How dentsu Norway helped their clients build better data models with Supermetrics and Google BigQuery

dentsu is a global group within media, CXM and creative, on a mission to drive sustainable business growth for brands and companies. dentsu’s network consists of several marketing agencies across the globe, such as Carat, dentsu X, iProspect, Merkle, and Isobar, just to name a few.

The Norway team is seen as one of the leaders in terms of marketing reporting across the whole dentsu network. Jarle Alvheim is Head of Data Technology at dentsu Norway and responsible for cross-platform analytics and reporting. A part of his day-to-day job is structuring and building data models that help dentsu’s clients uncover meaningful insights.

Looking to standardize client reporting

dentsu consists of multiple subsidiaries working on large-scale client marketing projects. As a result, the marketing reporting dashboards are challenging to build, and the amount of data they worked with was significant. As the business grew, dentsu Norway struggled to continue building high-quality reports for the agencies’ clients.

Jarle says, “We had quite many issues with client reporting. For example, the whole process took a lot of time, and combining cross-channel reports was difficult. In many cases, the reports looked different depending on who built them.”

The team understood that to improve both the speed and the quality of their reporting, they needed to streamline the process.

Jarle says, “We wanted to find a solution that could help us standardize our process and do reporting more effectively. Additionally, building dashboards on top of that to see the results of our work rather than spend time preparing the data.”

To solve the client reporting challenges, Jarle and his team started looking for a solution. Basically, when moving cross-channel data into a data warehouse or cloud storage, you have two options. You can either build and maintain your own data pipelines or buy them from a vendor.Ideally, the solution would meet these criteria:

• Stability.
• Connect to all the marketing sources that the dentsu team and their clients need.
• Integrate well with Google Cloud Platform and Google ecosystem.
• Can be combined with standardized scripting like Python.
• Up to date on all changes in the different marketing platforms.

According to Jarle, home-grown data pipelines were off the table from the get-go. After carefully weighing up different vendors, they chose Supermetrics. At first, they tested connectors from other vendors. However, they soon realized limitations regarding the metrics and dimensions one could choose. Additionally, since they processed the data under the hood, it was challenging for the dentsu team to understand what was going on with their data.

Jarle says, “The last thing we wanted to do is build the data pipelines ourselves. That could easily take 50-100 hours per connector, and then you have to take care of the maintenance yourself as well. We ended up looking for a vendor and decided Supermetrics was the best fit for us.”

From marketing reporting to data modeling that delivers better insights

When dentsu first used Supermetrics in 2019, they started with Supermetrics API. They were able to fetch data from different APIs to their data hub on the Google Cloud Platform. Then, they pushed the daily marketing data directly to clients’ reports in Google Data Studio or PowerBI.

Supermetrics API helped the team significantly increase the dashboard output by 50 times and shortened the time it took to create a report for each client. Besides, they also saw an increase in revenue and customer satisfaction.

Fast forward to today, dentsu focuses more on data science activities to provide better insights to their clients. As a result, they brought in BigQuery as their data warehouse solution, and Supermetrics continued to be a trusted partner.

The dentsu team uses Supermetrics to bring cross-channel marketing data to Google BigQuery. Once the data is there, they can easily enrich the data in several ways, such as creating custom columns, adding exchange rates or currency, and extracting specific campaigns for their clients. After the data enrichment step, they will fetch the data to BI tools for data visualization.

Jarle says, “The data transfer was quite easy to set up. Once the data is in BigQuery, we could enrich the data in many ways, such as creating custom columns, adding a new exchange rate or a currency, or extracting a certain campaign. Then, our client teams can use this data to build Data Studio or Power BI reports for the clients.”

Additionally, having the data ready in BigQuery enables them to do data modeling at a lower cost. They’re doing different data modeling for their clients, such as marketing mix modeling and sales modeling. Jarle says, “One thing that changed since we started is that we do more data modeling for our clients now. Supermetrics helps us move marketing data into Google BigQuery and enables us to do marketing mix modeling at a lower cost and less development time.”

Less time, less cost, and more valuable insights for clients

dentsu can now deliver better and faster solutions to their clients. This is also one of their biggest selling points. Jarle says, “I think one of the biggest selling points in several client meetings now is that the clients can look at the dashboards, and if they want to change something, we’re able to make the changes on the fly.”

Jarle continues, “In the early days, it took a specialist a week to make those kinds of reports. Now we can easily customize and adapt the reports in the meetings. The clients don’t have to wait very long for the results anymore, which is a wow factor. Solving these tasks with the clients also helps us nurture a better relationship with them.”

Besides the marketing reporting benefits, the new data infrastructure also allows the dentsu team to bring data science benefits to their clients, especially for big corporations looking to do sales and marketing mix modeling. Having an automated data infrastructure in place means less development time and costs.

Jarle explains, “Typically if a big corporation wants to do sales or marketing mix modeling, the process can take up one year. Of course, they may have some historical data already but bringing all data together manually and building the infrastructure is very time-consuming and expensive. It can easily cost €50k-€150k. However, if we automate the data transfer, it’ll take less time, less manpower, and less cost to do data modeling.”

As for the next step, dentsu is looking to put the data back into buying systems so they can automate their paid ad buying. The focus is not just on reporting or modeling but also on automating the buying process.

Want to get rid of data chaos and build better reports and data models like the dentsu team? Start your free trial of Supermetrics for BigQuery, or contact our sales team to discuss more.