all team members to independently build their own dashboards. The initial setup did take some time for the team, but it's empowering us to make data-informed marketing decisions.

Lightdash is incredibly user-friendly, especially for quickly viewing reports. It saves time by centralizing all data and facilitating easy report creation. The automated report generation for metrics is particularly valuable. I have no complaints and find it extremely helpful for consolidating data and producing straightforward reports, which is a time-saver. The ease of viewing reports and the automation of metrics tracking are features I highly value.

The precision in data extraction is excellent, allowing you to focus on exactly what you need to analyze. I also appreciate converting metrics into graphs and sharing links with colleagues. Overall, it's a solid reporting tool, though it can be clunky and confusing due to field naming conventions—for example, 'region' is labeled 'store', which isn't intuitive. Refreshing data manually instead of automatically when new information is pulled can be tedious, and copying tables isn't as straightforward as in Excel. I rely on Lightdash for extracting data and metrics, which serves well for reporting and analysis, helping to pinpoint issues by narrowing down necessary details. The graph translation and team sharing features make it a valuable reporting asset.

Lightdash leverages dbt and merges the strengths of various tools: it generates visualizations about ten times faster than Tableau, data modeling is considerably simpler than in Tableau or Looker, and models are automatically validated, breaking when they don't align with dbt, unlike in Tableau or Looker where these are separate systems. The team ships new functionality at an incredible pace, often based on user requests, and dashboards can be created through prompts in Cursor. So far, I haven't encountered any downsides: it does require a relatively high technical skill level and isn't suitable for a 'plug-and-play' BI analyst right out of the box. If cost is a concern, self-hosting is an option. I initially preferred Tableau over PowerBI, then recognized Tableau's compute limitations and lack of git integration compared to Looker. However, Looker is pricey and not dbt-integrated, meaning you must manage two separate repositories and duplicate changes. Lightdash is clearly the top BI tool available today.