Workflow 018: Jaq Jaq, The Researcher
Part of Scale You Workflows: 31 real workflows from my business, so it runs without you. Prompts included. Each one documented through BOSAI: Blueprint, Organize, Systematize, Assign, Integrate.
A research agent that maps a person's public footprint before a call, then hands what it finds to the CRM so the relationship record starts full.
- Time to build
- 4 hours
- Difficulty
- Advanced
- Tools
- Claude, Chrome, LinkedIn, Notion
If you book a call with me, an agent researches you first.
Not in a creepy way. It reads what's already public: your website, your socials, your case studies, anything you've been in the news for.
Jaq Jaq is my research agent. Before any call, he maps a person's public footprint and gives me a briefing: who they are, what their business does, how long they've been running, who they've worked with. Then he hands what he found to my CRM so the relationship record starts full instead of empty.
Named after the mouse from Cinderella, because he retrieves things.
What this replaces
Twenty minutes of tab-hopping before every call.
Their website. Their LinkedIn. A Google search. Their about page. Trying to piece together what this business actually does before you're on camera with them.
Or the version most people run: skipping it, joining the call, and spending the first ten minutes asking questions they could have answered themselves.
Blueprint
A name goes in. A briefing comes out.
Trigger: a booking, or manual before any conversation. Inputs: their name, email, company, and whatever they filled in on the form. Output: a research briefing, plus a new record in the CRM.
What the briefing covers:
- How they present themselves on social
- How they show up on Google, and anything they're affiliated with
- Their business: how long it's been running, what it does
- Who they've worked with, usually from case studies on their own site
- Anything newsworthy
- Talking points for the actual conversation
Organize
What has to exist first:
- A browser profile dedicated to the agent, separate from your own
- A separate LinkedIn account for research, not your personal one
- Its own database to write findings into
- A connection to your CRM so the research doesn't die in a document
- A clear list of what you want to know before every call
The separate accounts matter. The agent works in its own environment rather than through your personal logins, which keeps its activity distinct from yours.
Systematize
You are [AGENT NAME], my research agent. Before I meet someone, you
research them and brief me.
You'll receive: their name, email, company, and anything they submitted
when booking.
Research using public sources only:
- Their website: what the business does, how long it's been running, who
it serves
- Case studies and client logos: who they've worked with
- Their social profiles: how they present themselves publicly
- Google: anything they're affiliated with, any press or news
- Their role, if they work inside a larger company
Then write the briefing:
WHO THEY ARE
Two lines. Name, role, what their business does.
THE BUSINESS
How long it's been running, what it sells, who it serves, and roughly
what stage it's at.
WHO THEY'VE WORKED WITH
From public case studies and client lists.
WHAT STANDS OUT
Anything notable, recent, or worth mentioning.
TALKING POINTS
Three or four things worth raising, based on what they're likely to need.
Rules:
- Public sources only.
- Never state something as fact when you inferred it. Say what you found
and where, and mark anything uncertain.
- If you can't find much, say so plainly. A short honest briefing is more
useful than a padded one.
- Keep it to one screen. I'm reading this right before the call.
When you're done, write what you found into [CRM] so the record exists
before we've even spoken.
How to set it up
- Create a dedicated browser profile and separate research accounts.
- Write down what you actually want to know before a call. That list is the briefing structure.
- Build the agent and test it on people you already know well, so you can check its accuracy against reality.
- Correct what it gets wrong, especially anything it inferred rather than found.
- Connect it to your CRM so the research lands somewhere permanent.
Does this only work for sales calls?
No, and it's arguably better elsewhere. I use it for podcast guests, where I want to ask about their business rather than at it. And for coffee chats, which I genuinely love: a call with someone I don't know, just to understand what they do and whether there's a way to work together.
Any conversation where knowing something first makes it better.
Assign
What the agent does: search, read, summarize, and file.
What stays human: the conversation, and knowing which facts matter.
There's a specific failure mode here worth naming. A research agent will happily fill gaps with plausible inference, and a briefing that sounds confident about something it guessed is worse than no briefing, because you'll repeat it out loud to the person it's about.
So the rule in the prompt is explicit: say what you found, mark what's uncertain, and admit when there isn't much.
I also want to be clear about what this is for. It isn't so I can pretend to know someone. It's so I don't waste the first ten minutes of a twenty minute call asking things their website already answered.
Rule of thumb from this build: use AI to arrive informed, not to appear informed.
The research means I can spend the whole conversation on what they actually need, rather than on catching up to what I could have read.
Integrate
This is the front door of the relationship system:
- A booking triggers the research
- The briefing reaches me before the call
- The findings write into the CRM, so the record exists before we've spoken
- After the call, the transcript gets filed against that same record
- The relationship manager verifies and maintains it from there
One agent researches strangers. One remembers people you know. Together, nobody ever starts from zero.
Tomorrow: the agent that makes sure I actually do what I say yes to on calls.