Customer Data Platform: Is real-time a myth? Let’s untangle the true from fiction.

You’ve just completed a purchase in store. You leave the store, phone in hand, and a notification appears: an email from the brand offering you 10% off the item you just bought at full price. Frustrating, isn’t it? This is the consequence of unsynchronised marketing tools. In fact, for brands, the concept of a “real-time” CDP has become the ultimate promise to solve this type of problem. But behind this term, brandished by vendors, lies very different technical realities.  

Adrien Paul, Group Product Manager at imagino, lifts the veil on what’s really behind the notion of real time and explains how to rethink the approach to customer data. 

 
“Real-time should not be seen as a simple technical capability of the solution, but should be defined by specific, concrete use cases.”

How do you distinguish a true real-time Customer Engagement Platform from a solution that only imitates it? 

Today, there is real confusion in the market. Many solution providers use the “real-time” label when they handle “live data” (up-to-date data on a Cloud Platform such as Snowflake or BigQuery for example). But accessing up-to-date data does not mean that the system reacts instantly to an external event, it doesn’t react in real time. 

On top of that, the ability of a system to react in real time is not everything, it is also necessary to look at its ability to manage the flow of information. Most tools do simple logic: event -> reaction. A truly efficient solution, such as imagino, goes further: it masters the cycle of event -> enrichment (contextualisation) -> reaction. 

In concrete terms, the tool does not just trigger a message because an action has taken place. It captures the event in real time, queries the customer’s history to understand the context: 

• are they a premium customer?  

• is this their first purchase? 

and adapts the reaction instantly. 

 
“Live data is not the same as real-time data. Having access to an up-to-date database is one thing, reacting immediately to a customer event is another.” 

How can data architecture ensure a real-time CDP? 

In traditional data architecture, information spends its time travelling and being copied from one system to another. Each step of duplication, transfer, and synchronisation creates unavoidable latency that can take hours, or even days. The Zero Copy approach, on the other hand, is a game-changer. Instead of multiplying silos and copying data everywhere, the solution (like imagino) reads the data directly where it is, at its source (for example, within your Data Warehouse).  

You have instant access to information that is always fresh, minimising the discrepancies related to synchronisation and transfer times between the different technological bricks. By avoiding these unnecessary intermediate steps, you also protect yourself from the significant integration efforts that are needed to add missing or essential data to a new use case at a later date. In addition, the “Zero Copy” approach with Data Clouds (such as Snowflake and BigQuery) allows a direct reading of the data by the CEP, ensuring controlled and predictable costs thanks to the assignment of clear roles and responsibilities for both the Data Cloud and the CEP. 

 
“The Zero Copy approach makes it possible to avoid successive copies of the data in different solutions. The organisation can then rely on a single truth that is always up to date.”
 

Can you give us examples of marketing failures caused by lack of reactiveness in the data feedback?  

Latency problems are numerous and costly, both in terms of budget and brand image. A classic example is that of advertising retargeting (Display). You browse a site, you buy a pair of shoes on a Monday at 17.47. However, during the days that follow, you continue to see banner ads encouraging you to buy that same pair of shoes. The reason? Your purchase information wasn’t synced fast enough with advertising tools. The brand spends budget to show you an unnecessary ad, while also degrading your experience. 

A second example concerns consent management. A customer unsubscribes from your newsletter, but your system takes more than 48 hours to spread the information. In the meantime, they receive two new promotional emails. Today, consumers expect their unsubscribes to be taken into account immediately. A delay in updating this data generates real frustrations, and also complaints. 

Is real-time necessary for 100% of use cases? 

This is the biggest mistake – thinking real-time is needed everywhere. Real-time is neither necessary nor desirable for 100% of use cases. It is critical for transactional communications (order confirmation, password reset) or pure synchronisation (creating an account on the mobile application and expecting to be able to log into the website a second later). 

But for marketing actions, going too fast can actually cause problems. Let’s take basket abandonment. Reacting in real-time would mean sending a follow-up email the second the customer leaves the site. However, sometimes it is necessary to give them time to breathe. They may have gone to check a piece of information or compare a price and will come back later.  
“It is important to find the right balance. In some cases we need to react immediately, or sometimes we need to react the next day”. 
 

How does a real-time CEP optimise both data management and customer interactions? 

imagino has positioned itself as a Customer Engagement Platform based on a powerful CDP engine. Why? Because the CEP is the vital link between the data (live or real-time) and the activation. It supports the famous “contextualisation”, essential to not act blindly. Making the imagino choice means ensuring that the entire ecosystem (real-time and batch shipments) is centralised in the same place. If a customer calls customer service to say that they’ve not received their order confirmation, the agent should have an immediate overview, without juggling the database tool, the customer service history and the dispatch tool.