If you haven’t ever attended Salesforce World Tour, for 1 day per year its a bit like the travelling circus has taken over largest exhibition spaces you can imagine. Salesforces take a group of high energy staff and customers and has them make a frantic day of presentations. It repeats this in 8 places around the world. Thousands of Salesforce customers, admins, and eco system partners converge primarily to meet each other and learn what’s upcoming
There are star keynotes (Martina Navratilova in ’24), places where partners can sell you their wares, and this year about 86 sessions running in parallel over 1 day which are mix of case studies, demonstrations, workshops and hands on tutorials.
Working out what to see, whether you can fit it in, where each session is happening is a test of logistical prowess and not a small amount of patience. My method was to favourite the sessions on the computer a few days in advance, and then transfer to the phone app which was easier to use because it allows you to see your favourites chronologically and remove the clashes. It’s useful on the day too see what’s next and where you are rushing too. The app went offline at least once in the day, which left me relying on a paper printout of my expected agenda. This process should be much easier from people with as much tech and brainpower as Salesforce!
I think this was my fourth Salesforce Word Tour, and looking around, I was trying to work out who attends. For the external Salesforce eco system, it seems to be a place to meet existing customers and prospects. It’s not really a place you should go with a problem and hope to find a solution. For Salesforce admins, its a place to see some of the upcoming features, although care must be taken that the session titles meets your expectations and likely budget implications. An number of my chosen sessions didn’t!
So onto the main themes. This year, everything was again about Salesforce AI. Salesforce has a product called Data cloud. This allows external systems and the various Salesforce clouds to create a zero copy (doesn’t make another version of your external data) joined up version of your data. This has never been more important given the hunger of AI for good quality data to work from.
We saw an impressive case study from Aston Martin who are using Salesforce Einstein 1, where they had linked many systems from car telemetry to customers favourite communication channel as well as what looked like car manufacturing data. We saw examples of a customer coming to the likely end of their existing ownership, their browsing of the website of the new model, through to a WhatsApp invite to an exclusive event about the model, all the way to the customer asking mid build via WhatsApp whether they could still add some options. This was answered automatically and personally, and the customer happily added more features and spent more!
Nearly every session I saw talked about Copilot for your cloud. Unlike the consumer generative models we are all beginning to know, Salesforce has the trust layer. The data you submit for your model isn’t shared with others and so customer data doesn’t leak out of your organisation. The Copilots are AI assistants personalised to your specific company needs and come with standard actions and adding custom actions isn’t difficult. An action allows you to define a specific source record and a prompt that asks Copilot what you want to do with this.
We also saw an event management demonstration, where we listed all sessions with no speaker, and the Copilot was able to suggest speaker recommendations, and reasons why they were recommended. Useful example for some of our event clients presuming they have enough speaker data in their systems!
Staff from DPD logistics discussed that they have 40 departments getting various help from Salesforce AI, and that they store everything from assets to employee records in SF. This removes spreadsheets from their organisation and allows them to take away repetitive and tedious human tasks. They also use Copilot for thematic analysis to recommend a set of picklist values for new summary fields they are designing. They are starting to use the AI to write user stories and acceptance tests, allowing the technicians to concentrate on the parts of the development they prefer to do.
We heard lots of stats that most companies intend to use Enterprise AI in the near future and that Einstein 1 will allow companies to deliver this as quickly and trustworthy as possible.
At some points the show is a little more Ikea than World Fair. The logistics of handling this many visitors in this short of time must be an incredible challenge, but if you only do it once, I recommend you register your free ticket this time next year when the circus comes to town.
