Beam Analytics is a GDPR compliant Google Analytics (GA) alternative. There are many GDPR compliant GA alternatives out there. So on top of the standard web analytics you get from GA like visitors, bounce rate, traffic source, site duration etc, we also wanted to give users funnel analysis. And we wanted to be way more affordable than the alternatives. Our thinking was that it shouldn’t be expensive to stay compliant with the law.
Build vs Buy
We knew that we wanted to use ClickHouse because it’s ridiculously fast and has loads of out of the box features for working with web data. As always, the big question was whether to build or buy. There were 3 key reasons why we ultimately chose to go with Tinybird, which has all of the benefits of a managed ClickHouse with so much more:
Speed to market
We estimated what it’d take to build everything ourselves and the answer wasn’t great. Without using Tinybird, it would have taken us at least a full year to build a high quality service that we could take to market. With Tinybird, we were able to cut that right down to 3 months as we could just focus on building what mattered. In a world where most ideas fail to even reach the market, Tinybird played a huge part in keeping the momentum behind Beam Analytics going.
No operational overhead
We want to build an amazing service, and we want to offer it at an amazing price. We’re a small team, so we need to be able to focus our time and effort on continuously improving the product, and reduce any wasted effort just keeping the lights on. By using Tinybird, we don’t have to worry at all about backups, security, scaling, upgrades, monitoring or any of the things that don’t actually provide value to our users, but are absolutely necessary to deliver a great service. Instead, all of our time is spent building great analytics for our users!
Predictable and efficient pricing
We’re trying to build a profitable and sustainable business that our users can rely on for the long term. To be able to do this, we need to have clarity over our costs, and be able to predict how that cost will change as we grow. With Tinybird’s pricing structure, we have far greater certainty and visibility about how our cost structure will evolve over time. And knowing that meant we could be even more aggressive with our pricing and still make sure we were building a solid business.
How we built Beam Analytics using Tinybird
We primarily interact with Tinybird using these 3 features:
Tinybird CLI
We keep all of our Tinybird resources in a source control repository so that we can collaborate and version control our work. We use the Tinybird CLI to programmatically push our work from source control into Tinybird. This gives us total control to define and evolve the underlying ClickHouse schemas. For example, we specify each field type as well as SORTING and PARTITION keys. This control enables us to create highly optimized tables that reduce query latency and improve efficiency.
Tinybird Events API
We use the Tinybird Events API to send anonymized, one way hashed analytics data from our cloud running in the EU Hertzner datacenter to Tinybird. This is an amazing abstraction that enables us to focus on the product and not extraneous tasks like spinning up Kafka and thinking about batch sizes. Each pageview is sent to Tinybird with a simple HTTP request.
Tinybird Query API
Lastly, we use the Tinybird Query API to power our dashboards. They take care of SQL sanitization, authentication and query planning. All we do is construct the queries we want to execute and they respond with a nicely formatted JSON response that we use in our frontend dashboards.
Conclusion
We are really ecstatic to build on top of the amazing ClickHouse infrastructure Tinybird has built. They’ve served us well from the prototyping stage, to the MVP stage to now the scaling phase. If you’re at all interested in privacy preserving web analytics and/or have questions about our Tinybird setup, please reach out directly at hi@beamanalytics.io
Thank you!