Two inbound leads came into an agency I co-own. One of our guys looked at the second one, saw a long corporate email address he did not recognize, and called it "pretty sure spam."
Nobody asked my AI coworker to look at either of them. It was already in the Slack channel where the leads get discussed, so it went and checked who the names actually were.
The name on the "spam" address belonged to the EVP and Chief Sales and Marketing Officer of a publicly traded hospitality company. Appointed to the seat two months earlier. Same title at two other companies before that. If you are an agency, that is not a lead you delete on a hunch.
Then I Corrected It
I knew something it did not. The person is real, but that email address is not one his company uses. Which means the message was more likely someone using his name than the man himself.
So the score on this one is split. It was right that the name was a real executive and wrong that the lead was real. It researched the human and skipped the envelope.
I like that it took the correction without arguing and carried it forward instead of just apologizing. That is the part that makes it useful over months instead of minutes.
Why This Still Beats the Alternative
Our spam call was a coin flip on a name nobody recognized. Two minutes of research turned it into an actual decision with facts attached. That is true whichever way the decision goes.
Think about how leads really get triaged at a small agency. Somebody skims a form fill between calls, does not recognize the company, and moves on. Nobody Googles the name. Nobody checks whether the domain matches the company. The lead that looks weird and the lead that IS weird get the same treatment, which is none.
The second lead in that message was the boring one and it was the better prospect: a nonprofit crowdfunding platform that pays vet bills directly instead of routing money to the pet owner, 1,000+ animals helped, 1,000+ vet hospitals in the network. That is a donor-email program waiting to happen, and I would not have known any of it from a name and an email address.
The Two-Minute Check Anyone Can Copy
You do not need AI for this. You need someone to actually do it, every time, which is where AI wins.
- Search the person's full name plus their company. Press releases and appointment announcements are the fastest signal that the title is real.
- Check the email domain against the company's real domain. This is the step we skipped, and it is the one that decides whether the lead is a lead.
- Read what the company actually sells, not just what industry it is in. That is where the pitch comes from.
- Write the findings where the person handling the call will see them, not in a doc nobody opens.
- Note what you could NOT confirm. "Real person, unverified sender" is a much more useful handoff than "looks legit."
What I Changed After This
Inbound lead research is not something I ask for anymore, it just shows up in Slack with the name, the title, the company, what they sell, and an angle. My job is to read it and decide.
The honest limit: an AI can tell you who a name belongs to. It cannot tell you the person behind the address is that name. Domain checks, spoofing, someone borrowing a title to get a meeting, that is still a human judgment call, and this episode is exactly why I still make it.
Viktor is the AI coworker that ran this. It sits in my Slack, sees the same messages my team does, and can go research something without being asked. It also rebuilt this website and fact-checked a competitor's Instagram claims for me.
If you want to put it in your own inbox and lead flow, you can get $100 in free credits plus $50 off your first purchase here.
More real examples, screenshots and all: my full Viktor review.
