When people hear "AI-written message," they usually picture the same thing: a note that opens with "I hope this finds you well," drops the recipient's company name in a way that feels like a mail-merge, and ends with a soft ask that could have been sent to anyone. That skepticism is earned. That kind of outreach can be worse than no outreach at all, because it signals that you did not actually think about the person. It makes warmth feel manufactured.
But that is a description of AI outreach done badly, not a description of what AI-assisted outreach has to be. The gap between a hollow message and a useful one comes down almost entirely to context.
Why Generic AI Messages Are Generic
Generic outputs are not the technology's fault. They are the result of generic inputs. If you give a language model a name, a company, and a vague prompt, you get a vague-prompt-quality result. The output reflects what you put in. When founders are disappointed by AI-drafted messages, they are usually disappointed by their own level of context, not by the drafting capability itself.
Good outreach has always required knowing something specific about the person: what they are focused on right now, what changed recently in their world, what you talked about the last time you were in the same room, what you have been thinking about that might actually be relevant to them. When you have those things, a message almost writes itself. The AI, in that scenario, is helping you express what you already know in a way that sounds natural and like you.
The Three Layers of Context That Matter
Useful context for a check-in operates at three levels. The first is relationship context: how you know each other, when you last spoke, what you talked about, and what came of it. The second is situational context: what has changed for them recently, what they are currently working on, any news in their professional life that gives you a natural opening. The third is voice context: how you actually write, what your sentence rhythm sounds like, what level of warmth feels right with this specific person.
All three layers matter. A message that gets the situational hook right but sounds like it was written by someone else will still feel off. A message that sounds exactly like you but contains nothing specific about the other person will feel like a mass send. The combination of all three is what makes a reconnect message feel like it came from a real person who was genuinely thinking about the recipient.
How Vinly Approaches Drafting
When we built the draft generation feature, we were not trying to write messages for people. We were trying to write a first draft that a person would actually want to send, which is a harder and more specific target. A draft that needs to be rewritten from scratch is not useful. A draft that captures the right tone, uses the right hook, and gets you 80 percent of the way there is genuinely valuable.
Vinly's drafts are built from the relationship record it maintains on each contact: notes from past interactions, recent signals from their professional world, and context you have added over time. The draft is not a template with a name filled in. It is assembled from what the system actually knows about that specific relationship. You edit it before sending. Sometimes that means changing one phrase, sometimes more. But you start from something that sounds like you rather than from a blank page.
The Edit Step Is Not Optional
We are direct with everyone who uses Vinly: the drafts are starting points, not finished messages. The edit is not a formality. Reading the draft, adjusting the sentence that sounds slightly off, adding the specific detail only you would know, is the step that turns a good draft into a message that feels unmistakably from you. People who skip the edit tend to get fewer replies, and they usually recognize why when they read back what they sent.
The editing step is where your voice takes over. Think of it the way you might think of a rough sketch from a good collaborator: the structure is right, the proportions are good, and you are now adding the details that make it specifically yours. Thirty seconds of editing is not a failure of the technology. It is the step that makes the message actually yours.
The Problem We Are Trying to Solve
The real issue is not that founders lack things to say. It is the blank-page problem. Most people know roughly what they want to convey in a check-in. They just do not start the message, because starting is the hardest part. By the time they have thought about what to write, worried about whether it sounds too transactional, and navigated through their email, they have already moved on to something else and the moment is gone.
A draft that lands in front of you, already shaped around the right context and the right tone, removes the starting friction. You are not deciding what to say from zero. You are editing something that is already close to what you would have written anyway. That difference in friction is the difference between a reconnect that actually happens and one that stays on the mental to-do list for another month.
What This Cannot Do
AI drafts cannot create warmth that does not exist. If you have not genuinely thought about someone in two years and have nothing real to offer, no amount of contextual drafting will make the message feel warm, because the warmth is simply not there to express. The best version of AI-assisted outreach amplifies genuine interest. It makes expressing that interest faster and easier. It does not substitute for it.
If a draft feels hollow when you read it back, that is a signal worth paying attention to. It might mean the relationship needs more real investment before a check-in makes sense. It might mean the timing is off and there is no natural hook right now. Either way, the honest answer is to wait rather than to send something you would not be glad to receive yourself.