In a world of digital distractions, the CTO of Slack said he has to be intentional about carving out time for deep work.
“Sometimes my day is driven by my calendar or driven by my inbox or what’s coming at me,” Parker Harris, who also cofounded Salesforce, told Business Insider.
One strategy he uses to limit distractions when he turns to focused work: Keeping the sound of his notifications “completely off.”
“I hate that knock sound or any other sound,” Harris said about notifications in general. He said he also flips his phone over so he isn’t tempted by it.
“I don’t want to see my phone,” Harris said.
While the CTO said he normally keeps his notifications on, he doesn’t “jump straight” to addressing every individual message he receives, given he works in an “interrupt-driven” environment.
“We all need to find a way to concentrate,” he said.
The CTO said that when he’s doing deep work, he’ll go into a specific channel or company planning document and “immerse” himself in that. He said he’s able to focus when he intentionally sets aside that time to work on a specific task.
“Unless there’s some fire coming at me, like Slack has some issue, or Marc Benioff wants to talk to me,” Harris said, adding that in those situations, “I’ll drop what I’m doing.”
The company is rolling out a new AI version of its Slackbot tool on Tuesday. The CTO said the hope is that it will help users determine the most important task at any given moment.
Slackbot is a personal work agent, built directly into the platform. It will now do more, including surfacing relevant context; helping track notifications; and assisting users in prioritizing Slack activity, the company said.
Harris said the tool will help employees move from being “reactive to proactive” in the workplace.
“Slack is where work gets done. We’re going to continue to tackle the productivity challenge. We want to make employees more productive,” Harris said.
His best practices for using the new tool include being specific about context, audience, and outcome in prompts. He also suggested cross-referencing Slackbot with other data sources, providing feedback to the bot, and reviewing its output before sharing.
