What I learned at Slack City Tour, Sydney

If you suffer from vertigo, then the elevator ride to the 39th floor of Salesforce Tower in Sydney might put you off. Don’t be. It’s worth the ear-popping experience for the most incredible views of the Harbour Bridge and the Opera House. And, in my case, by friendly Slack staff offering you coffee and chocolate brownies.

This was all part of the Slack City Tour, an event that mixes power users with potential clients. For an hour and a half, the local Slack leadership shared insights, ran demos of the latest features and interviewed happy Slack customers.

I typically don’t attend marketing events, but I wanted to see from the inside what Slack is saying about itself. To get a feel for how this brand, perhaps the most B2B brand in the world, retains its identity now that it’s part of the giant that is Salesforce. And see what it has to say about AI, of course.

I’m glad I took that elevator ride, and not just for the brownies. Here’s why.

AI is the fabric that brings together Salesforce’s acquisitions

Derek Laney, Technology Evangelist at Salesforce, shared an incisive slide illustrating how AI can be used across Slack. It was something to this effect:

Before AI came along, you could call in data from your CRM or pull a BI report into a Slack thread through a myriad of applications. With AI, Slack is connecting those applications and deriving insights in a compound manner.

That means an answer to a prompt might be generated based on data from multiple applications instead of one, with all of them intrinsically connected. It’s powerful stuff. And why stop there, when Tableau augments this information with added insights – your CRM already has a wealth of account-based data but, done right, Tableau can provide a much deeper view of your entire business.

Of course, that’s assuming you are already a customer of Salesforce and Tableau, which just makes Salesforce as a software provider much stickier with customers. It’s unclear to me whether what Slack refers to as Slack AI is built with Einstein, but regardless, I doubt the above would work with, say, HubSpot.

Which brings me to my next point.

The “everything app” of B2B

At another moment in the keynote, the speaker shared that Slack was launching a feature called “Lists” later this year. A feature that “brings project management into Slack”.

Now, there are already plenty of project management applications in Slack. The description for Trello’s app, for example, literally reads “Collaborate on Trello projects without leaving Slack”. Here’s a preview of project management in the Slack app store:

These applications are typically meant to make certain tasks convenient, while still having you rely on the parent app to do whatever you need to do. If I were Wrike or Asana, I’d be fine with certain tasks being done through Slack if I’m still charging that user a subscription fee. The same is true of many other software categories.

But as Slack develops more features that are the same as what you can find in other SaaS, I wonder how many Procurement Managers and CTOs will start looking to cut their enterprise licenses in favour of “just Slack”?

The irony is that many of those SaaS vendors have gained success because they’re simple to use and appeal to many verticals. They’ve then built communication and collaboration tools on top, as an effort to further improve them. The problem is that simple, universally loved features such as “lists” are somewhat simple to replicate – and persuading people to use your messaging app is much harder to achieve. So it’s easier for Slack to persuade Asana customers to come over than it is for Asana to do the opposite.

As it builds more native features that “aren’t just chat”, I start thinking of Slack more the way that I think of a browser. Not literally – it’s no Chrome competitor – but in the sense that the browser is a central hub of activity with characteristics that make it irreplaceable.

Slack wants to be the “browser” you go to do work. From there, a mixture of native features and third-party applications means you mostly don’t need to use much else. In this regard, it makes Salesforce more akin to Google or Microsoft, with more vertical integrations and an ecosystem designed to keep you within its walls.

Come to Slack, stay in Slack

There’s a corollary to this. Doing more work from within Slack is great for gathering more data. Gathering more data is, in turn, great, if your goal is providing great business insights to customers. It’s therefore in the interest of Salesforce to have as much work done from Slack as possible.

And that includes persuading you to do as little through email as possible.

Oh email. So 2000s. You don’t need to attend a Slack event to sense that the idea of doing work via email must sound like the most “uncool” thing to Slackers worldwide (is a fan of Slack a Slacker? I don’t know).

To some extent, I can understand why. Email is slow(ish). There’s spam. Slack is fast, there are channels, staff can create their own emojis. At the event, there was a reference to “a legacy way of working” for most companies, presumably in reference to those that don’t use Slack – or similar – to modernise their operations.

I get it. I’m not here to defend email as a form of doing work. And given how much work often must be done via email, there’s little chance it needs defending. But it’s not lost on me that there’s little chance of an AI being trained by reading emails. At least for now, that is. For now, it’s the one workspace that Slack can’t tap into when feeding the above flywheel of AI-driven business insights.

And while it probably doesn’t need it to still arrive at those juicy business insights, it’s still one bastion it wouldn’t mind conquering.

A final anecdote

One of the day’s most celebrated AI features was being able to summarise unread Slack messages and send recaps. It’s an amazing and helpful tool, I’m sure. But it reminds me of all those proposed solutions – those hyper-intelligent email clients that never seem to arrive – that promise to summarise your unread email.

Call me what you will, but if you don’t have enough time in the day to process all the information sent to you, then surely you’re overworked? It’s not your fault. You probably don’t have the staff you need above, below and next to you to address the core reasons why you’re being sent that information in the first place. And if the information you’re sent isn’t useful, then the channels that got it there are broken and need fixing.

Of course, being overworked is the reality that many people have to deal with. So Slack AI is a welcome helping hand. Just like a feature that summarises my inbox would be great.

The more things change…

Worth a read

• Workday Adaptive Planning and Consolidation – what you need to know
• How genius coworker Salesforce Einstein Copilot AI can help your business
• Oracle to expand cloud and AI footprint in Japan with $8 billion investment