A software product costs essentially nothing to copy. Once a company writes the code, distributing it to one user or ten million users involves the same marginal effort: close to zero. This is one of the most economically unusual properties of digital goods, and tech companies have spent decades figuring out how to make you forget it entirely.
The strategy they landed on is artificial scarcity, the deliberate construction of limits that have no technical basis. Waitlists, seat counts, usage caps, token budgets, and “limited” tiers are not engineering necessities. They are pricing theater. And they work extraordinarily well, for reasons that are worth understanding clearly. This connects to a broader pattern we’ve covered before: how tech companies build better tools for themselves than they sell to you, and artificial scarcity is often the mechanism that enforces that gap.
The Waitlist as Marketing Tool
When Google launched Gmail in 2004, it was invite-only. You needed to know someone who already had an account. This was framed as a capacity constraint, but Google was not a company short on servers. The real effect was social proof at massive scale. Having a Gmail address meant you were in. Invites were traded on eBay for real money. The manufactured scarcity did something free advertising never could: it made an email client feel prestigious.
AI companies have revived this tactic with remarkable precision. When OpenAI launched ChatGPT, Midjourney enforced Discord-only access, and Anthropic drip-fed Claude invites, each company created the same psychological effect. The waitlist signals that something is worth waiting for. It converts the absence of access into a form of desire. And it costs the company nothing to maintain, because the scarcity is entirely artificial.
This is distinct from genuine capacity constraints. When a new restaurant is fully booked, there are real tables to count. When a software company puts you on a waitlist, the constraint is a number in a database. The decision to increment that number is a business choice, not an infrastructure one.
Usage Caps and the Illusion of Exhaustion
The next layer of artificial scarcity is usage limits. You’ve seen this everywhere: ChatGPT’s free tier offers a certain number of GPT-4 messages before throttling you to a slower model. Midjourney counts image generations. Notion limits AI queries. GitHub Copilot caps monthly completions on free accounts.
These limits are real in the sense that computation does cost money. But the relationship between the cap and the actual cost is not linear, and the primary function of the cap is not cost recovery. It is behavioral shaping. The cap is designed to be felt, to create the precise moment of frustration that motivates an upgrade decision.
This is related to how tech giants use friction against you. The usage limit is friction made quantifiable. You can see exactly how close you are to the wall. That visibility is intentional. A limit you never noticed would never drive conversions.
The math often confirms the strategy. When you look at what the incremental compute cost actually is for one extra GPT-4 query versus what a monthly subscription costs, the margin is enormous. The cap is not protecting a thin-margin business. It is protecting a pricing model.
Seat Limits and the Enterprise Playbook
For B2B software, artificial scarcity takes a different form: the per-seat model. Licenses are sold by user count, even when the underlying software would work identically if deployed to twice as many people. Slack, Figma, Notion, Linear, and virtually every enterprise tool on the market uses this model.
The seat limit has an interesting double function. First, it creates a natural expansion revenue mechanism. As a company grows, its software bill grows automatically, without any change in what it’s getting. Second, it creates internal budget pressure that actually helps the vendor. If adding a new team member requires a purchasing conversation, that conversation keeps the software visible and organizationally embedded.
None of this reflects genuine scarcity. The software does not degrade with more users. The constraint is contractual, not technical. And as we’ve explored before, software subscriptions cost more than one-time purchases for exactly this kind of deliberate reason, and per-seat pricing is the mechanism that makes subscription revenue compound.
The NFT Experiment and What It Revealed
The most explicit attempt to manufacture digital scarcity was also the most instructive. NFTs tried to solve the fundamental problem directly: how do you make a digital file scarce when anyone can copy it? The answer the ecosystem landed on was a blockchain receipt, a certificate of ownership attached to something that remained infinitely duplicable.
At peak hype, this worked. People paid millions for JPEGs. The scarcity was entirely social and contractual, dependent on a community agreeing that the receipt mattered. When that community’s belief eroded, the value collapsed almost instantaneously. The underlying files still existed. Nothing changed except the shared fiction.
This is the naked version of what software companies do more quietly. The scarcity is always a shared fiction to some degree. But the more embedded the product is in your workflow, the more durable that fiction becomes. A usage cap on a tool you depend on feels very different from a cap on something you could walk away from.
Why It Works on People Who Know Better
Here’s the uncomfortable part: artificial scarcity works even on technically sophisticated users who understand exactly what’s happening. This is not a bug in the strategy. It is the strategy.
The psychological mechanisms involved, status signaling, loss aversion, and social proof, operate below the level of rational analysis. Knowing that a waitlist is manufactured does not eliminate the feeling of wanting to be past it. Understanding that a usage cap is a pricing tool does not make hitting it less annoying. Recognizing that per-seat licensing is contractual rather than technical does not prevent the budget conversation it creates.
This connects to something worth being clear-eyed about. Tech companies are extremely good at behavioral design, often better than users are at resisting it. The same companies that engineered your brain’s reward system are now engineering your perception of scarcity. The tools are different but the underlying discipline is the same: applied psychology at product scale.
The defense is not cynicism. It’s awareness. When you hit a usage cap, it is worth pausing to ask whether the limit reflects a genuine constraint or a conversion funnel. When a product launches invite-only, it is worth asking what that choice is doing for the company’s positioning. The answers will not always change your behavior, but they will change the terms on which you make decisions.
And that, in the end, is the most valuable thing you can have in a market designed to manufacture the feeling of value from nothing.
