• Zero copy data integrations

    Zero copy data integrations with data lakes and data warehouses. Integrations to popular platforms like databricks and snowflake, without the need to move or copy the data (like a window into the data), is very exciting; the idea of not having to physically move data, yet still fully utilise it.

  • Unstructured data and structured data

    Linking unstructured data like images, emails and PDFs with other data in Data Cloud is very impressive. As this has only just been released, I wasn’t sure how this would work, but in the demo they showed that you can virtually scan in a document and it could take out the rich text and allow you to utilise it alongside your structure data with columns and rows.

  • Einstein Look-alike Segments

    Use Einstein look-alike segments to find audiences similar to an existing one. For example, you create a segment, maybe its your most ideal customers using their purchase history data for example, this is called your seed segment. Then Einstein uses data to find audiences with similar behaviour, profile etc, ensuring that it is most similar to the seed segment. The lookalike model runs every 7 days for each seed segment. You specify your audience size then Einstein first looks for extremely similar matches. Then, if the audience size isn’t reached, Einstein checks for moderately similar matches and so on…