Validate before you build anything
5 min read · 993 words ·
A friend of mine — smart person, good at execution — spent three weeks building an automated content system for a niche he thought was underserved. Polished workflow, clean data structure, proper affiliate research. Really solid work.
Launched it to silence. Not even negative silence. Just nothing.
He'd built a machine to serve demand that didn't exist. The niche wasn't underserved. It was just unpopular. Nobody was searching for it, nobody was asking about it, nobody cared enough to complain about the lack of good content in the space. He'd confused "I find this interesting" with "the market is interested in this". It's an easy mistake to make when you're excited. It's also the most expensive one.
The principle that doesn't change
You build systems to scale proven demand. Not to create demand.
That sentence is worth reading twice. Proven demand. Not hopeful demand, not theoretical demand, not "I think people must care about this" demand. Demand you can point to and say: here, people are already asking for this, and what's currently available isn't good enough.
This is especially true when you're building automated content systems — because the economics of these things only work when the volume is justified. If you're building a machine that produces and publishes content at scale, the cost of being wrong isn't one bad post. It's months of output into a void.
So you validate first. Every time.
The 30-Minute Demand Test
The method I use takes about half an hour and uses three signals together. No single signal is enough on its own, but in combination they give you a clear picture.
Signal one: LinkedIn engagement in the target niche. Search the specific topic you're thinking about. Not the broad category — the specific problem. Are people posting about it? More importantly, are they getting comments that look like real frustration or real curiosity? Comments that say "I've been dealing with this exact thing" or "nobody talks about this enough" are worth ten times more than likes. They mean there's appetite for a conversation and not enough of it happening yet.
Signal two: Reddit activity. Find the relevant subreddit or thread. Look at what questions are being asked, how often, and whether the top answers are actually useful or just the same vague advice rehashed by accounts with low post history. If a question gets asked repeatedly and never gets a satisfying answer, that's a gap. That's your room to exist.
Signal three: Competitor quality check. Are there existing resources in this space? Good — that confirms the demand is real. Now ask: are those resources actually good? There's a difference between a niche being covered and a niche being well-served. If the dominant content is thin, academic in the wrong way, or clearly written by people who don't do the thing they're writing about, there's still space for someone who actually knows what they're talking about.
You can add Google Trends as a fourth check if you want confirmation that the topic is stable or growing rather than past its peak. But for most niches the three signals above are enough to make a decision.
What this looked like for Sales Source Code
The SSC gap showed up on LinkedIn before it showed up anywhere else.
I kept seeing enterprise sales professionals posting the same complaints — about forecasting that felt like astrology, about discovery conversations that didn't actually uncover anything, about the gap between what they were taught in training and what the job actually demanded. The frustration was real and specific.
What was available to them? Certification programmes that skewed academic and took months. YouTube-style closer content that was all tactics and no framework. Nothing in the middle — no rigorous, practitioner-led resource for people doing complex B2B sales at scale who wanted to actually understand the mechanics of what they were doing.
That's a gap. A clear one. Not a gap I invented because I wanted it to exist, but one I could point to in actual conversations happening in public.
I built SSC to fill that gap. Not the other way round.
What this looked like for Nunlimited itself
I didn't spend three weeks building Zo.E before checking if anyone cared about automation-first side businesses.
I looked at what was happening in the "build an income machine" space first. Entrepreneurs complaining on LinkedIn about the time cost of content. Reddit threads asking whether n8n was worth learning. A lot of enthusiasm, not much practical documentation from people who'd actually built something and were honest about what it cost and what it earned.
The demand was there. The rigour wasn't.
Then I built the system.
What to avoid
Two failure modes to watch for when you're doing the test.
No signal at all. If you search for your topic on LinkedIn and Reddit and find almost nothing — no posts, no threads, no questions — that's not a gap. That's an absence of interest. There are topics that people don't care enough about to even complain about, and an automated content machine pointed at one of them is just noise.
Established giants with no opening. If the top results in every channel are authoritative, well-maintained, and clearly trusted by a large audience, the question isn't whether the demand exists — it does — but whether you can offer something meaningfully different. Sometimes the answer is yes (a more specific angle, a different audience subset, a format that isn't currently being served). Sometimes it's a genuinely crowded space with no room. Know the difference before you commit six months to it.
The sweet spot is what I had with SSC: clear demand, clear frustration, existing content that was either too abstract or too shallow. Small creators with audiences, not dominant institutions with moats.
Half an hour. Three signals. You'll know.
Ta,
James
Founder | Nunlimited
