Nobody wanted it

Nobody wanted my product at first.

Not founders. Not friends. Not my LinkedIn network. Not the people who owed me favors and not the people who had every reason to be polite. Nobody.

And I want to be clear about who was hearing all those nos. I have been building companies for almost 30 years. Perot Systems, Microsoft Consulting, then two decades running Henson Group. I had a network most first-time founders would trade a kidney for. And SocialPost.ai still got the same polite response from everyone: cool, but I wouldn't use it.

Cool, but.

If experience bought you immunity from early rejection, I had paid for the premium plan. It bought nothing. That is the first lesson buried in this story: the market does not check your resume before it shrugs at you.

What "cool, but" actually sounds like up close

Let me describe the texture of it, because if you are early, you are about to hear it too.

Nobody tells you your product is bad. That would be useful. What they say is some version of "this is really impressive" followed by a reason it is not for them. They already have a system. They will check it out later. They will send it to someone who needs it more.

The demos ended warmly. The follow-ups went quiet. And the worst part was the source: these were founders and operators, the exact people I built it for. If they didn't want it, the comfortable interpretations run out fast.

Most founders hear that for a few months and pivot. Some quit. I did neither, and I ignored it. But "ignored" is the wrong word for what I actually did, and the difference is the entire point of this post.

The signal buried inside every no

I stopped treating the nos as verdicts and started treating them as data. Because every no had a signal buried in it, and once I started writing them down, the signals sorted into three piles:

Onboarding was confusing. People who agreed to try it stalled before reaching value. That is not "I don't want this." That is "I couldn't find the thing I might want."

The value prop wasn't landing. People kept classifying us as another AI writing tool, a market they were already tired of. The real product, the operational layer that takes the caption your AI already wrote and turns it into brand-correct visuals, formatted and scheduled across platforms, was invisible in my own pitch.

Output quality needed work. Some rejections were honest quality feedback wearing a polite costume. Those went straight to the product backlog.

Notice what is not in any of those piles: "nobody has this problem." Not once. Every founder I talked to was drowning in the work of showing up consistently online without a marketing hire. The pain was confirmed in the same conversations where the product was declined.

That gap, confirmed pain plus declined product, is the most important diagnostic in early-stage building. It means the problem is on your side of the table. Which is the good news, because your side of the table is the only side you control.

I wrote it all down. Kept building. For two years, out of my own pocket, more than $1M before a single outside dollar came in.

I had seen this movie before, in 2002

Here is why I could sit inside two years of rejection without flinching: I had started a company nobody asked for once before.

February 2002. New York City, a few months after 9/11. I left my job at Microsoft Consulting Services and founded Henson Group from a one-bedroom apartment. I was the first and only employee for nine months. The city's businesses were displaced, shaken, and in no mood to buy IT services from a one-man company with no track record.

So I didn't sell. I gave. Free IT migrations for displaced companies, done well, done fast, no invoice. Nobody "wanted" what I was offering then either, until they experienced it. The demand was real the whole time. The trust wasn't built yet. That company eventually became one of Microsoft's largest global resellers.

I am not telling that story to flex. I am telling it because it installed a conviction that carried me through the SocialPost.ai winter: early rejection measures your positioning and your proof, not your market. In 2002 the fix was free work that built proof. In 2023 the fix was rewriting the pitch and the onboarding until the value was impossible to misread. Different decades, same underlying mechanic.

Even so, I will admit the SocialPost.ai version was harder on the ego. In 2002 I was a nobody, and nobodies expect nos. In 2023 I was a veteran operator watching my own network pass on my product. Experience does not spare you that feeling. It just teaches you not to make decisions while you are feeling it.

What two years of nos bought

Every logged rejection became a work item. The onboarding complaints drove rebuild after rebuild of the first-session experience. The classification problem drove the positioning we run today: keep your AI, we handle everything after the caption. The quality complaints drove the product itself.

Now we've done hundreds of demos, most of them run by me personally. The product is sharp, and we brought on a VC partner, after the proof existed, on the strength of a real business rather than a story. The company is venture-backed and profitable, with more than 15,000 users. Demand is obvious today. Investors see it, users see it, the numbers show it.

At the start, nobody wanted it. Same product idea. Same founder. Same market. The only thing that changed in between was everything the nos taught me.

How to read rejection in your first 90 days

If you are in the "cool, but" phase right now, here is the discipline I would hand you. Run it for 90 days before you let yourself conclude anything:

Log every no in writing, verbatim where you can. Memory smooths rejection into mush. Text preserves the signal.

Sort each no into one of three buckets: couldn't reach the value (onboarding), didn't understand the value (positioning), or reached it and found it lacking (product). Force a choice. No "other" bucket.

Separately, track whether each person confirmed the PROBLEM, regardless of what they said about your product. Ask about their current workaround. Pain plus a duct-tape workaround is a live market.

Fix the biggest bucket first and re-test with new people. Never re-pitch the same people with the same pitch and call it iteration.

Only if the problem itself keeps failing to confirm, if people genuinely do not have the pain and are not working around it, do you get to say "bad market." That conclusion must be earned by the log, not by your bruised morale.

The lesson underneath it: early rejection is almost never a market problem. It is a messaging problem, sometimes an onboarding problem, occasionally a quality problem. All three are fixable by you, this quarter. A dead market is rare, and you will know it by the absence of pain, not the absence of yes.

Not "they said no." Why they said no. That is the whole craft.

Your turn: In your first 90 days, how did you tell bad idea apart from bad positioning, and what did it cost you to find out?