Learn how Avenso uses the Data Virtuality Logical Data Warehouse to build a central data warehouse and connect it to the business’s data marts to reduce reporting time to less than 5 minutes.
'With the help of the Data Virtuality Logical Data Warehouse we brought multiple sources together and we managed to create data marts which contain data from a lot of sources, ready to be used. And once we connected the LDW to our BI tools for analysis we already reduced a lot of dependency on us. And now, instead of downloading data from different sources, and then joining and cleaning them e.g. in Excel, we simply have these tables ready-made in Data Virtuality.'
Head of Business Intelligence
ABOUT AVENSO AG
Avenso GmbH operates an art gallery network and online retail for photographic art. Founded in 1999 and based in Berlin, Germany, the company has set out to democratize the art market. Avenso unites two brands in its business – LUMAS and WhiteWhall, both located in Berlin. LUMAS is a retailer for museum-grade, curated photo artworks, as well as contemporary art. And, since 2004, the brand creates a unique experience for a wide audience with 20 LUMAS galleries around the world. Under the name WhiteWall, Avenso operates a photographic printing brand with museum grade quality, serving artists, professional photographers, and hobbyists online and in WhiteWall Stores in Berlin, Munich and Dusseldorf.
As an ecommerce business that offers custom-made products in gallery quality as well as a range of online services for high-end photo processing, Avenso depends on current and clean data that is available at all times. Indeed, every department, or roughly 80% of the company, works with data every day. In the past, this presented the BI team around Raghav Sehgal, the Head of Business Intelligence, with substantial challenges.
A Lot of Data in Disparate Data Sources
With two brands operating in different countries around the globe, Avenso generates a lot of data in vertically spread sources: Spryker, Google Analytics and other data from each of the 23 web shop’s MySQL database, marketing data, offline data from the 14 galleries, SMB data, and data from the companies’ producers, who also work in different countries.
Compiling data that is spread across disparate data sources was a time-consuming work for Raghav and his team as they had to manually gather, cleanse and then consolidate the data in Excel spreadsheets. It took a heavy toll on the company’s speed of execution in terms of exhausting the full potential of the data. Not being able to access the different data sets speedily, Avenso’s BI team could only analyze selected data sets at a time.
Restrictive Access to Data and Dependency Bottlenecks
Another challenge for Avenson’s BI Team was the management and supply of data to all other departments in the company. As Raghav recounts, “The original process was very restrictive because everyone always had to go through the BI department as we were providing the data. Also, access to different data sets was restricted. We only gave access based on the department the request was coming from and the role the person requesting access has in that department.”
This restriction on access and knowledge also led to a significant bottleneck of dependencies: If Raghav and his team were not available, the rest of the company had to wait for the BI team to return to work on the system before they could get the data they had requested.
Looking for an Out-of-the-Box Data Integration Solution
Faced with these challenges, Avenso’s BI team set out to find a data integration solution that would enable them to become truly data-driven.
Avenso decided against building a traditional data warehouse as this would have been a time and resources intensive endeavor. Also, as Raghav points out, “maintaining a traditional relational database system is very costly from an operational point of view. Because every time something changes you have to adapt the system to these changes. So, you need to keep an eye on the data warehouse at all times.”
Instead, the company began looking for an out-of-the-box solution with which the BI team could start pulling in the data right away. Further to this, Raghav and his colleagues also wanted the perfect-fit solution to
- Offer many different connectors, especially to standard databases such as MySQL, PostgreSQL, Google Analytics and SAP,
- Centralize the data and build a single source of truth platform for reporting and other usages,
- Validate, cleanse, and standardize the data for analysis
- Come with an inbuilt job scheduling feature to set up automatic processes.
The Data Virtuality Logical Data Warehouse was the first solution Avenso put to trial – and it turned out to be just the solution Raghav and his team were looking for.
Easy, Straightforward Implementation
It took Raghav and his BI team 6 weeks to set up the first basic reporting process with Data Virtuality Logical Data Warehouse including the traditional business KPIs. With a little bit more time, the initial structure and all connectors to their webshops, marketing tools, CRM system, etc. could be brought in. As Raghav reports, “Normally, setting up the initial structure is challenging and time-consuming. But with Data Virtuality it was a very clean and pretty straightforward process.”
And with this foundation, they could start digging right into the data.
Daily Business: Central Data Warehouse Development
On a day to day business, Avenso’s BI team uses the Data Virtuality LDW for reporting and data validation as well as for central data warehouse development, specifically for building a business data mart from the sources that are connected to the Data Virtuality platform.
The BI team also has to do daily standard tests to check that all jobs are working properly and that all data is going into the correct business tables.
And, of course, they get new requests from other departments at Avenso all the time for which they have to develop new procedures, if it can’t be covered with other existing ones.
A Single Source of Truth That Everyone Can Draw On
With the Data Virtuality Logical Data Warehouse, Avenso was able to significantly reduce the time it takes to get the data and use it in their BI tools, such as Tableau, for analysis and reportings.
But having a single source of truth not only improved the speed of access to data at Avenso. Raghav’s team is also able to thoroughly and efficiently validate, cleanse and standardize the generated data, so that it can be used right away.
“With the help of the Data Virtuality Logical Data Warehouse we brought multiple sources together and we managed to create data marts which contain data from a lot of sources, ready to be used. And once we connected the LDW to our BI tools for analysis we already reduced a lot of dependency on us. And now, instead of downloading data from different sources, and then joining and cleaning them e.g. in Excel, we simply have these tables ready-made in Data Virtuality.”
Time savings thanks to automated processes
And while it’s only the BI team who have access to the LDW, most of the business reports have been automated. Meaning the whole process is now much more efficient, less time-consuming, and doesn’t require manual work.
This way, Avenso was able to reduce the time it takes to get a data set by more than 90% – from two hours to less than 5 minutes. According to Raghav: “In the past, if someone asked for a raw data set with annual sales of all countries for a brand and then wanted to join that data with data from Google Analytics, it would have taken two. Because you’d have to manually get the data. But now, if you’re downloading it directly from Data Virtuality, you get the data in less than 5 minutes. Because it’s standardized and the data is already prepared and ready. No matter the skill set. DV is pretty straightforward and easy to use. So, even a junior analyst would need not more than 5 minutes – just like an expert.”
Raghav's next project is to enable cost attribution analysis for multi channels so Avenso can fully understand the cost associated at each stage along the customer journey - from the beginning to the end. This will help Avenso to gain competitive advantage by focusing their resources on the most profitable channels.