You've got a meeting in ten minutes with someone you've never met, at a company you know almost nothing about. So you do what everyone does: skim their homepage in the car park, catch three buzzwords, and walk in hoping nobody asks a pointed question.

There's a much better version of those five minutes.

Give an AI tool the company and the person, and you can walk in sounding like you spent the afternoon researching them. Not with generic flattery, with specific, relevant observations that build rapport and sharp questions that move the conversation somewhere useful.

Here's how.

 

Why This Beats a Quick Website Skim

A homepage skim tells you what a company says about itself. It doesn't tell you what matters to the person in front of you, what they're likely worried about, or what to actually ask.

An AI tool with web search can pull together the public picture - recent news, the person's role and background, the company's priorities - and turn it into something you can use in a conversation rather than just recite. The difference between "I saw you do marketing software" and "I noticed you've just expanded into the ANZ market, how's that changing your go-to-market?" is the difference between a vendor and someone worth talking to.

It also settles the nerves. Walking in prepared changes how you show up, and that shows.

 

What You Need First

  • The person's name and role, and the company website URL.
  • Their LinkedIn profile URL, if you have it.
  • Claude or ChatGPT, ideally with web search turned on. No web search on your plan? You can paste their About page text and LinkedIn text in manually instead - it still works, it just relies on what you give it rather than live results.

 

A Quick Note on Privacy

You're putting a real person's name and professional details into an AI tool, so a couple of sensible habits.

Stick to public, professional information - their company site, their public LinkedIn, published news. That's the same material anyone could look up, which keeps this firmly in fair territory. Don't paste in private data, notes from back-channel sources, or anything the person wouldn't expect to be public. And if you're on a free or personal AI account, be aware some train on submitted data by default; an enterprise or team account avoids that. For public-profile research this is low-risk, but the habit is worth keeping.

 

How to Set It Up

Step 1: Gather Your Inputs

Have the person's name and role, the company URL, and their LinkedIn URL ready to paste. Two minutes, tops.

 

Step 2: Run the Prompt

Paste this into Claude or ChatGPT, with web search on if you have it:

"You are my sales strategist. I'm about to meet [name, role] at [company website URL]. Here's their LinkedIn: [URL]. Give me a one-page brief: their likely top three priorities right now, two specific things about their business I can mention to build rapport, the objection most likely to come up about working with us, and five sharp questions I should ask."

No web search? Swap the URLs for pasted text: their About page, their LinkedIn summary, any recent announcement you can find. The prompt works the same way on text you provide.

 

Step 3: Read It With a Critical Eye

The brief will be good, but treat it as a starting point, not gospel. AI with web search can occasionally pull outdated or mismatched information, especially for people with common names or companies with a light web presence. Sanity-check the specific claims before you repeat them in the room - nothing undoes rapport faster than confidently referencing something that isn't true.

Keep what's sharp, ditch what's generic, and make sure the two rapport points are things you'd genuinely be comfortable raising.

 

Step 4: Walk In With Three Things

You don't need to memorise the whole brief. Take three things in with you: one priority you can speak to, one specific rapport point, and the one question you most want answered. That's enough to change the entire tone of the meeting.

 

Where This Fits Alongside Your CRM

This hack is for the meetings where you don't have history yet - first meetings, cold-ish prospects, new stakeholders at an existing account. It's built on public research, not CRM data.

Once someone's an active deal in HubSpot, you've got a different and better source: the notes, stages, and history on the deal record. That's a separate prep move. Think of it this way - public research for the first meeting, CRM data for every meeting after. Use the right one for where you are.

The teams that do this consistently tend to be the ones who log what they learn straight back into HubSpot afterwards, so the second meeting starts from real history instead of another car-park skim.

 

Want AI Working Across Your Whole Sales Process?

A five-minute brief is a great habit. Wiring AI into your sales process so preparation, follow-up, and CRM hygiene happen without anyone remembering to do them is a bigger win.

Neighbourhood is a Diamond HubSpot Partner and AI services team. We build AI into HubSpot workflows to handle the repetitive work. If your team is winging its prep, let's fix that properly.

Talk to us about AI for your sales process.

 

Get More Hacks Like This

Ops Intel is our newsletter for teams putting AI to work in their CRM - practical, no hype, no filler, straight to your inbox. Subscribe to Ops Intel.

We're also on Facebook and YouTube breaking down hacks like this one. Come find us there.

Happy optimising!