How to find thousands of high-intent leads without buying a list

The insight underneath this is the part I'd pull out: you're not finding more leads, you're finding leads with context already attached. "Saw your comment on X's post" works because the person already has the topic loaded in their head. A cold list has zero context, so the first line has to manufacture relevance from nothing, and most people can tell.

I've been seeing a version of this from the SEO side. A brand new domain gets shown for hundreds of relevant queries pretty quickly, Google understands the topic fine. What it won't do is rank high enough to get clicked, because nobody trusts an unknown source yet regardless of relevance. Feels like the same failure mode as the static database: matched on topic, missing the signal that makes someone actually act.

Question on the mechanics: if there's no LinkedIn account connection, where's the like/comment data actually being pulled from? Trying to understand whether this holds up as LinkedIn changes how much of that is publicly scrapable, or if there's a more durable data source behind it.

The 7-day window is the part people are arguing with in the replies and I think you're right to hold it. Decayed intent is exactly what makes a template obvious — the first line is still technically accurate, the person just no longer cares about the thing they said three weeks ago, and that gap is what reads as automation. Personalization at scale isn't really a writing problem, it's a recency problem.

Where I'd mark a boundary: I run LeadGrid, which builds local-business lists off Google Maps, so I'm coming at this from the segment your playbook doesn't reach. The owner of a three-van HVAC company isn't liking a deliverability post. He's often not on LinkedIn at all, and if he is it's a profile from 2019 he's never opened. Run the engagement play against trades and you get an empty set — not a small one, an empty one. So for anyone reading this and selling to local SMBs rather than SaaS, the input has to be something the business does exist in, which is usually the map, the reviews, and its own website.

The more interesting difference is what "fresh" is measuring. Your unit is a person, and it decays in days, because they change jobs and their attention moves on. Mine is a business, and it decays in months, on a completely different axis — the number goes dead, the site disappears, the owner's name changes. Same word, two refresh cadences, and I think people conflate them and then wonder why "fresh data" advice contradicts itself depending on who's giving it.

One measurable thing you could add to the A/B test you described: log how many of those 1,000 commenters already received a "saw your comment on X's post" email from someone else that same week. That's Gregory's channel-burn point, but as a number rather than a worry. If the overlap is high, the window isn't the variable that matters any more — the queue in front of you is.

I love the approach. Can you get numbers on actual performance?

Intent signals will almost always outperform static lists. The challenge is scaling personalization without making outreach feel automated.

Static Apollo/Instantly lists vs people who liked or commented on competitor posts in the last 7 days is a clean intent cut. Comment > like as signal, and using that specific post as the first line (“saw your comment on X…”) is the part most list tools skip.

No LinkedIn login for the pull is a real trust unlock for anyone who’s burned an account.

In the 3–5x reply lift, how much is the intent list vs that first-line personalization?

Interesting approach. The part that caught my attention was turning engagement into a lead signal instead of relying on databases. Curious, what part of the process still takes the most manual effort for you now finding the right posts, qualifying leads, or converting them?

How did you identify your first target users before scaling? I'm especially interested in how indie founders find the first group of people willing to try a new product.

The 7-day window is the detail most people will skim past, and it is the whole play. Intent decays fast: in my Henson Group days the rep who reached a prospect the same week they raised their hand won deals that a better pitch two months later could not. I would also cap how many commenters you pull per post, because when everyone scrapes the same viral thread, the 50th "saw your comment" email burns the channel for everyone.

IbexAI targets a valuable pain point by using AI to identify high-intent LinkedIn prospects, moving beyond static list generation. To improve conversion and trust, the platform should specifically define its intent signals, proactively address LinkedIn scraping risks, emphasize lead quality over quantity, and offer an interactive, data-driven preview on its landing page.

Static databases are basically graveyards at this point. You're spot on.

I learned this the hard way. Coming from a heavy engineering background, my first instinct was to just out-build the problem. I set up a massive n8n automation sequence running on my $0 split-brain architecture (local workstation tunneled via Cloudflare Zero-Trust), thinking sheer volume would solve my pipeline issues.

It didn't. Bad data + high-speed automation just means you hit the spam folder faster.

Pivoting to "Guerrilla Outbound"—hijacking high-intent threads and targeting the exact people engaging with my competitors—completely changed the game for my SaaS (Vitabase). Now, I realize intent is the only metric that actually matters.

I love the premise of IbexAI. Quick technical question: since it doesn't require a LinkedIn account connection to scrape those commenters, are you bypassing rate limits through a specific rotating proxy setup, or is it purely a frontend scraping play?

This is such a smart shift in mindset — going after people who've already signaled intent instead of cold-guessing an ICP. The point about a comment being a stronger signal than a like really stands out; it makes so much sense that engagement quality should shape how you rank/prioritize outreach.

Also really appreciate that it doesn't require LinkedIn account access — that's a real concern with a lot of these tools. Curious how you handle it when the same person keeps showing up across multiple competitor posts — do you treat that as an even hotter lead?

Thank you! This is where I am stuck right now. I’ve mainly been using Instagram because I am trying to get my app into college campuses so this will be helpful.

The engagement-signal approach is solid — someone who comments on a competitor's post about cold email deliverability is probably dealing with cold email deliverability.

One thing I'd add: the real bottleneck isn't finding these leads. It's following through at scale without the personalization degrading. You can pull 200 commenters off a viral LinkedIn post, but if your first line is a template with their name swapped in, you've wasted the intent signal.

The operators I've seen make this work treat each batch like a micro-campaign: same source post, same pain point, but the outreach connects their specific comment to a specific thing you can help with. Five emails that land > fifty that get archived.

Is this useful? Just find the people who have interest in other application or product?

The fact that it pulls intent data without needing an active LinkedIn account connection is the real game changer here. Account bans from scrapers are such a headache. Definitely a solid approach compared to dry Apollo lists!

This lines up with what I've been seeing too, engagement signal beats firmographic fit almost every time. One thing I'm curious about: how do you handle personalization at scale once the list gets into the thousands? "Saw your comment on X's post" works great for the first line, but does the rest of the email still need a manual touch, or have you found a way to keep it feeling non-templated past line one?

Curious what counts as "high-intent" in your framework. I've found that signals like recent hiring for a specific role, or a company just raising a round, tend to correlate way better with actual buying intent than firmographic filters alone. What's been working for you?