Inflation Stopped Being About Money When Tech Companies Started Pricing Like Utilities

The CPI Doesn't Measure What We Actually Buy Anymore

The Consumer Price Index tells us inflation is under control. Meanwhile, my SaaS subscriptions increased 40% last year, cloud hosting costs jumped without warning, and every API I depend on introduced new pricing tiers that somehow cost more while delivering less. The official numbers and the lived experience have diverged so completely that citing inflation statistics feels like gaslighting.

This isn't about conspiracy theories or government manipulation. The CPI basket was designed for an economy where you bought physical goods and occasional services. It wasn't built to track software licenses, API calls, cloud compute, or the subscription economy that now dominates professional expenses. When housing is 33% of the index but software infrastructure is barely represented, you're measuring the wrong economy.

The tech industry figured out how to inflate prices without calling it inflation. They just renamed it "pricing optimization" and "value-based pricing" and acted like annual 20-30% increases were normal market dynamics.

Tech Inflation Doesn't Show Up in Official Statistics

Traditional inflation happens when too much money chases too few goods. Tech inflation happens when switching costs get high enough that vendors can raise prices arbitrarily and customers have no choice but to pay. These are fundamentally different phenomena, but only one gets measured.

Consider what happened with cloud hosting over the past three years. AWS, Google Cloud, and Azure all raised prices on core services. Not because their costs increased—their economies of scale improved. They raised prices because they could, because migrating infrastructure is expensive enough that most companies will absorb a 15-20% increase rather than rebuild their entire deployment pipeline.

The same pattern plays out across enterprise software. Salesforce, Adobe, Atlassian—every mature SaaS company eventually discovers that existing customers are captive customers. The initial pricing was customer acquisition cost. The real business model is the annual price increases that come after you've built your workflows around their platform.

The Subscription Economy Made Inflation Invisible and Continuous

One-time purchases had natural price visibility. You knew exactly what something cost, and price increases were discrete events you could evaluate and resist. Subscriptions turned pricing into a continuous variable that vendors can adjust monthly, annually, or whenever they introduce a new "tier."

The genius of subscription pricing isn't recurring revenue—it's pricing opacity. When Slack moves features from the Standard plan to the Plus plan, is that a price increase or a product change? When GitHub introduces a new tier between Free and Team, are they raising prices or adding options? The answer is both, and the ambiguity is intentional.

This matters beyond individual companies. When every tool you use can adjust pricing continuously, inflation becomes ambient. There's no single moment to push back against, no clear price increase to reject. Just a gradual upward drift in your monthly burn rate that feels inevitable rather than chosen.

Developer Tools Are Where Inflation Gets Weaponized

API pricing deserves special attention because it represents inflation's purest form in tech. You build something that depends on an API, it becomes load-bearing infrastructure, and then the pricing changes. Not because the API got better or costs increased, but because the vendor realized you can't leave.

OpenAI's pricing has been relatively stable, but watch what happens as models become commoditized. The current race to the bottom on token pricing is temporary—once a few providers dominate, prices will stabilize and then creep upward. We've seen this exact pattern with cloud hosting, with CDNs, with every infrastructure service that started cheap and became essential.

The AI API market is currently in the customer acquisition phase. Companies are pricing aggressively to build dependency. Once that dependency is established, the real pricing begins. Anyone building on these APIs should understand they're getting introductory rates, not sustainable prices.

Inflation Expectations Changed When We Stopped Owning Software

The shift from perpetual licenses to subscriptions fundamentally changed how we think about software costs. With perpetual licenses, you paid once and could use the software indefinitely. Price increases meant choosing whether to upgrade. With subscriptions, price increases just happen, and your choice is pay or lose access to your own data.

This psychological shift matters more than the economic one. We've normalized the idea that software costs should increase annually, that maintaining access to the same functionality requires paying more each year. That's not inflation in the traditional sense—it's rent-seeking dressed up as service improvement.

The companies that successfully made this transition—Adobe being the canonical example—didn't just change their business model. They changed customer expectations about what software should cost. Now every B2B SaaS company follows the same playbook: hook customers with reasonable initial pricing, then increase 10-15% annually and call it "market alignment."

The Real Inflation Story Is What We Stopped Measuring

Official inflation statistics capture housing, food, energy—the necessities of personal life. They barely touch the necessities of professional life in 2024: cloud infrastructure, software tools, API access, data storage. For knowledge workers and tech companies, the latter category often exceeds the former.

This measurement gap means policy discussions about inflation completely miss the costs that matter most to the tech economy. When the Fed adjusts interest rates based on CPI data that doesn't include SaaS subscriptions or cloud hosting, they're optimizing for the wrong variables.

The companies experiencing real inflation pressure aren't the ones buying more consumer goods. They're the startups watching their AWS bills grow faster than their revenue, the agencies seeing design tool costs consume an increasing percentage of margin, the developers building on APIs that reprice annually.

What Actually Works Against Tech Inflation

Complaining doesn't work. Vendors know you're not leaving, and they've priced in customer complaints. What works is maintaining optionality: keeping switching costs low, avoiding lock-in, building abstraction layers that let you move between providers.

The companies that resist tech inflation most effectively are the ones that treat vendor relationships as temporary by default. They build infrastructure that can migrate between cloud providers, use open-source tools where possible, and maintain internal expertise rather than outsourcing everything to SaaS platforms.

This requires upfront investment that many companies skip. It's cheaper initially to go all-in on a single vendor's ecosystem. The costs come later, when that vendor realizes you're locked in and adjusts pricing accordingly. By then, the switching costs are high enough that you'll pay almost anything to avoid migration.

The Next Phase Gets Worse Before It Gets Better

AI tools are currently in their customer acquisition phase. Prices are low, capabilities are expanding, and everyone's racing to build dependency. This is the honeymoon period. What comes next is the same pattern we've seen with every infrastructure service: consolidation, then optimization, then price increases.

The difference with AI is the switching costs will be higher. When your entire product is built on GPT-4 or Claude, migrating isn't just technical work—it's retraining your system, adjusting your prompts, potentially rebuilding your entire user experience. The vendors understand this, and they're pricing accordingly for the long term, not the short term.

Anyone building on AI APIs right now should assume current pricing is temporary. The real cost structure will emerge once a few providers dominate and switching becomes prohibitively expensive. That's not cynicism—it's just how infrastructure markets work once they mature.