Home Blog First Party Data: wh ...

First Party Data: why it is essential today, in the Cookieless Era


Companies own a great amount of First Party Data and too often they don’t exploit it in the right way. Only a structured Data Strategy allows you to understand their real value. In fact, they are really precious, since they enable companies to analyze and optimize their business, directing the most strategic decisions.

First Party Data is all the data owned by a company: online user behaviors, brand’s stores location, production times, performance of paid campaigns… A common mistake is thinking that the final goal is data collection itself, as a quantitative value: this is not quite true, since a data strategy is not just to collect it, but rather to interpret it in order to decide which business cases should be activated. In particular, data from users and marketing campaigns allow the best analysis and optimization of marketing performance.

Index of topics

  • First Party Data and Third Party Data: what changes today
  • Marketing Mix Models: the new performance evaluation
  • Customer Lifetime Value: from segmentation to action

First Party Data and Third Party Data: what changes today

Historically, digital marketing has made limited use of First Party data, relying more on Third Party Data instead (for example, ad platforms such as Google, Facebook, Criteo, etc.). This kind of data has always been in charge of measuring the performance of the various campaigns, describing and analyzing users. However, today the collection and sharing of data is becoming much more complex for these adtech players, due to increasingly restrictive privacy regulations and the progressive abandonment of Third Party Cookies and Mobile Advertising ID.

So what can we do today, considering that 3rdparty cookies and Mobile Advertising ID are disappearing according to the provisions on Privacy? The giants of the web are choosing a Cookieless perspective, thus the advertising based on user data – the so-called programmatic advertising – is undoubtedly the most affected channel, although it has experienced a growth of 6% compared to last year, reaching a global turnover of 588 million euro. Thus, the urge of companies to analyze user behavior internally and exploit the incredible power of combining data has become imperative: data from CRM and Surveys, transactional data and navigation data – which usually reside in separate silos – are a real “Treasure” of internal knowledge and they need to be shared accordingly.

Therefore, in this new era of digital advertising, companies must help themselves for two main activities:

  1. the measurement of campaign performance for advertising;
  2. the segmentation and analysis of its users to customize marketing strategies.

Basically, it is not enough to collect First Party Data, but you need to know how to interpret and activate it for very specific purposes. Of course, the collection phase is essential: the more accurate and relevant the data, the more personalized marketing actions, so the company will be able to predict the behavior of the public, both for a positive outcome (for example, knowing in advance which customers will proceed to a purchase) and a negative one (which customers will be lost or what the churn rate on a product / service will be).

This information allows companies to take a series of effective actions to improve performance and increase ROI.

Marketing Mix Models: the new performance evaluation

To define a good Data Strategy, it is necessary to identify what are the kinds of the First Party Data that we need to collect and organize: impressions, clicks and investments of the various digital marketing activities, GRP, marketing budget for campaigns and general turnover of the company (even offline).

These simple kinds of data, completely compliant with user’s privacy, allow companies to activate these new advertising performance evaluation models: the Marketing Mix Models (MMM).

MMMs base their effectiveness on econometric models, consolidated over the years and still used today in traditional marketing, combined with predictive models based on machine learning, such as Meta’s Prophet: thus, they support decision makers and marketing managers in correctly evaluating the contribution of every single marketing activity in the company’s global turnover. Moreover, they help companies make informed decisions about budget distribution and forecasting results, as well as suggesting the optimal budget. Today, the most mature open source framework of MMM is Facebook Robyn, preferred by the advertising world over the MultiTouch attribution methods, which need to follow the entire user journey.

Customer Lifetime Value: from segmentation to action

Above all, First Party Data helps companies in a fundamental and often underestimated activity: the retention or exploitation of the customer base, instead of always struggling to chase after new business. In fact, one of the most effective methods to improve business performance is to engage with those existing, most valuable customers.

How to spot them? The Customer Lifetime Value models are key, since they segment customers according to their behavior: the most used model today is the so-called RFM or Recency-Frequency-Monetary Value.

RFM identifies and segments users based on parameters such as:

  • revenues generated for the company (Monetary)
  • purchase frequency (Frequency)
  • how much time has passed since the last purchase (Recency).

By deciding how many user segments we want for each metric, the model will automatically determine the segmentation: for example, if we want to get 3 segments for each parameter, the system will divide our users into 27 different clusters (3x3x3). These can be further grouped and used for direct marketing and strategies through emails and SMS, for remarketing, as well as for navigation and advertising personalized activities; exploiting the potential of look-a-like models, to find users similar to our best customers.

The strategies can be customized for each segment, maximizing their effectiveness and making the most out of the data:

  • for more loyal users, who buy often or with high carts, you can set up rewarding activities, with discounts or dedicated events;
  • for users with medium-high carts, who buy infrequently, you can engage them by recommending the repurchase of their favorite products or even complementary products;
  • for users who have not purchased for a long time, but who often return to our site carrying out searches without buying, you should understand why, analyzing their searches: you may discover that a specific product is out-of-stock or decide to discount the object of their search.