As expected Genie is the common theme throughout every session and the messaging is consistently centred around “creating customer magic”. The only problem though is that I don’t believe in magic and want to understand how Genie actually works and how it is different to similar announcements in previous years.

In 2019, Salesforce announced its Cloud Information Model (CIM) with the aim to create a standardised data interoperability layer that effectively made it simpler to map various schemas into a common canonical data model. CIM came along with a data mapping solution that looks very similar to the data mapping solution that is in Genie now. In the following years, we then got introduced to CDP which found its place embedded in the Marketing stack of the platform. Having implemented CDP with our team at J4RVIS, we found that it was difficult to integrate CDP data back into core Salesforce to drive meaningful interactions, automation and customer insights.

Essentially Genie has the overarching goal of bringing data from disparate data sources together into a single view. Imagine you have data from Salesforce CRM (e.g. accounts, contacts, cases, opportunities etc.), your ERP (e.g. invoices, shipments etc.), POS (loyalty, transactions etc.) and more and want to bring this together so that you can look at a customer record and see all of these datasets related it. Once combined you can then use this data to drive automation (e.g. trigger a flow or a marketing journey), insights and feed AI models.

To make the above idea work, Genie is designed to allow you to connect data sources through a set of pre-built connectors. Such connectors range from Salesforce Cloud data, to Snowflake data (this was also announced as a new strategic partnership moving forward) to 3rd party data storage systems, to open APIs and MuleSoft connectors.

Once connected, data can then be ingested either via batch or real-time streaming. Once data is coming in, users can then prepare data through recipes and transforms, which by the way reminded me a lot of the CRM Analytics data preparation screens, and make it available through spaces to add additional access controls.

Through harmonization – the process of bringing all data sources together into a single profile, Salesforce creates what they refer to as the customer graph, a centralised view of customer data. To achieve this, users can use declarative mapping tools to map source systems to a common data model in Genie and to then define matching rules that associate various profiles with each other. My biggest question here was:

How and where is the data stored?

I have found out that it does not sit in the traditional Salesforce database (and hence doesn’t blow out your data storage there) and instead sits in an entirely new datastore that is built using lake architecture principle to support the processing of billions and trillions of rows of data. Genie then works away to crawl through the data to bring it together by consolidating customer identities from various sources into a single profile (think of contacts in Salesforce, Subscribers in Marketing Cloud and Profiles in Commerce Cloud).

Lastly, you can then use the data to drive insights and predictions and to act on the data through flows, journeys and APIs.

In my opinion, the real power comes off the back of the ability to act on the centralised data. What I found incredibly painful with CDP was the inability to use the data to drive automation and this is what is changing now. Essentially with Genie, you can trigger flows and journeys that drive business automation. This becomes incredibly powerful as the only way of achieving this today is to either sync the data into Salesforce or to invoke a process via API calls/platform events. As an example, imagine you have an external ERP system and you want to notify your sales team members every time their sold item has been shipped or paid, you can do this by connecting your ERP to Genie and to then leveraging the ERP data to trigger a flow that perhaps generates a notification.

There are still many open questions in particular about data security like encryption. From a hosting perspective, data is hosted in Hyperforce which essentially means that data residency is less of a concern. I am also curious how this will be licensed and who the target customer is as I can see benefits here for both large enterprises and mid-market customers.

Whilst I initially found it challenging to get deeper insights that go beyond the marketing slides I went to a session at the end of the day that had the entire Genie product team presenting on how it is architected. I will write a separate blog for that.

I hope this is useful – thanks for reading and I am looking forward to many comments and discussions.