Here’s a pattern I keep seeing.
Someone opens a new tab to look something up. Twenty minutes later they’ve half-read four articles, clicked through six links, and have no idea how they got there. Work still unfinished.
It’s not a willpower problem. It’s a workflow problem.
The Fix: Drop It and Forget It
Here’s what I’ve been doing instead.
At the January investor workshop, I demoed connecting Perplexity to Slack for automated background research.
The setup is simple: you create a dedicated Slack channel — something like #research-queue. Connect Perplexity to it via Lindy or Zapier. Then, whenever you hit something during your work that needs a deeper look, you drop the topic into that channel and keep moving.
Perplexity picks it up, researches it, and posts a summary back to the channel.
You check when you’re ready. The brief is waiting.
No tab opened. No rabbit hole. No context switch. Your focus stayed where it was, and the research happened in the background.
Why This Works Better Than It Sounds
The problem with most research workflows is that they’re synchronous. You need information, so you stop and go get it. That stop is expensive. Research on attention shows that context switching can cost 20+ minutes of refocus time per interruption — even for short switches.
The Perplexity integration makes research asynchronous. You queue the need, the AI handles the retrieval, you review the output when you have natural downtime.
And Perplexity specifically is very good at this. It’s built for real-time research with sourced answers. This is what I think of as the multi-tool native idea: you route the work to the tool that’s best at it. Perplexity is great at real-time research. Route your research there. Don’t ask ChatGPT to do it badly when there’s a tool built for exactly this.
The Team That Stopped Doing Their Own Research
One of the teams I worked with had a problem that a lot of knowledge workers have: roughly a third of their day was going to research tasks that didn’t require human judgment. Finding background on a prospect. Looking up a competitor’s pricing. Pulling context on a market. Smart people doing work that a well-prompted AI could handle in seconds.
When we automated the intake — drop a topic, get a brief — their workday changed. The research still happened. They just stopped being the ones doing it.
They moved from reactive to strategic. Same information flowing in. Different relationship to it.
The Setup
Getting this running doesn’t require an engineering background. The basic version:
- Create a Slack channel (e.g., #research-queue)
- Connect it to Perplexity via Lindy or Zapier — both have Perplexity integrations
- Set the trigger: when a message is posted to the channel, send it to Perplexity as a research query, post the result back
- Start dropping topics in when you’re working
For a more customized version, you can add output formatting — ask Perplexity to structure results with a summary, key facts, and sources. Or set it up to tag you when results post so you know when to check.
The whole setup is under an hour. Perplexity has a free tier that works for this.
One Tweak This Week
There’s a framework I come back to often: one tweak a week. Not “overhaul your entire workflow.” Not “implement a 12-step AI system.” Just: do one more thing with AI this week than you did last week.
This is a good one-tweak-week candidate. It’s small to set up, the payoff is immediate, and it changes your daily experience of doing knowledge work.
The context switch might feel minor. But if you’re switching contexts to research 5-10 times a day, and each switch costs 20 minutes of refocus time… you’re looking at hours of lost focus time every single week. For a 45-minute setup.
The research rabbit hole is optional. Most people just don’t know they can opt out.
If you want to learn more about setting up AI workflows that run in the background while you work, the 4-Day AI Sprint covers this kind of automation hands-on.
