ETL & Cloud Unique: Customized Connection and Hosting

Scal-e has just been listed in two reports by the Forrester International Research Institute : “Now Tech: Customer Data Platforms, Q4 2021”  Asia Pacific (APAC) report, and “Now Tech: Customer Data Platforms, Q1 2022” world report. The marketing & technology French Blog Martech.cloud took advantage of the launch of the new Scal-e website to organise an interview.  

Scal-e´s CEO Christophe ALVES spoke with the Martech.cloud about the company since its takeover in 2018, he discussed Scal-e´s strategy, the evolution of the market, its team and his vision of the company’s future.

This article is the first of a series of 5 entries published throughout 2022. Find the links to all the articles at the end of the page at the end of the summer.

Martech.cloud: Thank you, I understand your approach much better now.

Christophe Alves :The Gartner Institute pointed out, that the market is already looking for 3rd generation tools and that gathering and unifying data it is no longer enough. It is now necessary to be able to identify the origins of data and be able to exercise consumer rights (to forget, erase, etc.), anywhere and anytime, and be able to process the customer’s data for customers (statistics, marketing devices or services).  

This accentuates, the need to have flexible tools that are easily configured, instead of amassing lots of different technologies to recreate what a CDP offers. To make data available to both marketing teams and external tools, it must be possible to collect, process, unify or de-duplicate the data. For this purpose, we have set up an ETL natively connected to our CDP.

Martech.cloud : When you talk about an ETL, do you mean a tool like Zapier, Segment, Integromat, Leadsbridge or Mulesoft ?  

Christophe Alves : Yes, an ETL is intended to facilitate the collection and sharing of data. The difference is that our in-house ETL is no- code and offers connectors to nearly all the solutions you mentioned with other functionalities, like the possibility of making requests to external databases, and an API dedicated to our CDP, etc…

Martech.cloud : Ok, so like these ETLs or IPAAS on the market, your CDP has implemented different standard connectors?  

Christophe Alves : Yes indeed, I invite you to look at the non-exhaustive list on our new website! We take advantage of each project to identify new ways to connect and develop standard and custom connectors. 

Martech.cloud : What do you mean by custom connector? 

Christophe Alves : So, when we talk about a standard connector, we mean a connector that we can connect with in a few minutes. Just like a user who needs a login and a password to enter and fill in the preferences for using a solution. 

A lot of tools allow you to modify their data model, for example, most of the CRMs on the market allow you to add objects or fields to the initial data model. The problem is that via their API, these CRMs do not always allow you to recover data from objects or fields added afterwards, therefore we call our API at Scal-e a polymorphic API. 

Martech.cloud : What do you mean by polymorphic API? 

Christophe Alves : Our CDP allows the implementation of customized B2B, B2C or even B2B2C data models, as we wanted to make work easier for the business teams. As soon as an object or field is configured in the CDP without any intervention, it will be accessible in read or write format via our API. Hence the idea of polymorphism as, the API adapts to the data model implemented in the CDP. 

We have already had requests to add additional fields or calculated fields to the customer data. An example is to calculate the age from their date of birth. The idea is that, as soon as we add a field to map or calculate data, this field can be used on the platform to create segmentations, scores, personalize content or recommend products.

Martech.cloud : Okay, so the approach is for you to create tailor-made connectors that will be implemented or adapted during the project.

Christophe Alves : Exactly, as many other software editors announce standard connectors and yet they indicate that it will take X days to implement it.  

However, if the system does not allow you to easily recover all the data you want, you must then adapt your connector or find and use another technology to recover this information.

Martech.cloud : You mean in addition to your connectors? 

Christophe Alves : Yes exactly. Sometimes to connect with a single data source, you must use a connector and be able to make database queries. 

Martech.cloud : Do you have an example in mind? 

Christophe Alves : Not for a database query using a connector. However for Shopify, if a brand wishes to collect behavioural data such as page views or product views, this type of data cannot be retrieved via the Shopify API.  So all tools like Scal-e that have a Shopify connector will use their connector to retrieve contact and transactional information, and will have to use another connector to a tagging tool like Piwikpro , Mattomo or Google Analytics to retrieve behavioural data.  

So, for a merchant website, you will need 2 different tools. This is one of the reasons we wanted to develop an ETL natively integrated into our CDP. An ETL that offers different options as standard, such as the polymorphic API, connectors, the ability to query external databases or even import CSV streams to facilitate the collection of data to be mapped in the CDP.

Martech.cloud : Isn’t this the case for all CDPs of the market? Do they not all offer an integrated ETL and this kind of flexibility?  

Christophe Alves : No, a lot of CDPs on the one hand come “off the shelf”. They are ready to use but for only one type of use and we see a lot of CDPs oriented towards e-commerce. Basically, their model allows them to collect behavioural or transactional data from e-commerce platforms like Shopify. These are good solutions, the problem is if a brand wishes to aggregate more customer knowledge by coupling their CDP with other tools, for example, a POS tool, or if the brand decides to change from Shopify to Prestashop, Magento or any other similar tools. It will be at best very complicated, requiring a lot of development, and at worst a reason to change CDP. And I’ll tell you even more : most of the time, they will push to use another ETL or IPAAS as they have not developed their own ETL. The risk will be the cost of stacking different technologies.  

Likewise, few CDPs today natively integrate their own privacy policy management module and tend to offer other complementary solutions.

Martech.cloud : So, with a Scal-e CDP you can’t use a solution like Didomi, Axeptio or OneTrust?  

Christophe Alves : No, it’s not like that. 

Our philosophy is to not tell our customers to change tools, especially if they are satisfied with their solutions in place. As solutions like Didomi or Axeptio are very good and answer to other use cases. If a brand has X websites and has an offline activity (store for example) it will anyway need to connect its CMS like Axeptio, Didomi to a CDP 

For One Trust it’s the same, even if their offer is a little more developed since they offer more than what a CMS offers. Our privacy module can be used in master or slave mode. If the customer has a solution like OneTrust, he will be able to benefit from our ability to recover his settings. Otherwise, he can benefit from our module. That often the problem is being able to set up a privacy policy and ensure all the tools of the IT system can respect it. This is where our natively integrated approach offers many advantages. Even if our privacy module is used in slave mode from a master tool developed internally or externally like OneTrust, our platform natively integrates with all these mechanisms (erasure, anonymization, blocking, portability, etc) because we also know how to manage them. 

Cloud History

Martech.cloud : Nice! You often talk about the concept of  “Single cloud” Does this mean that you are in competition with solutions like Snowflake or BigQuery or hyperscalers like AWS, Azure or OVH?   

Christophe Alves : Not at all. We have many customers, who use Snowflake, BigQuery, Databrics etc to create data lakes or even data marts. Our function will be to connect to these tools and retrieve the data that will allow teams to implement their marketing strategy independently. These solutions are very good, as shown by their success. However, Snowflake or Bigquery are better suited to an IT consultant than a marketing manager. 

Of the hyperscalers, you mentioned AWS, Azure and OVH and others. We are beginning to see sovereign clouds and there are a lot of other interesting companies dependent on the country, such as Alibaba cloud in China. At Scal-e, we use these companies to either host customer data or to host our platform.  

So, we are more complementary rather than competing! 

Martech.cloud : So, a unified database with all the brand’s data can be hosted by the customer (on-premises) or by a host of their choice? 

Christophe Alves : That’s right. 

Each customer will be associated with a data server (tenant isolation) for security reasons to avoid the risk of data leakage. This data server may reside with the customer, a host of their choice, or managed by the Scal-e team. 

Martech.cloud : Does the hosting choice impact the cost? Even more, what is the minimum budget for a CDP? 

Christophe Alves : If the hosting of the platform stays as we chose, and if we speak only about the location of the server that contains the unified data : this will not affect the price. We wanted to offer this service, so it does not become a deciding factor for the customer.  

Martech.cloud : That’s what I thought I saw on your site regarding the private cloud offered as standard.  

Christophe Alves : Exactly. 

Follow up for the rest of the interview next week!

Become a Scal-e Customer or Partner

If you are familiar with customer data management and acquisition and retention issues, and would like to become a Scal-e customer or partner, contact us!

Contact us