The "partner with me and we'll both win" pitch is everywhere in AI right now. Some are genuine. Many are a high-ticket sale dressed up as a partnership. These are the patterns I've come to watch for — opinions formed from my own experience and from comparing notes with others. None of these alone proves anything; a few of them stacked together is when I slow down.
The proof is all personal brand, none of it verifiable
Big income numbers in video titles, a confident on-camera presence, "here's how I made $25M." It's persuasive, and that's the point — it's produced to persuade. When I can't find a single client I could actually call, a real case study with a name attached, or a working business behind the personality, I treat the impressive numbers as marketing claims, not facts.
It's all about your leverage, with "or you pay me" as the floor
A real partnership is mutual. A red flag for me is when the entire qualifying conversation is about what I must bring to the table — money, connections, an audience, an existing agency — and the explicit fallback whenever I might not have "enough" is that I pay them to build me one. That structure puts all the risk and most of the cost on you.
"Partner if you have leverage, otherwise pay me" is a sales funnel, not a partnership. Know which one you're in.
The prep is one-directional
Being asked to watch a video before a call is normal. Being asked to invest hours into someone's content as the price of admission — before they've shown you anything verifiable — is a subtle power move. It conditions you to feel you've already invested, which makes you more likely to keep going. Your time is worth as much as theirs.
Urgency without paper
Five-figure decisions pushed on a live call, with urgency ("spots are limited," "the price goes up") and no written agreement, deserve a hard pause. Legitimate operators are comfortable putting the offer, the deliverables, and the terms in writing and letting you think. Manufactured urgency exists to stop you from doing exactly the diligence this site recommends.
Vague deliverables
"I'll help you build an agency just like mine" sounds great until you ask: build what, exactly? By when? With what support, what milestones, and what happens if it doesn't work? If the answers stay fuzzy while the price stays firm, that asymmetry is the flag.
How they handle friction
This is the big one, and it's why I value the small reversible test. Watch how a counterparty behaves when something goes wrong — a scheduling mix-up, a technical glitch, a pointed question. Do they communicate, take responsibility, and make it right? Or do they go quiet? How someone treats you in a small moment of friction is exactly how they'll treat you in a big one. In my own experience, that moment told me everything.
None of this means "never partner"
Plenty of people are building real things in AI and genuinely want collaborators. The point isn't paranoia — it's pattern recognition. Bring the same rigor you'd bring to any significant business decision: verify the claims, get it in writing, test small, and pay attention to how people behave when it's inconvenient. The good ones welcome it.
These are my opinions, informed by personal experience. For the specific account behind them, see My Experience and the dated emails on Proof.