I don't use LinkedIn as the search tool anymore. I connect Claude and Gemini to large databases and Apollo, give them what we know about the business, its conversations and the people it wants to work with, and AI searches for those exact profiles. We check them, enrich them with LinkedIn data, then build the LinkedIn audience from those URLs.
I don't use LinkedIn as the search tool anymore. We connect Claude and Gemini to large industry databases and to Apollo. We tell them everything we know about the business, what its conversations have taught us and who it wants to work with. AI then searches for those exact profiles. We check them, enrich them with LinkedIn data, and build the LinkedIn audience from those profile URLs. We call it sniper targeting.
I was walking a client in the health tech space through this on a recent check-in call. They had been watching our work closely, so I took them through the whole process step by step. Here is how it fits together.
Why not just use LinkedIn's own search?
That is how we started. Our first process used LinkedIn as the search tool. We picked filters like keywords, industries, geography and a bunch of other things. The aim was to make the audience as close as it could be to a possible client. That audience then went into our automation.
It works. But the targeting is only ever as good as LinkedIn's filters. As far as AI has come right now, there was no reason to stop there.
What happens once the audience is built?
The outreach itself has a simple rhythm. LinkedIn allows us to send around 800 connection requests a month. We send them at intervals, so LinkedIn does not see it as a problem.
As soon as someone accepts, they get the next message after an hour. In between, our automation likes one of their posts and views their profile. It feels like one profile dealing with another, not something extremely automated.
We also try to keep regular human behaviour. That means withdrawing some requests and sending them again later. If you have 500 connection requests pending, LinkedIn starts asking who this person is and why they are sending so many.
If anyone replies at any point, the sequence breaks itself and we come in manually. That part is not AI. It is not reading messages and replying in real time. It is just an if and then formula: if someone replies, cut the sequence. If the reply is just thanks for connecting, we resume it. If there is a specific question or any sign of interest, we try to get them to book a call.
How does AI find the right people instead?
This is where we go one step backward. For this client, we looked into some of the largest health tech and health sector databases in the US, along with Apollo.
We connected Claude and Gemini to those databases. We use them to understand everything about the company we are doing this for. They learn from conversations we have had, and from the conversations we are having on LinkedIn right now. Then we tell them what kind of people the business is working with now, or would like to work with.
From all of this, it goes to Apollo, where AI searches for those exact profiles and sends them back to us. We review them. Once we feel they are okay, we enrich those profiles with LinkedIn data. Then we take all of those LinkedIn URLs, create an audience on LinkedIn from them, and put it into our automation tool again.
So rather than using just LinkedIn as a search tool, we use AI and whatever else is available on the internet first. The outreach sequence stays the same. What changes is who goes into it.
Why does the targeting matter so much?
Because messages only go to people who accept the connection request. Everything after that depends on who you asked in the first place.
That is the whole point of sniper targeting. As we get better at understanding who the target audience really is, the rate at which connection requests are accepted starts to go up. And so do the messages. Better targeting feeds every step after it.
What if there is one company you really want to work with?
Sometimes it gets even narrower than a database search. If a client tells me there is one company they really want to work with, we just go manually and find whatever we need to. We find the right person's contact details and start contacting them.
AI does not replace that kind of focused effort. It sits alongside it.
Will this process stay the same?
Probably not, and that is on purpose. Our team keeps learning every time we do this for someone. So we are always looking for newer tools. What is the newest technology that can get us even closer to the right audience?
At the end of the day, the goal is simple. Find prospective clients and get them on calls. Everything that happens beneath it, above it and around it, we are willing to do to make sure of that.
“Rather than using just LinkedIn as a search tool, we go one step backward to use AI and whatever else we have available on the Internet.”
Said on the call, 23 April 2026
Open this article in your AI assistant and ask how it applies to your business.
Add mananmehta.com as a preferred source and Google will show my posts higher in Top Stories and AI Mode for you.
I run MarketinCrew, a marketing agency of around sixty people across Mumbai, Lyon, Paris and Toronto, and AI Crew, where we set up AI and marketing systems for businesses. I work with founders as a Virtual CMO, teach Search Engine Marketing and Social Ads at YNOV Campus in Lyon, and host The Marketin Podcast. Before all of that I co-founded American Joint, a restaurant in Mumbai that grew to three outlets.
If your LinkedIn outreach starts and ends with LinkedIn's own filters, there is probably a better audience you are not reaching. My team at AI Crew can audit how you find and contact prospects and show you where AI fits.
Get your audit withaicrew.com