From chaos to control: how online metrics will evolve in 2025

[shukla7789]

ANDBy 2025, two key trends will stand out in the world of online metrics and transform the way companies understand their customers: the end-to-end processing of customer data using advanced language models and the widespread adoption of marketing mix modeling (MMM).


With the help of Joan Miró , co-founder and general manager of Kraz , an agency specializing in MMM, we analyze these two points and their impact on Spanish eCommerce. This article is part of our “2025 Trends Special” , in which 28 industry experts preview everything that is coming in the world of marketing and eCommerce for the coming year.

Download it for free
Comprehensive data processing with language models
The advancement of language models , such as those from OpenAI , has made it easier to analyze large volumes of text from a variety of sources, such as social media, satisfaction surveys, reviews on websites, and calls to customer service. These tools not only classify and categorize comments, but also analyze sentiment , translate text , and generate summaries , adapting to specific business needs.

Organizations face the challenge of managing data from both fusion database sources (such as surveys and customer service) and external sources (blogs and social media). With language models, all this information can be integrated into single repositories , allowing common patterns and key themes to be identified. For example, in the automotive sector, these models have been used to classify feedback into categories such as design, comfort, or technology, helping companies focus their efforts on specific areas of improvement.

These capabilities are especially useful for companies with multiple departments interested in customer data, such as marketing, product development, or customer service. With these tools, teams can access specific insights that drive informed decisions and more effective strategies.