Snowplow

I have been using Snowplow for the past four or five years. We migrated from Google Analytics to Snowplow, and our main use case for Snowplow is for tracking events for instrumentation. We also use it to understand user journeys and funnels, measure conversion, retention, and engagement, and it helps us to send clean, structured behavioral data into a data warehouse, such as BigQuery or Snowflake. The main purpose, to summarize, is for event-level behavior tracking where we capture how users interact with the product, and then we use that data to analyze funnels, engagement, and feature performance.

I can walk you through something we do on a day-to-day basis using Snowplow to track how many users are exposed to the feature, how many interact with it, where they drop off, whether they complete the intended action, and how downstream metrics such as conversion or engagement change.

Since we have visibility and control over our event tracking and the raw data we store, it helps us to quickly understand if there is a data quality issue rather than just thinking it could be due to seasonality or a promotional period. It simplifies our work when KPIs are trending down or up exponentially, making faster root cause analysis possible. We can quickly isolate the affected step, market, or platform rather than investigating the entire journey, resulting in measurable improvements in conversion, sometimes in the range of ten to twenty percent at specific funnel steps.

What makes Snowplow so unique and different is its flexibility. There are multiple use cases, and you can basically come up with your own customized events and properties, and track everything that you could think of. The challenge we had with Google Analytics and other tools where there were certain limitations does not apply, but Snowplow is quite an open-source tool for me where we can have our own structure, and we can make sure it works as per our taxonomy. We usually use it for event instrumentation and storing that data in our backend, where we have data models to transform that data. We have a silver and bronze layer where all the raw data sits, and we use DBT models on top of it to transform it, and we get our silver layers, with the silver layer being event-based row tracking. On top of that, we have built a gold layer where we get all our KPIs which powers our dashboards directly.

Snowplow offers very interesting best features because we see a huge difference in how we track data now and how much we trust it to make decisions. The strongest features would be that we get very granular data on what users actually do, define events and properties specific to our use cases, reconstruct sequences, and clearly identify drop-offs. Raw behavioral data goes into our warehouse rather than being locked in some other analytics platform, and it is also real-time, which we appreciate for monitoring our new launches and product changes. You can scale up or down your models quickly, as we used to have spikes in traffic for events such as the Met Gala or Amazon Prime Day, and Snowplow is very flexible for such use cases. It is very easy to validate the data; the schemas help ensure that events are consistently structured, so the biggest value for me is the granularity and flexibility of the event data.

I would pick granular event-level data because as a product analyst, we are often trying to understand exactly where users drop off or how they are interacting with a new feature. Having detailed events lets me reconstruct the user journey, build funnels, segment by device or market, and drill down when a KPI changes, making my analysis much faster. I can go beyond knowing that conversion dropped and actually understand where and why it happened.

I would also highlight schema validation, as having consistently structured and validated events gives me more confidence in the data and reduces the time I spend checking data quality before starting an analysis.

On a broader level, Snowplow has brought a lot of credibility to the analysis and dashboards that I have created, with the biggest positive impact being that it gives us reliable, centralized behavioral data that teams can use consistently. It has made it much easier and faster for us to understand user journeys and make product decisions based on actual user behavior rather than assumptions.

Snowplow could be improved in a couple of areas. The Snowplow team is readily available and proactive, always jumping on calls to make changes, especially how we track consent and non-consent data for the EU market. One area of improvement could be its vast canvas, which might feel overwhelming and confusing to people who are less technical or are not sure how to best structure their data. Although Snowplow is powerful, getting value from this granular event data does require strong SQL skills and knowledge of the underlying data model. Therefore, making the data more accessible to less technical users could enhance intuitive self-service exploration, funnel visualization, and easier debugging of tracking issues, allowing product teams to gain insights without relying heavily on analysts.

Snowplow could also automate most processes and implement smarter monitoring alerts for fallback. Better documentation and easier debugging when event tracking or schemas change would help, given that managing event definitions centrally can be a hassle when there is a breaking change. Having automated systems to inform users about changes in events or properties would make things smoother across teams.

I have been using Snowplow for the past four or five years.

Snowplow is stable.

Snowplow's scalability is pretty robust and fast. The Snowplow team handles a large volume of event-level data across multiple markets and user journeys, and I have not experienced significant scalability issues in our day-to-day analysis. It reliably accommodates increases in data volumes during major events when the traffic spikes by ten times, handling it smoothly without hiccups or data failures.

The customer support is diligent and proactive, always trying to come up with better solutions. They help us optimize our current instrumentation processes and understand how we want to use this data based on different organizational needs, providing amazing solutions or quick workarounds.

We previously used Google Analytics as our solution before switching to Snowplow, as we were unable to achieve the granularity in event tracking we desired and needed more flexibility in defining custom events while maintaining control over raw behavioral data.

Snowplow is a much cheaper tool compared to Google Analytics. While I do not have access to financial ROI, I have seen ROI in terms of time saved and faster decision-making. Now, when we see a conversion drop, the event data allows us to quickly identify the exact funnel step, platform, or user segment affected, which could otherwise take several days to narrow down. Now it takes just a few hours, allowing the product team to respond much faster and fix problems quickly.

I would advise those looking to use Snowplow to ensure that if it is a small organization, they can rely on tools such as Google Analytics or Apps Analytics for smaller use cases since they may not have a lot of data. I recommend being very clear about their tracking strategy before implementing it, properly defining key user journeys, events, and properties. Initially, it is better not to track everything—being focused and clear about needs is essential. Investing in a strong and consistent event taxonomy and schema from the beginning is crucial, as Snowplow offers much flexibility, but the quality of insights depends on the design of instrumentation. I would rate this product an eight out of ten.

My main use case for Snowplow is user behavior tracking at DPG Media on all their websites and apps as well as on television streaming. We stream the data to Snowflake in real time and stitch the data as well. From there, it is used for many use cases in the company.

For example, we tracked the scroll depth of pages and the time on pages, sending it to Snowplow. Many other use cases were developed. The data was stitched in a profile service on AWS and then utilized to segment for advertisements.

Snowplow offers the best features in that you are completely free to make your own data model, track the way you want to track, and control the way the data comes to Snowflake so you are the complete owner of the raw data without being forced into a certain data model.

Having that level of flexibility impacts our team's work and projects as we needed quite a lot of people that were really good in Snowplow. However, from the moment that you completely understood the technical aspects, you were completely free to set up your own Snowplow environment and track the way you wanted to track and what you wanted to track, which is not possible with Google Analytics and Adobe Analytics, for example.

Snowplow has impacted my organization positively, very well. For DPG Media, it is the most important data source that we have, delivering a lot of value in the company. We are making data mesh products on this data all over the company. I think for advertisement, it is really the most important source of data.

I do not have many improvements to deliver for Snowplow. Perhaps for smaller companies that do not need to customize so many things would be helpful. In the beginning, we also did some things wrong. The data model that we used was not as clean everywhere we were tracking. We made some mistakes, but I think we learned from them, and perhaps for smaller companies that start now, it could be made a little bit easier, although I think Snowplow has already worked on that.

I have been working in my current field for over seven years.

Snowplow is stable.

Snowplow is very scalable.

Snowplow is deployed in my organization on a private cloud managed by Snowplow on AWS.

My experience with pricing, setup cost, and licensing was very good. The only issue that we had at a certain moment was that Snowplow was offering more services and asking us to pay more, but we did not use all these services. We had a discussion with them, but they were very open to discussing with us and negotiating a new contract.

Snowplow's customer support is also very good. We had an account manager available, so the experience was very good.

I previously used Google Analytics in the company, but now at the end we only use it for very small B2B websites.

We did quite a lot ourselves. Snowplow's whole setup was done by ourselves. I know that Snowplow now offers a stitching service, but at DPG Media, we did that ourselves. We did the streaming ourselves and the whole data model, etc. Snowplow's whole setup was custom-made for DPG Media.

We did quite a lot ourselves. Snowplow's whole setup was done by ourselves at DPG Media.

I have seen a return on investment, but that is difficult to quantify. I would not say fewer employees, but we saved money because Google Analytics was much more expensive. The time to deliver real-time data and value helped a lot.

My experience with pricing, setup cost, and licensing was very good. The only issue that we had at a certain moment was that Snowplow was offering more services and asking us to pay more, but we did not use all these services. We had a discussion with them, but they were very open to discussing with us and negotiating a new contract.

Before choosing Snowplow, I evaluated other options, including Google Analytics, which was another option, but it was very expensive, along with Google BigQuery.

My advice to others looking into using Snowplow is to learn first from others that are also using it. I would rate this review nine out of ten.