Dance reduces customer acquisition costs by 20% with Fivetran

German-headquartered startup Dance is on a mission to create more connected and livable cities through sustainable mobility. The company offers electric bikes and mopeds on subscription basis and operates in Berlin, Hamburg, Munich, Vienna and Paris. 

In the beginning, Dance had just a database which was connected to a replica database from where some basic KPIs could be extracted with a simple reporting tool. Dance’s data analytic capabilities were tightly constrained due to a lack of access to backend data and because of the poor performance of the replica database. 

Accelerated data mobility with Fivetran

The company looked to build a data stack that would enable self-service analytics across the company. The demand for information was high at the company, not only for reporting but also to better understand the business and the markets. Specifically, Dance’s marketing team wanted to drive more ROI with its budget. Its product teams desired to optimize pricing, staffing and vehicle availability as well as prevent revenue loss. 

Bruno Pinto de Sá, a senior data engineer at Dance, needed a solution that would enable him to easily centralize data into a data warehouse for analytics.

“We tried a variety of tools for automated data movement. Fivetran performed perfectly on Postgres RDS database, Stripe and all the other sources that are important for us. It was obvious that Fivetran could handle maintenance in the most efficient way – an important factor for us since we are a small team.”
— Bruno Pinto de Sá, Data Engineer, Dance

In just a few days, data sources were integrated. “This would have taken us three to six months without Fivetran,” Pinto stated. Since then, he’s added TikTok Ads, LinkedIn Ads, Braze and Google Analytics for its marketing team.  

Dance also uses Papershift to plan the shifts of the mechanics who repair the ebikes. There was no pre-built connector for that in Fivetran, so Pinto had to create one. But even in this case, he was able to rely on Fivetran. “The support of the company on setting up this new connector was really impressive. The process was fast and easy.” By integrating the tool, Dance was able to track how many repairs could be done during a shift, allocate the repair costs to the vehicles and thus better calculate them.

For data transformation, Dance migrated to dbt Core. This opened the possibility of scaling the number of transformations and data sets that are customized. This enabled Dance to quickly and easily analyze HubSpot marketing performance and provided detailed insights into their financial performance, including subscriptions, payments and coupons. By utilizing Fivetran’s fully integrated scheduling for orchestration, data is transformed automatically upon load of new data from the connector in the warehouse. This reduces data latency and Dance‘s reliance on third-party orchestration tools.

Business and operational data like vehicle performance, weather and battery tracking are ingested and transformed from the Dance database and Google sheets and modeled by custom dbt models. Dance uses Looker as its BI tool for analysis. 

Outcome

Today, about 80 percent of Dance employees use Fivetran data. All departments get their respective core metrics with just a few clicks. If someone wants to drill down into a specific issue or metric, even colleagues without any technical knowledge can easily adjust the report themselves to get all the insights they need.

Key business outcomes include:

• Increased ads performance: With new insights, the marketing team was able to increase conversion rates and reduce the cost per customer acquisition (CAC) by 20 percent within four months

• Optimize pricing, staffing and vehicle availability: All business decisions are data driven, based on the combination of various data points, such as demographics for each city district, customer data, seasonality and specific events as well as amount, duration, and maintenance work costs for the vehicles and much more. 

[CTA_MODULE]