To make a complex product simple.
Customer case Skistar
SkiStar AB is the leading player in mountain tourism in Scandinavia. Until a few years ago, it was a company focused on winter tourism, but has expanded its offering to be "all-year". The vision is to create memorable mountain experiences, and the experience starts already in the purchasing process. Cypoint has been working with SkiStar for more than a decade, where we have had the privilege to be part of an organization that, despite its size and market value, is still characterized by an entrepreneurial spirit.
Challenge
To measure is to know - but are we using the same yardstick?
SkiStar's product for the customer is very complex, the idea is to be a "one-stop-shop" for accommodation, ski passes, ski rental, ski school, events, and related products. Hiding the complex backend from the customer is key to providing the best possible shopping experience.
A complex and wide range of products also presents challenges in analysis and business intelligence. In a company with data-driven decisions, new requirements and data sources are constantly emerging.
SkiStar's biggest challenges were that the needs for analysis evolved faster than the existing toolset could accommodate. There was a great need for centralized data, clear strategies for data processing, moving away from risky person dependencies, establishing a development process that is efficient and enjoyable to work with as a BI developer. Together with SkiStar, we at Cypoint have been involved in developing the technical foundation that solves these problems today, while also looking ahead to how the landscape may change in the medium to short term.
Solution
Centralized data, clear development process
Together with SkiStar, we developed frameworks around a shared data warehouse. We established rules on how each system hands over data, as well as strategies on how we extract data from external systems. The technical platform, consisting of a number of frameworks, patterns, and tools, formed the basis for a predictable way of working and democratized maintenance and troubleshooting.
All transformational data flows are versioned like other code in the organization, CI/CD pipelines enable automatic deployments of new features.
Orchestration of data flows is done in tools that allow more people to troubleshoot and restart failed jobs.
Result
Streamlined data flow
Thanks to the technical platform, today we have a centralized data warehouse where source data is verified. Transformations and analysis take place in several steps, and data consumers can choose to use data from the different stages based on the acceptable latency and the level of data aggregation. In PowerBI, there is a set of curated datasets that are verified, where the most advanced users can build their own reports following the self-service model.
Documentation takes place in the same location where the transformation occurs and becomes a natural artifact in the form of a data catalog. Test environments were introduced which, together with version control, provide a simple and smooth experience for developers where they can reach their full potential. Instead of relying on individual developers, SkiStar's back office support can now also monitor and resolve issues related to BI flows.
More customer cases
Polarn O. Pyret customer case. Successful development of e-commerce and PIM through close collaboration.
JULA customer case. How we helped Jula with the implementation of a new modern PIM.
Tykoflex customer case. Successful launch of a new strategic website on a flexible platform.
Regin customer case. New PIM, seamlessly integrated with web and business systems.