Cases 07.12.2022

Valio and Dagmar studied the anatomy of launches through modelling

Data-driven leadership Marketing development Marketing effectiveness Sales growth Marketing communications Marketing optimization Marketing research Data Science Marketing optimization Sales modelling

“How should Valio’s media investments develop in the coming years, taking into account the change in media consumption and aiming to improve the effectiveness of marketing?” This question inspired us to consider how more information could be obtained on the effectiveness of Valio’s measures without reviewing and modelling each product brand separately.

Valio is familiar with using data to support marketing decisions – and the importance of data is only growing at the same time as major changes take place in consumers’ purchasing and media behaviour. Building an effective encounter, and even effectively, will be increasingly demanding in the future.

When there are many measures and investments are significant, opportunities to increase efficiency should be sought where most of the investments are directed.

At Valio, like most grocery brands, various launches and product variants play a key role in marketing communications. Up to ten brands can launch new products each year.

We estimate that if the launch measures alone can be made even ten per cent more efficient, the savings will be significant. The saved investments can then be transferred, for example, to the long-term development of the brand. We were curious to experiment and learn when Dagmar suggested to us the possibility of participating in a pilot project.”

Outi Lindberg, Valio´s person responsible for consumer marketing, Head of Marketing

Analysis of about 20 brands and nearly 100 launches

Together with Dagmar, Valio carried out a pilot project to study the effects of the measures of nearly a hundred launches on sales. The aim was to find out the ROMI of launch advertising and understand how Valio’s launches typically work.

In addition, the aim was to find out whether, for example, there are any laws by product group that can be utilised in planning future launches and allocating investments.

The analysis included the sales of the launches of about 20 brands and the actions taken during the launches. To find out the connection between them, a statistical model was built.

Clear differences in the impact of the measures between different product groups and brands

The modelling provided Valio with valuable information on how different product groups and brands benefit from different launch measures and how effective the different measures are. The modelling was used to study, among other things, paid advertising measures, pricing and measures in stores, as well as other advertising in the same product group.

According to Outi Lindberg, the differences between different product groups and brands surprised the most in the results. Some product groups clearly benefited from paid advertising measures, while other product groups increased sales thanks to price reductions and an appropriate price point.

Within product groups, the differences between brands could be very large. To understand the differences, the modelling results were extracted in joint workshops and mirrored with other brand-specific data. This created a vision of how different actions should be assigned to each brand in launches.

We immediately applied the lessons learned from modelling. Dagmar has played a major role in interpreting the data and translating it into practical actions.

Outi Lindberg, Valio´s person responsible for consumer marketing, Head of Marketing

Dagmar designs new tools to increase the impact of marketing

In addition to the launch modelling done for Valio, Dagmar’s advanced analytics unit has designed many other new ways to determine the effectiveness of marketing. The most famous of them is ROMI modeling.

The unit has developed contribution modelling that can be used to verify the effectiveness of marketing in digital channels. In the future, contribution modelling will also take offline channels into account as data pipeline automation improves.

Piloting is an important part of our product development. We use it to test the models and algorithms we have developed to support our customers’ marketing in practice. After the first tests, we will continue the development work to ensure that the new tools meet our customers’ needs even better.

Ilkka Keskiväli, Director, Advanced Analytics, Dagmar

There are several tools for measuring the long-term impact of brand marketing, such as Brand Marketing Optimizer, Brand Impact Analysis and Brand Driver Analysis. They provide answers to questions such as:

  • How to optimally distribute brand and tactical marketing investments?
  • What impact does the brand have on sales?
  • How do marketing measures affect the brand and how much should be invested in maintaining the brand?
  • What brand attributes influence the main brand metric and how big is this impact?